Tungsten for Defence Applications: Why Western Procurement Officers Are Looking at Pakistan
China controls roughly 82% of global tungsten mine supply. That single number is why my inbox looks the way it does in 2025.
Three years ago, when I told procurement contacts in Europe we held tungsten-prospective ground in Gilgit-Baltistan, the reply was polite silence. Now it's a phone call within 48 hours. Something shifted — and it wasn't the geology. The geology's been there for 40 million years.
What changed is that a US Army program manager can no longer sign off on a kinetic penetrator contract without someone in the room asking where the ammonium paratungstate came from. And the honest answer, nine times out of ten, traces back to Jiangxi or Hunan.
Why tungsten is the awkward one
Antimony gets the headlines. Gallium and germanium get the export-control briefings. But tungsten is the metal that quietly sits inside almost every serious defence platform — kinetic energy penetrators, armour-piercing cores, rocket nozzles, radiation shielding, the counterweights in every rotary-wing aircraft you can name. There's no drop-in substitute. Depleted uranium works for some penetrator applications but comes with political baggage most NATO ministries won't touch anymore.
So the material is essential. The supply is concentrated. And the stockpiles held by the US Defense Logistics Agency are a fraction of what they were during the Cold War. I've seen the numbers quoted publicly — the DLA tungsten inventory is measured in hundreds of tonnes, not thousands. Annual US consumption is somewhere north of 8,000 tonnes.
Do the math. It's uncomfortable.
What Pakistan actually has
Here's the thing — Pakistan isn't a tungsten story anyone in London or Washington grew up reading about. The published resource base is thin compared to what's already known in Portugal, Bolivia or Vietnam. But the geology is right, and most of it has never been drilled to modern standards.
The Karakoram batholith and its contact aureoles host classic scheelite-bearing skarns. We see it in the field. Our teams have logged scheelite fluorescence under UV in multiple showings across our concession block — coarse crystalline scheelite in calc-silicate skarn, associated with granite intrusions of Cretaceous to Eocene age. Grades in grab samples from surface showings have run between 0.4% and 1.8% WO₃. That's not a resource statement — those are rock chips, and I'll say that plainly because serious buyers know the difference. But it's enough to justify the trenching and diamond drilling programmes we're scoping now.
There's also wolframite in the pegmatite fields we're mapping for lithium and tantalum. Two mineralisation styles, one district. That matters for a processing flowsheet down the line.
Compare that to the tungsten alternative to China conversation happening in Portugal (Panasqueira is a legacy operation running lower grades every year), Spain (Barruecopardo restarted but modest scale), South Korea (Sangdong ramping, but almost entirely spoken for), and Vietnam (Nui Phao — excellent asset, but the offtake book is full and the geopolitics aren't uncomplicated either).
That's the entire non-China tungsten supply universe of any consequence. It's a short list.
What procurement officers actually ask me
I've had maybe 20 serious conversations with Western buyers in the last year — defence primes, specialty metals traders, one sovereign strategic-reserve entity I won't name. The questions cluster around four things:
Can you deliver tungsten concentrate Pakistan-origin with a clean chain of custody? Yes, in principle. Concessions are held under valid GB mineral titles. Export licensing runs through the federal Ministry of Commerce with GB provincial mining authority sign-off. We can document origin from pit to port.
What grade of concentrate and what impurity profile? For scheelite skarn material, we're targeting 65%+ WO₃ concentrate through gravity plus flotation. Impurity work — molybdenum, arsenic, phosphorus, tin — is ongoing. Arsenic is the one everyone asks about. Early indications from our showings suggest low-arsenic mineralogy, but I won't quote numbers until we've run proper composite assays through a certified lab. I made the mistake early on of quoting a single ICP result to a buyer and having to walk it back. Don't do that.
Logistics? Karakoram Highway to Islamabad, then either Karachi or Port Qasim for seafreight. It's a 1,400 km haul from the concession, mostly on paved road, with the usual weather closures December through February. Air freight for smaller lots is viable through Skardu or Gilgit. Not fast. Not cheap. But functional and we've moved sample tonnages this way already.
What about the security narrative? Fair question and I get it every time. Gilgit-Baltistan is not the tribal belt. It's a tourism region — the same valleys foreign trekkers visit every summer. Different security posture entirely. That said, any serious JV partner should send their own risk team and I encourage it.
Where this goes next
The tungsten defence supply problem doesn't get solved by one concession in one country. It gets solved by five or six non-China sources each contributing 500 to 2,000 tonnes of contained WO₃ annually, backed by long-term offtake at pricing that reflects the strategic premium rather than the LME-adjacent APT spot number.
Honestly, that pricing conversation is the one Western buyers still find hardest. The Chinese domestic tungsten price sets the reference. Any non-China producer needs a floor above that to make the capex work. Procurement officers know this intellectually. Getting it through their finance committee is a different exercise.
But it's happening. The US Defense Production Act Title III allocations, the EU Critical Raw Materials Act offtake provisions, the UK MoD's own supply-chain review — the money and the mandates are lining up. What's missing is the assets on the ground, drilled and permitted and ready.
That's the part we're working on. If you're on the buy side and want to see the field data, come to Skardu. Bring boots.
Discuss a JV or off-take →