How We Took an Indus Placer From First Pan to a 200-Tonne Bulk Test

The first colours came off a pan on a gravel bar about 40 minutes' drive downstream of Bunji, where the Indus slows and drops its coarse load. Nothing dramatic. Fine flour gold, a couple of small flakes, maybe eight or nine colours in a pan that took me twenty minutes to work properly because I'm slower than the local washers who've done it their whole lives.

That's how a placer gold operation actually starts. Not with a drill rig. With a pan and a lot of walking.

I want to lay out how we moved this particular block from that first bar sample through to a proper 200-tonne bulk test, because most of what gets written about alluvial gold in Pakistan is either wildly optimistic or dismissive. The reality sits in the middle, and it's the middle that decides whether an off-take partner should bother getting on a plane.

Reconnaissance, and why the boring part matters most

We worked roughly 6 km of active and terrace gravels over two field seasons. The plan wasn't to find the richest single pan — anyone can cherry-pick a pan. It was to build a picture of where the gold sits in the profile and whether it's consistent enough to feed a plant every day for years.

Honestly, I got the sampling density wrong at first. Early on I spaced pits too far apart, chasing ground, and the numbers jumped around so much they told me nothing. So we tightened up. Test pits on a rough 100-metre grid across the bar, hand-dug down to bedrock or refusal where we could reach it, with the full section logged and each horizon washed separately.

What came out of that:

• The gold isn't evenly spread top to bottom. The bottom 30–40 cm sitting on the clay-rich false bedrock carried most of it. Above that, the gravels ran thin.
• Grades across the pit programme averaged something in the range of 0.18 to 0.35 grams per cubic metre for the productive horizon. That's a working figure, not a headline figure. Alluvial numbers lie easily, and I'd rather quote you the range than a single flattering point.
• The gold is fine. A lot of it in the sub-0.5 mm fraction, with the occasional flake up to 1.5 mm. Fineness ran high — good, clean colour, low base-metal contamination in the concentrates we assayed.

Fine gold is the thing that catches people out. It travels. It floats over a badly set sluice and walks straight out the tailings end. More on that below, because it's where I lost recovery.

Bulk sampling gold — the step nobody wants to pay for

Test pits give you grade indications. They don't give you recovery, and they don't tell you if you can process 40 or 50 cubic metres an hour without the whole thing clogging on Indus silt. For that you need volume. Real volume.

We ran a staged programme. First a 20-tonne trial through a small trommel-and-sluice setup we trucked in, just to shake out the mechanical problems. Then, once we trusted the kit, a 200-tonne bulk test drawn from the productive horizon across three separate pit locations so the feed represented the bar and not one lucky spot.

Here's the part I'll be straight about. Our first recovery run through the sluice was poor. We were losing fine gold off the tail — I could pan the tailings and still get colours, which is exactly what you don't want to see. The riffles were carrying too much water velocity for the fineness of our gold.

So we changed the back end. Added a set of finer matting and a slower final run, and put a small centrifugal concentrator on the sluice tails to catch what the riffles missed. Recovery on the corrected setup came up sharply. The centrifuge alone pulled back a meaningful slice of what we'd otherwise have dumped in the river.

The 200-tonne test gave us a defensible recovered grade and — more important — a recovery percentage we could stand behind in front of a partner. That number is what you finance an operation on. Not the pan.

What the bulk test actually told us about a real operation

A few things came out of this Indus placer case study that matter if you're thinking about off-take or a JV on alluvial ground here.

Water and season run the show. The Indus is glacier-fed. High flow through summer means the active bar is under water when melt peaks, and the working window for the wettest ground is genuinely limited. Terrace gravels above the current channel don't have that problem — they sit dry — but they need water pumped up to them, which changes your fuel and pump costs. We're now weighting our development toward the terraces for exactly this reason.

Strip ratios are kind here. This isn't lode mining. There's no drill-and-blast, no crushing circuit, no cyanide. You're moving unconsolidated gravel with an excavator and washing it. Capital per tonne is low compared with a hard-rock gold operation, and that's the genuine attraction of placer for a partner who wants cash flow while the bigger lode concessions in our portfolio go through longer permitting.

But — and it's a real but — the grade is modest and the margin lives or dies on throughput and recovery. You need to move volume, and you need to catch the fine stuff. Get sloppy on either and it's not a business.

Licensing sits under Gilgit-Baltistan's own minerals framework, administered provincially rather than under the federal setup you'd deal with elsewhere in Pakistan. Alluvial working is recognised and grantable. We hold the ground, the title's clean, and I'll walk any serious party through the concession documents rather than wave a map around and ask you to trust me.

Look, the honest read on this bar is that it's a solid, unspectacular placer gold operation with room to scale across the terrace ground once we settle the water logistics. It won't make anyone rich off one season. It will produce consistently, it responds well to a properly-tuned recovery circuit, and it can run alongside the harder projects rather than competing with them for attention.

The Indus and its tributaries — the Gilgit and the Hunza feeding in upstream — have been shedding gold into these gravels for a very long time. Local washing has gone on informally for generations, which is itself a signal worth reading. The question was never whether there's gold. It's whether you can recover enough of it, reliably, at a cost that works.

After 200 tonnes through the plant, I've got a number I trust. If you're running an off-take book and want to see the pit logs and the recovery data before you commit to anything, that's a conversation I'm happy to have — what fraction of the feed are you comfortable financing against?