Water and Power for Remote Mine Sites in Gilgit-Baltistan: What a JV Partner Needs to Plan For

Last winter we ran a diesel gen-set at 2,850 metres for six weeks straight while waiting on a transformer that was stuck at Rawalpindi dry port. Burned through 41,000 litres. That's the number I want any serious JV partner to sit with before we talk grades and tonnage.

Because here's the thing — the geology in Gilgit-Baltistan is genuinely world-class. Copper-gold porphyry, antimony vein systems, tungsten skarns, pegmatite fields that haven't been properly walked over. But the mineral in the ground is maybe 40% of the conversation. The rest is water, power, road, and whether you can keep a crew fed and warm from November to March. I've watched more than one foreign visitor go quiet on the drive back from a site once they've worked out the logistics on the back of a napkin.

So let me lay out what we actually plan for. Not the brochure version.

Water: abundant, but not where you need it

Gilgit-Baltistan sits on some of the largest ice reserves outside the poles. The Batura, Hispar, Biafo and Baltoro glaciers feed the Indus, Hunza, Shigar and Gilgit rivers. Water is not scarce in the region. Water at your specific portal or pad, at the elevation you need it, at the flow rate a flotation circuit demands — that's a different problem.

A few things I tell every partner who visits:

Seasonal flow variation is severe. A nallah (side stream) that carries 200 litres per second in July can drop to under 15 in February. If your process design assumes summer flow, you'll be trucking water or shutting the mill for four months.

Sediment loading is brutal. Glacial melt carries fine rock flour — the Hunza River can push over 4,000 mg/L of suspended solids during peak melt. Any process water intake needs settling ponds, and you'll be dredging them. Budget for it.

Community water rights are real and they matter. Villages downstream of your intake have been using that channel for irrigation for centuries, often under a rotational system called warabandi locally. We negotiate water-sharing before we drill, not after. Every time. Skipping this step is how projects die in year two.

Groundwater is generally not the answer. Fractured-rock aquifers in the Karakoram are unpredictable and low-yield. Surface water plus storage is almost always the right design.

For a mid-scale operation — say 2,000 tonnes per day through a mill — we're typically planning around 25 to 35 litres per second of make-up water, with a lined reservoir sized for 60 days of winter operation. That reservoir is often the single biggest civil earthwork on site after the tailings facility.

Power: hydro is the honest answer

The grid in Gilgit-Baltistan is thin. The main 132 kV line runs the Karakoram Highway corridor and branches sparingly. Load-shedding is a fact of life — even Gilgit city sees 8 to 14 hours of outages in winter when hydel plants ice up. Connecting a mine to the public grid is possible in some locations but you cannot rely on it for continuous mill operation.

What actually works, in order of preference:

Run-of-river micro and mini hydro. This is the play. GB has an estimated 40,000 MW of hydro potential and less than 5% is developed. A well-sited 2 to 8 MW run-of-river plant on a tributary near your operation is cheaper per kWh over 25 years than anything else, and the government's Water and Power Department (WAPDA and the GB Power Department) has a workable framework for private generation with wheeling arrangements. We've scoped three of these on our concessions. Lead time is 30 to 42 months from feasibility to commissioning — that has to be in your Gantt chart from day one.

Diesel plus solar hybrid for early works. For camp, exploration drilling and initial development, a containerised diesel plant with a solar array handles it. Solar irradiance in GB is excellent — Skardu averages over 5.8 kWh/m²/day annually — but panel soiling from dust and snow load on structures both need engineering attention.

Grid connection as backup, not primary. Where a 33 kV or 132 kV tap is within reasonable distance (we've got two concessions within 18 km of usable lines), a grid connection makes sense as redundancy. Not as base load.

The honest number: for a 2,000 tpd operation you're looking at 8 to 12 MW of continuous load once the mill is running. That's not trivial to build in a valley three hours off the KKH.

What I got wrong early on

I used to think power and water were engineering line items you solved after the resource was proven. Drill first, JORC-compliant estimate, then figure out infrastructure. That's how it's taught.

Honestly, that sequence doesn't work in GB. The infrastructure timeline is longer than the drilling timeline. If you wait until you have a bankable resource to start on the hydro licensing and the community water agreements, you're adding four years to first production. We now run infrastructure scoping in parallel with early-stage exploration — sometimes before, if the geology already looks obvious from surface sampling.

For a serious JV partner, this means the term sheet needs to cover infrastructure capex explicitly. Not as a contingency. As a defined workstream with its own milestones, its own budget line, and ideally its own project manager who has actually built something above 2,500 metres before.

What we bring to the table

Sixteen concessions, and for each one we've done a first-pass infrastructure assessment — nearest hydro-suitable stream, elevation drop available, distance to nearest 33 kV tap, community stakeholder map, road access class by season. It's not a full feasibility. But it means when a partner walks in, we're not starting from a blank page on the question of whether the site is actually buildable.

A copper-molybdenum prospect that needs 15 MW and sits 60 km from the KKH with no side road is a very different project from an antimony vein that needs 800 kW and a 12 km spur. Both can work. They're not the same conversation, and the JV structure shouldn't pretend they are.

If you're evaluating Pakistan for critical minerals off-take or equity, ask the operator about water and power before you ask about grade. The answer tells you whether they've actually been on the ground, or whether they're selling a map.