Photovoltaic silver consumption has tripled since 2015 and forecasts for 2030 keep climbing. We unpack TOPCon, HJT, and what thrifting really means for ounces per panel.
Contents3 sections
Solar panels now consume more silver than the entire photography industry did at its 1999 peak. And the curve hasn't flattened.
Every major forecast issued before 2022 has been revised upward at least once.
The technology mix is shifting fast
Standard PERC cells use roughly 10mg of silver per watt. TOPCon cells, which captured over 60% of new capacity in 2024, use closer to 13mg. Heterojunction (HJT) panels, the rising premium tier, demand 20mg or more per watt despite years of thrifting research.
Why thrifting has limits
Silver paste sits on the front of solar cells as a conductive grid. Manufacturers have steadily reduced the amount per cell, but each reduction trades off cell efficiency. Below a certain threshold, resistance losses overwhelm the silver savings.
- 2015: ~150mg of silver per cell average
- 2020: ~100mg per cell across PERC
- 2024: ~80mg PERC, ~110mg TOPCon, ~140mg HJT
- Floor estimated at 50-60mg before efficiency degrades unacceptably
'You can thrift silver out of solar paste, or you can keep cell efficiency climbing. You cannot do both forever.' — PV materials engineer at a tier-one Chinese manufacturer
The 2030 demand math
The Silver Institute pegs 2024 PV silver demand near 230 million ounces, roughly 20% of total fabrication demand. Bloomberg NEF projects annual solar installations hitting 800GW by 2030, more than double 2024 levels. Even with aggressive thrifting, that pencils out to 350-450 million ounces of annual PV demand by decade's end.
For context, total global mine supply runs around 820 million ounces. Solar alone could absorb half of that within five years, before accounting for medical, electronics, brazing, jewellery, and investment demand.
The supply side has no comparable growth lever. Primary silver mines account for less than 30% of supply; the rest is byproduct from copper, zinc, and lead operations whose production is governed by base metal economics, not silver prices. Higher silver prices barely move the needle on byproduct output.
That mismatch is what keeps analysts revising 2030 projections higher. Demand has visible technology-driven growth. Supply does not.
Bottom line: solar isn't a silver thesis on its own, but it's the structural floor under every other silver thesis. Don't ignore the panels on your roof.
Reader Letters
The mailroom is empty.Be the first to write in.