The Solar Boom Is Using Less Silver Every Year

Yes, there is silver in solar panels, and somewhere there is a well-paid engineer whose whole job is getting it out. Look closely at a panel on a sunny roof and you’ll see the silver: a fine grid of bright lines across each dark blue cell, like frost on a window. That grid is silver paste. It’s there because silver is the best electrical conductor of any element, which engineers love and their accountants do not. Picture the budget meeting: the engineer gushing about conductivity, the accountant holding the silver invoice like a parking ticket.

The twist most silver headlines leave out is that the panels use less of it every year. Silver demand from photovoltaics was 197.5 million ounces in 2024. It fell to 186.6 million in 2025, and it’s forecast at 151 million for 2026, a 19% drop in a single year. Installations hit a record in 2025 and silver use fell anyway, because silver per panel keeps shrinking. That 19% is about 35 million ounces. At the October 24, 2026 price, it’s roughly $1.9 billion of silver the panels stopped eating, while the rest of us don’t have the willpower to skip dessert.

So “solar will need more and more silver” is the weak half of the case for silver. The strong half is about where silver comes from, and where the world goes when it runs short. It ends somewhere nobody puts in a forecast: your grandmother’s silverware drawer.

Is solar using more silver or less?

Less, and you can blame one person for it. Call her the Thrifter. She isn’t real: she stands in for the cell engineers of the world, and she has a desk, a microscope and a quarterly target with silver’s name on it. Hand her a recipe whose priciest ingredient keeps getting pricier and she won’t admire the price. She’ll get it out of the recipe.

The trade calls this thrifting, which sounds like a Saturday at the thrift store and is actually a campaign fought one hairline at a time. First she prints the silver lines narrower. Then thinner. Then she redesigns the cell so it needs fewer of them. When that runs dry she swaps the metal itself: silver-coated copper powder is already documented to cut silver use by 30% to 50% at comparable cell performance. Every milligram she saves gets its own slide at the quarterly review, and possibly a sheet cake.

Not every change goes her way. The industry’s move to TOPCon cells, a newer design, pushed back hard: a TOPCon cell uses roughly 50% more silver than the PERC design it replaced. That’s like switching to a recipe that calls for a stick and a half of butter instead of one. Thrifting won anyway. The result is the drop at the top of this page, less silver per unit of sunlight, every year, on purpose.

That is the honest bear case, and you may as well hear it from people who like silver. If the case for silver rests on “solar will need more and more of it,” that case has a Thrifter problem. The better case doesn’t depend on solar at all. It’s underground, at a copper mine, being ignored.

Where does the world’s silver come from?

Mostly from mines that aren’t looking for it.

In 2025, only 26.1% of the world’s mined silver came from mines that were primarily silver mines. The other three-quarters came up as a by-product, meaning it rode along while somebody dug for something else: 29.4% alongside lead and zinc, 28% alongside copper, 15.9% alongside gold.

That matters because most commodities fix themselves, the way a busy restaurant does. A dish gets popular, the price goes up, the owner hires another cook, more plates come out of the kitchen, and the price drifts back down. It’s why most commodity booms end.

For three-quarters of the world’s silver, nobody in the kitchen is cooking silver. They’re cooking copper, lead, zinc and gold. A copper mine’s board doesn’t approve a billion-dollar expansion because silver looks attractive. It approves it because of copper. Silver shows up at the smelter in whatever quantity the copper decided, and it would show up in the same quantity at twice the price or half of it. Silver is the parsley on the copper plate.

We don’t have the minutes from any real copper board, but we’re pretty sure they read something like this. (“Minutes of the Copper Mine Board Meeting. Item 3: copper price, discussed for two hours, with charts. Item 4: silver price, up again, apparently. Noted. Item 5: lunch orders, discussed for two hours.”)

That leaves roughly a quarter of supply that can respond to price, attached to a metal that has now come up short five years in a row.

Is there really a silver shortage?

You’ll often read that industry alone now uses more silver than the mines produce. It doesn’t, and the real version is better.

In 2025, industrial demand was 657.4 million ounces. Mine production was 846.6 million ounces. So industry on its own takes about four-fifths of what’s mined. A big bite. Not the whole plate.

What empties the plate is everybody else: jewelers, silversmiths and investors, who brought total demand to about 1.13 billion ounces. Recycling hit a 12-year high of 197.6 million ounces. Add that to what the mines produced, plus a few smaller sources, and supply still fell 40.3 million ounces short.

That was the fifth deficit in a row. The Silver Institute expects the sixth, in 2026, to be wider, at 46.3 million. Imagine a household that has overspent its paycheck five years running and is budgeting, with some confidence, for a sixth.

Forty million ounces is a number nobody can picture, so here’s its shape. It’s about 1,253 tonnes of silver, and silver is dense enough that the pile is smaller than you’d expect and heavier than you’d like. Poured across a tennis court, it would come up to about your knees. One year’s shortfall, one court, everybody wading.

And none of it was mined. It was withdrawn. Every year for five years the world has wanted more silver than it produced and covered the difference from metal already sitting somewhere: a vault, a warehouse, a coffee can of pre-1965 dimes, somebody’s grandmother’s silverware drawer. Added up since 2021, the drawer has given up roughly 762 million ounces, about five years of solar’s silver at the 2026 rate. That’s the silver in about 4.2 billion pre-1965 quarters, roughly a dozen for every person in America, slipping out of coffee cans, sock drawers and the back of the china cabinet, one butter knife at a time.

That is a fine arrangement right up until someone asks how much is left in the drawer. Nobody can tell you, because silver above ground doesn’t register its address. What the numbers do say is that the world’s shortfall is being covered by people who already hold silver, and each of them decides whether to let it go. Grandma, it turns out, is a swing producer, a title oil analysts usually save for Saudi Arabia. Unlike Saudi Arabia, she makes her production decisions at Thanksgiving.

Will solar’s silver demand grow again over the next 5 to 20 years?

Nobody knows, and the forecasts disagree with each other by about a factor of three, which is a lot even for forecasters. It comes down to a race between how fast the world builds panels and how fast the Thrifter takes the silver out of them.

The arithmetic fits on a napkin. The main industry outlook has installations about 30% higher in 2030 than in 2025. For solar’s silver demand to rise, silver per watt has to fall by less than about 23% over the same stretch. Silver-coated copper alone cuts it by 30% to 50%, and copper-plated cells in the lab are cutting further. The Thrifter is winning right now. The big forecasts assume she takes a long vacation.

Into the weeds: the installation forecasts and the competing silver forecasts

Installations. The world installed a record 664 gigawatts of solar in 2025, according to SolarPower Europe. For 2026 it forecasts the first contraction in more than twenty years, down 8% to 612 gigawatts, mostly because China is slowing. Its medium scenario then has the world installing about 864 gigawatts a year by 2030, roughly 30% more than in 2025.

The large forecasts. A 2025 academic study of silver supply and demand to 2030 projects photovoltaic demand of 10,000 to 14,000 tonnes a year, about 320 to 450 million ounces, because it assumes silver-hungry TOPCon and heterojunction cells take over. An LBMA Alchemist article from 2024 put solar silver demand at 546 million ounces by 2030, from a different baseline. Meanwhile the Silver Institute’s 2026 forecast for this year is 151 million.

The copper counterweight. In 2025 researchers at Forschungszentrum Jülich reported heterojunction cells with copper on both sides, using about 70% less silver. They reached an average efficiency of 22.4%, with a peak of 23.08%, and the team says it next plans to test the design in full modules. No commercial date was given. Oxford Economics, in a December 2025 study for the Silver Institute, notes that copper plating is a minimal share of sales today but is expected to “increase significantly” in the years to come.

As for ten or twenty years out, we couldn’t find a credible forecast of solar silver demand over that horizon, and the further out the number, the more likely it’s an extrapolation wearing a lab coat. The Silver Institute’s own summary of the 2026 survey, as reported by others, expects the structural deficit to “gradually erode” beyond 2026, though it also warns that supply stays tight. Treat all of it as a range, not a date.

Could batteries, EVs or data centers make up the difference?

They could add to the pile. Whether they could replace what solar is shedding depends on which one you mean.

Electric vehicles: real, steady, modest. An EV carries about 25 to 50 grams of silver, up to 79% more than a gasoline car, and automotive silver demand is forecast to reach about 94 million ounces by 2031. That’s about a seventh of today’s industrial total. EVs are the meat and potatoes of new silver demand: filling, reliable and nobody’s idea of a feast.

Solid-state batteries: the wild card. One widely repeated scenario has a silver battery design adding 16,000 tonnes a year of demand, about 514 million ounces, or roughly 60% of a year’s world mine production. If it happened, it would be the biggest thing to hit the silver market in decades. No manufacturer has confirmed it, which is why it’s filed under “possible” and not “forecast.” It’s the soufflé on the menu: spectacular if it rises, and nobody is taking orders yet.

Into the weeds: the EV and battery numbers

An electric vehicle contains about 25 to 50 grams of silver, 67% to 79% more than a gasoline car, which holds about 15 to 25 grams. Oxford Economics expects global EV production to grow about 13% a year from 2025 to 2031. It expects automotive silver demand to grow about 3.4% a year, to roughly 94 million ounces by 2031, with EVs overtaking gasoline cars as the biggest automotive source by 2027.

Several articles say a silver-carbon anode design associated with Samsung SDI uses about 5 grams of silver per cell. That works out to roughly 1 kilogram per 100 kilowatt-hour pack, against 25 to 50 grams in a car today, with mass production targeted for 2027. The scenario says that if 20% of the world’s cars (about 16 million vehicles) used it, silver demand would rise by 16,000 tonnes a year. It comes from articles that do not cite a source, no manufacturer has confirmed a silver loading at scale, and several of the designs may not use silver at all.

How much silver do AI data centers use?

Nobody knows, and be wary of anyone who quotes you a precise figure. Global IT power capacity grew from 0.93 gigawatts in 2000 to nearly 50 gigawatts in 2025, and chips and packaging do use silver, all of it counted inside the 657.4 million ounce industrial total. But no one has published a reliable silver-per-gigawatt number. We won’t invent one, though we were tempted. It would have been a beautiful number, round and confident, the kind that lands on a conference slide and gets repeated until everyone assumes somebody checked it.

The honest ranking goes like this. Solar is the largest single industrial user, 29% of industrial silver demand in 2024 according to Oxford Economics, and shrinking per watt. EVs, grid investment and AI electronics add steadily. Batteries are a very large maybe, the kind of maybe your mom gave when you asked for a pony.

What does all this add up to for silver?

An argument narrower than the usual one, and much harder to dismiss. Solar, the biggest industrial user, is shrinking its silver bill on purpose. Supply mostly can’t respond, because three-quarters of it answers to copper, lead, zinc and gold. And the gap between the two has been filled, five years running, out of the drawer.

None of this is a forecast, and it’s certainly not a price target. It’s a description of two kinds of silver. On the roof, silver is a cost, and very good engineers are paid to shave it off a milligram at a time. In the drawer, silver is the point. If Grandma’s flatware is stamped “sterling,” a federal rule says it has to be at least 925/1,000ths pure silver. It owes nobody anything, nobody can print more of it, and it weighs exactly what it weighed the day it went in the drawer.

The Thrifter can thin the frost on the roof to a whisker. Nobody has ever thrifted the gravy ladle.

Sources

• 2025 final totals (total demand 1.13 billion ounces, mine production 846.6 million, recycling 197.6 million, industrial demand 657.4 million), the fifth consecutive deficit and the 2026 deficit forecast of 46.3 million ounces, released April 15, 2026: The Silver Institute, World Silver Survey 2026. The 2025 deficit of 40.3 million ounces, the 762-million-ounce drawdown since 2021 and the 2026 industrial forecast of 639.6 million are as reported by Mining Visuals and Investing News (secondary sources). Earlier, preliminary figures (665 million ounces of industrial demand, 813 million of mine production, a 95-million-ounce deficit) were replaced by these.
• Solar installations of 664 gigawatts in 2025, the 2026 forecast of 612 gigawatts and the roughly 864 gigawatts a year by 2030 in the medium scenario: SolarPower Europe, Global Market Outlook.
• Silver in EVs, automotive demand to 2031, data-center capacity, solar’s 29% share of industrial demand and copper plating: Oxford Economics for the Silver Institute, Silver: The Next Generation Metal, December 2025.
• Photovoltaic demand projected at 10,000 to 14,000 tonnes a year by 2030 and the 1.5 to 2 times silver intensity of TOPCon and heterojunction cells: “Forecasting silver demand and supply by 2030,” ScienceDirect, 2025. The 546-million-ounce figure: LBMA Alchemist, issue 117.
• Copper-metalized heterojunction cells: pv magazine, September 24, 2025.
• Solid-state battery silver claims (about 5 grams per cell, 1 kilogram per 100 kWh pack, the 16,000-tonne scenario, a 2027 production target), which are unverified: SilverSeek, August 2024 and Investorideas, 2026.
• Photovoltaic demand of 197.5, 186.6 and a forecast 151 million ounces, the TOPCon comparison and the silver-coated copper figure: the Silver Institute’s World Silver Survey 2026, as reported by Mercom.
• The by-product split — 26.1% primary, 29.4% lead and zinc, 28% copper, 15.9% gold, for 2025: Statista, compiled from the World Silver Survey.

Figures read September 23, 2026; supply, demand and outlook figures were updated to the April 2026 survey and checked October 3, 2026. The 1,253-tonne conversion of 40.3 million ounces, the tennis-court comparison, the 23% break-even, the 30% rise in installations, the 514-million-ounce conversion and the one-seventh comparison are our own arithmetic. So are two added October 26, 2026: the one-year cut of about 35.6 million ounces at $54.05 (silver’s price on October 24, 2026, gold-api.com), about $1.9 billion; and 762 million ounces ÷ 0.1808 ounces per pre-1965 quarter, about 4.2 billion quarters, about 12 per person at the Census Bureau’s 342,916,055. The by-product percentages are as compiled by Statista and were not re-checked against the 2026 survey.

Questions people actually ask

Is silver used in solar panels?

Yes. Silver conductive paste forms the fine lines that carry current across the face of a photovoltaic cell. Silver is used because it is the best electrical conductor of any element.

Is solar using more silver or less?

Less, and by a wide margin. Photovoltaic silver demand was 197.5 million ounces in 2024, fell to 186.6 million in 2025, and is forecast at 151 million for 2026 — a 19% decline — even as installations grew. Manufacturers are deliberately reducing silver per cell.

How much of the world’s silver comes from silver mines?

Only about 26.1% in 2025. The rest is a by-product of mining other metals: 29.4% from lead and zinc, 28% from copper and 15.9% from gold. This is why a high silver price does not reliably bring on new supply.

Is there really a silver shortage?

There has been a structural deficit for five consecutive years, estimated at 40.3 million ounces in 2025 and forecast to widen to 46.3 million in 2026. But the common claim that industry uses more silver than is mined is not accurate: industrial demand was 657.4 million ounces against 846.6 million ounces of mine production. It is total demand, about 1.13 billion ounces, that exceeds supply.

How much silver is in a solar panel?

The amount varies by cell design and falls every year as manufacturers print thinner, narrower conductive lines. Industry-wide, photovoltaic silver demand was 197.5 million ounces in 2024, 186.6 million in 2025 and is forecast at 151 million for 2026, even though installations kept growing.

Why is silver used in solar cells?

Silver is the best electrical conductor of any element. Solar cells use silver paste to form the fine lines that collect current from the cell surface. Manufacturers are working to use less of it, including by substituting silver-coated copper, which can cut silver use by 30% to 50%.

Does a TOPCon solar cell use more silver than a PERC cell?

Yes. A TOPCon cell uses roughly 50% more silver than the PERC design it replaced. Even so, thrifting, the process of printing thinner lines and using less metal per cell, outweighed that increase in the net demand figures reported for 2025 and forecast for 2026.

Will AI data centers cause a silver shortage?

There is no reliable public figure for silver demand from AI data centers. Silver is used in electronics, which sit inside the 657.4 million ounce industrial total for 2025, but precise data-center numbers quoted online are usually unsourced. Treat them with caution.

Will silver demand from solar panels increase in the future?

Not necessarily. SolarPower Europe's medium scenario has installations rising about 30% by 2030, but silver per panel is falling fast. Demand rises only if silver per watt falls by less than about 23%. Forecasts range widely, and no credible 10- or 20-year forecast exists.

Will electric vehicles or solid-state batteries increase silver demand?

EVs use about 25 to 50 grams of silver against 15 to 25 for gasoline cars, and automotive silver demand is forecast to reach about 94 million ounces by 2031. Solid-state battery claims are far larger but unconfirmed: no manufacturer has verified a silver loading at scale.

Why doesn't a higher silver price bring more silver to market?

Because most silver is not mined for its own sake. In 2025 only 26.1% came from primary silver mines; the rest was a by-product of mining lead and zinc, copper and gold. Those mines expand or shrink based on their main metal, so a higher silver price brings out comparatively little new supply.

Where does the extra silver come from when demand is higher than supply?

From silver that already exists above ground: vaults, warehouses, coins and household silver. The world has run a silver deficit for five straight years, and from 2021 those shortfalls added up to roughly 762 million ounces, all covered from existing stocks. Nobody knows precisely how much above-ground silver is left to draw on.