THE hydrogen economy was supposed to be platinum’s salvation, but it is turning out to be the lightest of lifelines.
Deployment of electrolyser capacity, the technology that positioned green hydrogen as an energy storage miracle, is expected to be 7,000MW this year. That’s not an insignificant addition, yet green hydrogen of that scale absorbs only 45,000 ounces in platinum group metal (PGM) supply, says Metals Focus, a UK-based industry research house. That’s a small return considering the broad-based policy support given to hydrogen energy and storage applications since 2020.
Platinum miners are therefore acknowledging that ‘green hydrogen’ – which is a fraction of the 7.7 million oz in annual platinum demand (excluding other PGMs) – is completely insufficient to remedy its long-term demand concerns.
Current PGM demand is heavily reliant on the automotive sector, particularly from internal combustion engines which use PGMs heavily in their autocatalysis compared to hybrid cars, which combine petrol and battery technology and use less. Both technologies are threatened by battery electric vehicles (BEV) which don’t require autocatalysis. While BEV adoption has slowed lately in favour of hybrids, giving PGMs precious time to find alternative uses, the expectation is that BEVs will dominate the drive-train eventually.
Luckily, the PGM sector has always managed to find other specialised industrial uses whenever the question of future demand comes up. “With green hydrogen dreams steadily diminishing, the PGM sector is now turning its attention to other potential sources of demand, this time in the tech space,” say analysts at BMO Capital Markets, George Heppel and Helen Amos.
Enter ruthenium and iridium, minor elements in the PGM family of metals. Demand is growing for iridium and ruthenium, which are produced in relatively small proportions to the overall basket. Exponentially, some say, pointing to the infrastructure rollout behind AI’s advance.
The World Platinum Investment Council, an industry-funded marketing and research organisation, estimates that some 500,000oz in incremental PGM demand was created in the past two years from data storage, optical networks and semiconductor manufacturing. “On aggregate, the iridium deficit is expected to increase year-on-year in 2026 while ruthenium deficit is expected to decline but remaining substantial at 14% of demand,” says Investec analyst Nkateko Mathonsi in a recent note.
It’s not just ruthenium and iridium; fibre glass manufacturers are also buying more platinum. The metal is used in machines that make micro-processors. Bushings coated with fibre glass containing platinum vastly improve component connectivity. “The two minor metals together with platinum are expected to be the biggest beneficiaries of AI-linked demand in 2026 and beyond,” says Mathonsi.
In addition, AI componentry constitutes recurring demand as each bushing needs recoating. That is quite unlike PGM-bearing internal combustion engines and hybrid BEVs, which once driven off the lot, become last year’s demand statistic. “The industry forecast suggests that platinum demand from the glass sector alone could increase by 83% in 2026, driven by new fibreglass capacity expansion from stronger demand for applications including AI-related technologies,” says Craig Miller, CEO of Valterra Platinum.
[Minor metals] markets are small, concentrated and supply-constrained, meaning incremental demand does not need to be large to move prices materially – Arnold van Graan, Nedbank Securities
Valterra is entering into more new supply agreements with end users than ever before, partly driven by iridium and ruthenium. The market is “at a higher level of contractedness than we’ve had before” for these metals, Miller says.
The minor metals also bring with them an extremely helpful revenue boost. “These markets are small, concentrated and supply-constrained, meaning incremental demand does not need to be large to move prices materially,” says Arnold van Graan, an analyst for Nedbank Securities in a recent report. “Even a modest acceleration in technology-driven demand could provide a meaningful tailwind.”
“Everyone is joining the opportunity to really think through how do we diversify the demand into different buckets, rather than being overly reliant on just the automotive sector,” says Miller.
Good news surely, but with this new demand comes the trickiest of challenges for PGM miners. While the industry is well positioned to supply and catalyse nascent technology, it also runs the risk of destroying its own investment case by committing to new production. “The PGM sector must telegraph its ability to guarantee long-term supply to growing sectors without placing too much emphasis on looming oversupply, which could erode the recent price support found for platinum and palladium,” say Heppel and Amos.
Bad history
Investment and fund manager analysts are fearful of miners embarking on new projects for future, untested demand. It raises the stakes on capital allocation potentially drawing funds from future dividends. Something of a knife-edge, you could call it.
The market remembers the last time PGM prices flew high. Share prices ended up being severely damaged. That’s because the market was snatched from under the feet of those miners who were too enthusiastic on growth. “I want to take you back to 2019, when I sat across the table from investors who basically said the company had become ‘uninvestable’,” said Nico Muller, CEO of Impala Platinum, earlier this year. “I’d rather have a strong balance sheet, so we don’t get into that position again.”
Miller thinks those lessons continue to sting, somewhat usefully. Valterra cut R12bn in costs in 2024/25 and is ever watchful of how inflation has become a factor again post the attack on Iran in February, and the wider Middle East conflict. New, unexpected inflationary pressure has been heaped on miners. Valterra is tracking its cost guidance for 2026, but 6% of costs are impacted by diesel, chemical, and explosive prices.
Yet there is a place for production growth. “If you operate in the lower half of the cost curve, if you have well-capitalised assets and discipline around the execution of that, then you should be able to sustain new production coming into the market,” says Miller.
Valterra’s flagship mine in Limpopo, Mogalakwena, can grow output a fifth – for a net growth of 10% – by mining the giant resource underground. So far, however, Valterra hasn’t committed to it. The board is due to opine on an investment decision next year. Other miners are less forthcoming. Implats has an option to build a palladium mine, also in Limpopo, but won’t do it; not yet.
Demand from AI adds an interesting dynamic to the PGM picture but it, too, might not be a “silver bullet”, though perhaps times have changed since expectations around green hydrogen were first raised six years ago. The US has risen as a hoarder of critical minerals; the EU is also considering its options.
Miller says it’s important to interrogate why PGMs are critical to each bloc: in the US, it’s for rhodium; in Europe it’s platinum, both owning to the unique circumstances of their automotive industries. There’s no comforting single source of demand but at least there’s really only one source of primary refined PGM supply, which is South Africa.
If you operate in the lower half of the cost curve, if you have well-capitalised assets and discipline around the execution of that, then you should be able to sustain new production coming into the market – Paul Miller, Valterra Platinum
Richard Stewart, CEO of Sibanye-Stillwater, believes governments could support PGMs by offering US-style price floor guarantees. “Iridium is a good example,” he said in an interview. “They aren’t going to develop the [AI] technology until they’re certain of supply, but likewise, supply is not going to develop until it’s certain of demand.
“So you get caught in this catch-22 between who’s going to invest first to drive the other. That’s where being able to bring in floor prices – to say we’ll protect that to a point, to get a certain amount of critical supply to drive comfort in terms of demand levels and ultimately balance the market – that’s another mechanism where I think it can work,” he says.
Introducing price floors, especially provided by the public sector, is not to everyone’s taste, and difficult even for governments to adopt. The EU doesn’t favour them, even in lithium, where the US has provided a price protection mechanism.
The PGM market is not in a bad place. Metals Focus, UK semi-fabricator Johnson Matthey, and chemicals firm BASF agree that the platinum deficit of the past two years is set to continue. Views on palladium diverge sharply between deficit and surplus but rhodium will be in balance. “We think you’ll have some growth in secondary supply – recycled volumes – but ultimately the deficits I spoke about in platinum, we do see those continuing,” says Miller.
The market is divided on the future for palladium. Says Miller: “Every time we sit down and talk about a palladium surplus, it seems to just move out another year or so. I think rhodium, because of some of its applications, will either be in deficit or in balance. And ultimately those three metals support a higher price environment than what we saw back in 2024 and 2025.”
This article first appeared in Miningmx’s The Mining Yearbook 2026. Click here for more articles like this one.
