The obstacles to working with red mud

Ireland doesn't usually come up in conversations about mining, given its limited mineral wealth. Yet there's a facility on the river Shannon, Aughinish, about half an hour from where I grew up, that refines close to a third of all the alumina produced in Europe.

China, unsurprisingly, refines around 60% of the global supply, with Australia contributing 14% and Brazil, India and Russia accounting for most of what's left. Assuming you'd rather not source feedstock from China or Russia, that leaves roughly a quarter of the market available to you. 

Aughinish illustrates just how odd this industry can be. The plant is owned by Rusal, Russia's aluminium giant. Its founder faced personal sanctions in 2022, though the company and its Irish operation were exempted. Reports have since surfaced suggesting most of its alumina output makes its way to Russian smelters, feeding a supply chain that ultimately reaches sanctioned defence contractors. The EU has allowed Aughinish to keep operating on the grounds that closing it would damage Europe's own aluminium supply chain and increase reliance on external imports. So a plant on European soil continues supplying Russia, in part because shutting it down would leave Europe more dependent on Russia. That's the strange logic governing critical metals right now.

The second obstacle is iron. Anyone trying to recover the higher-value metals in red mud such as titanium and rare earths, has to work around it first. Standard practice is acid leaching, but iron oxides dissolve easily, so targeting even 80 grams of scandium per tonne means dissolving a large share of the surrounding iron too. That pushes up acid consumption and adds extra separation steps. What's needed is a way to strip the iron out early and cleanly, so the rest can be separated more efficiently. The third obstacle concerns the scale and value of what can actually be extracted. As noted, red mud carries plenty of iron, titanium and aluminium in oxide form, all essential inputs for domestic manufacturing. It also contains scandium and gallium. Scandium is a tiny fraction of the total weight, yet it accounts for somewhere between 50–70% of the material's value per tonne.

That figure reflects raw contained value at current prices, and it's worth distinguishing from what a producer can realistically capture. If iron, titanium and aluminium can be extracted cheaply enough, those higher-volume metals generate their own margin. This means scandium's share of actual profit may end up smaller than its share of headline value. 

On scandium specifically, global output sits at around 40 tonnes annually, well short of demand. It sells for roughly $4,000 per kilo as an oxide, among the priciest of the rare earths, largely because there are no dedicated scandium mines, supply exists only as a byproduct of titanium, nickel and uranium production. China tightened access further last year by placing export controls on all forms of scandium. Set against iron ore, which is produced in the billions of tonnes annually, this is a minuscule market. But scandium punches well above its weight. Adding just a few grams to a kilo of aluminium alloy boosts tensile strength by roughly 30%. For rocket or missile applications, where cutting weight without sacrificing strength matters enormously, that's a significant advantage. It also improves weldability and reduces cracking, making the resulting alloy easier to repair and less dependent on a fragile supply chain. 

So why hasn't it caught on more broadly? Because supply has never been consistent or large enough for engineers to design around. Committing to a metal that could be cut off by a single trade decision, forcing a product redesign, is not an appealing bet. 

The fourth obstacle is what happens to the market once a reliable, large-scale scandium supply materialises. One view is that flooding a small market crashes the price. I'd argue the more probable outcome is a Jevons-style effect, where stable, growing supply actually drives up demand. Run the numbers: if just 0.1% of global aluminium output (roughly 70 million tonnes) used a 0.5% scandium alloy, that alone would require 350 tonnes of scandium annually — about 10 times current global production. Balancing price as supply scales will be delicate, but aluminium's own history offers a precedent. In the 1800s it was among the most expensive metals on earth. Once the Hall-Héroult process eliminated its scarcity, prices collapsed, demand surged, and aluminium went from a luxury item to something we now use to wrap sandwiches. 

Which means the technology is only one half of what makes a company here work. The other half is a strategy for offtake and pricing, plus the groundwork of building demand with buyers, so that growing supply doesn't erode your own margins. Nail the extraction chemistry but misjudge the market, and the margins shrink considerably. To sum up the challenges: it's a $3 trillion mine that's already been dug up, but the supply is concentrated in countries you may not want to depend on, the iron complicates the chemistry, a substantial chunk of the margin depends on a fragile market, and the whole bet only works if that market grows in step with supply. That's a heavy set of demands for a single company to meet, which brings me to the one I believe can pull it off.

The elephant in the room

Ganesh, the elephant-headed god of new beginnings (fitting, since no one has cracked red mud yet) and remover of obstacles (also fitting), makes for an apt avatar to FAST Metals' branding. 

Red mud is a market where founder-market fit genuinely matters, it demands real expertise on both the technical and commercial sides. Sumedh and Anthony, who met during PhD research focused on red mud, bring both. Sumedh has spent over a decade in technical roles at Glencore and Air Liquide, two companies that understand how to run industrial processes profitably over the long term. He's led development on lithium-ion and battery recycling programmes and designed metallurgical processes for gold, copper and lithium extraction. 

On the commercial side, Anthony most recently served as CEO of Nyrstar's US operations, running five mines, two processing plants, and the country's last remaining primary zinc smelter. Over more than twenty years he's scaled production at Li-Cycle, Freeport-McMoRan and Newmont. 

Their process uses oxalic acid as the leaching agent, which they're able to produce from waste industrial carbon monoxide streams, cutting both cost and environmental footprint. The oxalic acid reacts with iron first, pulling it out in soluble form, before capturing the remaining metals as oxides. Those metals still require further refining to reach higher purity, but the key advantage is that the reagents can be recycled at a high rate, keeping costs down, and the leftover waste is inert silica making up only about 10% of the original mass. In other words, the iron is removed early and cleanly, solving one of the biggest bottlenecks to most startups’ success. 

Their plan centres on mobile skids, each capable of processing about a tonne of red mud daily, which can be added cheaply on-site as operations scale. The feedstock itself is potentially free or even negative cost, since red mud is currently a liability that refiners pay to store. With low input costs and largely recycled reagents, the iron, titania and alumina recovered can cover most of the margin before a single gram of rare earth is sold. That's what meaningfully de-risks the open question around scandium pricing — because the cheaper metals are extracted profitably on their own, FAST isn't staking everything on scandium prices holding firm. 

One caveat I want to mention. I've seen plenty of startups build unit economics around feedstock that starts out free or cheap simply because the supplier treats it as waste. That tends to change once the startup demonstrates the feedstock has real value — at which point the supplier has every incentive to reprice or demand a cut. In the worst case, you become entirely exposed to their pricing power, forcing you to move upstream to lock in supply yourself. 

I expect the same dynamic will eventually play out with red mud, but likely not for at least a decade. Part of that is down to storage costs (raising the walls isn't cheap), but more significant is the liability risk. In 2010, a tailings dam wall failed at an alumina refinery in Hungary, releasing roughly a million cubic metres of red mud into the surrounding valley. Ten people died, many others suffered chemical burns from the alkalinity, and forty square kilometres of land were left barren. With more than 150 million tonnes of red mud still being produced every year, that liability exposure isn't disappearing anytime soon. 

The range of metals FAST can recover should also help cushion against any single price swing, and their low cost base means they can plausibly stay profitable even when prices dip. Some upside still depends on scandium, gallium and rare earths becoming larger markets over time, but the lower-cost metals carry the margin while that plays out. 

Just as important as the technology is the mindset of the founders. We were initially sceptical when we first came across FAST, it seemed like an extraordinarily difficult market to solve for given the combines technical and commercial hurdles. But just as we'd underestimated the difficulty of the market, we'd also underestimated this team. 

Beyond their obvious technical and commercial credentials, what stood out most to me was the founders' mindset. Sumedh is constantly learning and adapting, with a quiet, intense focus. Anthony brings hard-won experience from turning around some of the toughest mining operations in the West. Neither has a traditional startup background, so much of this world is new to them, yet the speed at which they've improved week over week has impressed me more than anything else. 

FAST has closed a $4.3 million pre-seed round to fund this work, backed by investors and advisors who understand the challenge, including Azolla, NCV, and Humba Ventures, alongside advisor Nick Popovic, former Head of Copper and Zinc Trading at Glencore International. 

In the end, the team's success will hinge on keeping costs low, volumes high, and finding a way to sell into a market resilient enough to absorb price swings. What gives me confidence in FAST's ability to manage that is their recognition that this is as much a commercial challenge as it is a chemistry one. 

Red mud is the trillion-dollar, billion-tonne elephant in the room.