This website is designed only for use in portrait orientation on mobile devices. Please rotate your device.
Back to insights
February 19, 2026

Our material world – the benefits of unlocking circularity in material supply chains

Report

Part 1: The Real Cost of Gold

The material supply chains that underpin our economy are consuming a significant – and ever‑increasing – share of our global carbon budget. Materials like polymers and plastics, fibres and textiles, and metals and minerals are essential to modern life, yet they exact a heavy price in both economic and environmental terms.

In a series of posts, we will look at the impact of these materials and explore the novel technologies offering scalable, robust solutions to decarbonise our world.

To start the series, we’re looking at gold. Crucial to the functioning of our modern world- electronics, aerospace, medicine and dentistry all rely on gold for circuit boards, switches, connectors, satellite components, medicinal compounds and dental fillings. And while each smartphone we buy may only contain tens of milligrams of gold, with billions of smartphones in the economy, the stocks of gold locked up in products and waste materials are quietly increasing year on year.

Gold mining is incredibly energy intensive. To produce a single standard gold bar (400 troy oz or 12.4 kg), anywhere between 5,000 and 10,000 tonnes of rock may need to be moved, depending on the ore grade. This process typically requires roughly 1.5–3.5 terajoules (TJ) of energy per bar. That’s approximately equivalent to the annual electricity consumption of about 100–300 average UK homes for each bar of gold produced. And that’s just the direct energy.

Gold mining also relies on a lot of energy‑intensive resources and materials (steel, chemicals, cement, tyres, explosives) as well as heavy machinery (trucks, excavators, mills). Industry data suggest that around half of production costs for a typical gold mine are effectively energy‑related – either direct fuel and power or the embedded energy in inputs like reagents and consumables – meaning that roughly 50% of the mine’s total production cost is controlled by direct energy costs.

Then there are the chemicals.

Mining for gold isn’t just blasting rock; it also requires chemicals throughout the process. Ammonium nitrate is used to create powerful explosives for blasting through rock; sodium cyanide is used to dissolve and separate gold from crushed ore; lime is added to stabilise the potentially lethal cyanide; and nitric and sulfuric acids appear in different parts of refining and downstream processing.

These chemicals are often described as “frozen energy” because they are so energy intensive to produce. Industry estimates suggest that up to around 75% of the cost of sodium cyanide is bound up in energy. Lime production requires extreme heat in large kilns, with energy accounting for about 65–70% of its price. Blasting explosives are made mostly from fuel‑based materials, giving them an energy exposure of more than 80%. Even refining acids depend on high‑temperature, fossil‑fuel‑driven processes. Taken together, a large share – often cited at around 70% – of a mine’s operating costs can ultimately be traced back to energy- direct and embedded.

And ultimately, there’s the cost in terms of natural capital and biodiversity loss.

Both large‑scale and artisanal gold mining drive deforestation, especially in tropical forests; globally, gold and coal together account for just over 70% of mining‑related deforestation between 2001 and 2019. From 2001 to 2020, mining activities were responsible for the loss of about 1.4 million hectares of forest cover, of which roughly 450,000 hectares occurred in tropical primary rainforests – among the most carbon‑dense and biodiverse ecosystems in the world. Add to this the chemicals used in gold mining, which often end up contaminating rivers and wetlands, killing aquatic life and toxic materials such as mercury bioaccumulating in food webs.

From a natural‑capital perspective, gold mining frequently converts high‑value, multifunctional ecosystems into simplified or degraded landscapes, and many losses – such as old‑growth forest structure and endemic species – are effectively irreversible on human timescales.

What Are We Using This Gold For?

Much of the approximately 5,000 tonnes of gold entering the global economy each year goes into jewellery, investment products (bars, coins, ETFs) and central bank reserves. Around 7% of annual gold demand – roughly 300–330 tonnes – is used in high‑reliability electronics, most of which ultimately ends up as e‑waste as devices reach end‑of‑life. Global e‑waste in 2022 was about 62 million tonnes, yet only 22.3% of this was documented as properly collected and recycled; this e-waste total is expected to climb to 82 million tonnes by 2030 with a significant contribution from the rapid rise of AI and cloud computing, while the documented collection and recycling rate looks likely to drop to 20%. Crucially, even this does not guarantee that embedded gold (and other critical minerals in e-waste) is actually recovered – only that the waste has entered a formal recycling system.

The Potential for Urban Mining

The current global stock of gold embodied in existing waste electronics (discarded devices and low‑grade e‑scrap, not counting still‑in‑use equipment) is huge. Conservative estimates place it in the low‑thousands of tonnes, and it could plausibly be on the order of 2,000–4,000 tonnes, given annual flows of several hundred tonnes into e‑waste and historically low recovery rates. This stock is sitting in “urban mines” – landfills, scrap yards and processing centres. Recent assessments indicate that in 2022 alone, around USD 15 billion worth of gold was discarded in e‑waste.

Consider this: e‑waste contains much more gold than most mined ore. Even taking the low end of estimates – around 100 grammes per tonne of mixed e‑waste – it is still roughly 20–100 times richer than typical gold ore at 1–5 g/t. At 800–1,500 g/t, high‑grade e‑waste such as computer RAM and certain circuit boards can be 160–1,500 times richer than average ore.

If the existing global stock of electronic waste were fully mined and processed, it could realistically yield around 1,800–3,600 tonnes of gold. The technology sector currently uses about 320 tonnes of gold per year, including roughly 270 tonnes in electronics, meaning the recoverable gold already sitting in this “urban mine” stock is equivalent to around 6–10 years of total technology‑sector demand, or approximately 7–13 years of electronics demand alone.

Put another way, the energy value embedded in that gold is comparable to the annual electricity consumption of roughly 1–3 million UK homes.

While processing e‑waste does generate emissions, a key advantage of urban mining is that much of the material is already collected and above ground. Collection systems exist today, but weak recovery economics limit sorting and processing. With improved technology and incentives, significantly more value could be unlocked without the need for new extraction. Crucially, the economics of urban mining are not driven by gold alone: e‑waste also contains copper, platinum, nickel and other valuable metals that can be recovered using similar technologies.

In total, recent UN estimates suggest that metals in today’s e‑waste stream contain in the order of USD 60–90 billion in unrealised value. Recovering all these materials makes urban mining compelling economically, while also avoiding the broader impacts of conventional mining: high energy use and carbon emissions, natural‑capital depletion, biodiversity loss, water pollution, heavy reliance on hazardous chemicals and, of course, deep social impacts on communities and countries least able to address the environmental damage.

When we talk about the technologies we’re backing in metal recovery, we’re not talking about a sector with marginal, peripheral impact. We’re talking about cutting emissions across Scopes 1, 2 and 3, preserving natural capital and unlocking serious economic value. These are technologies that hyperscalers like CoreWeave and AWS can deploy on site to extract and retain value from exhausted circuit boards and GPU stocks, sharpening their edge against the competition. They are tools that industrial giants can use to diversify away from concentrated processing hubs and regain control over strategic materials. This is not waste management; it is the difference between technology that quickly becomes redundant and depreciates in value, and assets that contain value that can be extracted and monetised, and rapidly brought back into supply chains.

The new frontier of mining for metals is urban mining – right here in front of us in our landfills and processing centres.

Later in this series of posts, we’ll look at the impacts of other metals, the impacts on oceans and waterways, and some of the novel solutions deep‑tech firms are developing to access and process this e‑waste that are already redefining the metals industry.

Further Reading

  1. https://www.unepfi.org/themes/climate-change/new-research-identifies-remaining-global-carbon-budget-for-twelve-main-industries/
  2. https://www.mckinsey.com/industries/metals-and-mining/our-insights/aligning-the-value-chain-to-decarbonize-plastics
  3. https://www.mgsrefining.com/blog/how-gold-powers-todays-technology/
  4. https://www.wri.org/insights/how-mining-impacts-forests  
  5. https://news.mongabay.com/short-article/mining-drove-1-4m-hectares-of-forest-loss-in-last-2-decades-report/
  6. https://www.iucn.nl/en/blog/minings-impact-on-forests-a-growing-threat-to-biodiversity-and-climate/  
  7. https://ewastemonitor.info/electronic-waste-rising-five-times-faster-than-documeted-e-waste-recycling-un/
  8. https://prism.sustainability-directory.com/scenario/e-waste-as-urban-mining-resource-potential/    
  9. https://www.miningnewswire.com/q2-saw-tech-industry-demanding-more-gold-wgc-data-shows/
  10. https://www3.weforum.org/docs/WEF_A_New_Circular_Vision_for_Electronics.pdf