BHP and Amazon Pilot a Credit Market for Low-Emissions Copper
A pilot between BHP and Amazon would let buyers purchase emissions attributes from Escondida copper separately from the metal itself — the model that already reshaped green power markets.
BHP and Amazon have established what they describe as an industry-first pilot to buy and sell credits representing lower greenhouse gas emissions attached to copper concentrate and cathodes produced at Escondida, the world's largest copper mine, which runs on renewable electricity and desalinated water.
BHP and Amazon have set up a pilot programme to buy and sell credits representing the lower greenhouse gas emissions associated with copper concentrate and cathodes produced at Escondida in Chile, the world's largest copper mine. The two companies describe the arrangement as an industry first, and it borrows a structure that will be familiar to anyone who has followed renewable electricity procurement over the past decade: separate the environmental attribute from the physical commodity, and let each trade on its own terms.
Escondida is the natural place to try it. The mine sources its electricity demand from renewable energy and runs on desalinated water rather than drawing on the scarce freshwater of the Atacama — two of the biggest levers on the emissions and resource footprint of a large open-pit copper operation. That gives BHP a physically differentiated product. The problem, until now, has been that the copper market does not price it.
Why copper cannot simply be sold as "green"
Copper concentrate and cathode are fungible industrial inputs. Concentrate from Escondida is blended with concentrate from elsewhere on its way to a smelter; cathode is drawn into rod, wire and tube by fabricators who buy on grade and price, not provenance. By the time copper reaches a data centre busbar or a transformer winding, its origin is untraceable in any practical sense.
That is precisely the problem renewable energy certificates were invented to solve. Electrons on a grid cannot be labelled either, so the market created a separate instrument — a certificate attesting that a megawatt-hour of clean generation was added somewhere — and allowed buyers to purchase it independently of the physical power. A credit tied to the emissions intensity of a tonne of copper works on the same logic: the metal flows through the ordinary supply chain, while the attribute is sold to a buyer who wants to claim it.
For a miner, this is a route to monetising decarbonisation spending that the LME price will never reward. For a buyer, it is a way to act on supply-chain emissions that sit far upstream of anything it directly controls.
What Amazon gets out of it
Amazon's exposure to copper is indirect but enormous. Data centres, substations, transmission connections, grid-scale batteries, electric delivery vans and charging infrastructure are all copper-intensive, and the build-out continues to accelerate. That copper consumption lands in the hardest category of corporate emissions accounting — the upstream, supplier-side footprint that a company neither owns nor operates and usually cannot measure with confidence.
A hyperscaler can sign power purchase agreements to deal with the electricity it uses. Dealing with the embodied carbon in the metal, concrete and silicon it buys is far harder, because the emissions occur several tiers up the chain at companies with no contractual relationship to the buyer. A credit purchased directly from the mine collapses that distance. It also gives Amazon a demand signal to send, which is arguably the more consequential part: a repeat buyer of emissions attributes at scale creates a revenue line that justifies the next renewable retrofit or desalination investment at a mine site.
The pilot was reported by International Mining.
The verification question sits at the centre of it
Everything about whether this becomes a market rather than a one-off depends on how the credits are measured, certified and retired. Several questions will determine that, and none of them has a settled industry answer yet.
- The baseline. A credit representing "lower" emissions requires a reference point — an industry average intensity, a regional benchmark, or a site-specific historical figure. Each choice produces a different credit volume from the same tonne of copper.
- Boundary of the calculation. Mining and concentration sit at one end; smelting, refining and transport at the other. Concentrate and cathode are at different processing stages, so a single methodology has to handle both without double counting.
- Double claiming. If BHP sells the attribute, BHP cannot also claim the reduction, and the physical copper buyer cannot claim it either. Registries exist to enforce that in power markets; copper has no equivalent infrastructure.
- Additionality. Escondida's renewable supply and desalinated water are already in place. Critics of attribute markets argue credits from existing assets fund nothing new. Supporters argue the revenue is what makes the next project bankable.
- Third-party audit. Without independent verification against a published standard, the credit is a bilateral claim between two companies rather than a tradable instrument.
Everything about whether this becomes a market rather than a one-off depends on how the credits are measured, certified and retired.
Where this fits in BHP's copper strategy
BHP has spent years positioning copper as its growth commodity, on the argument that electrification, grid investment and data centre construction create structural demand that supply cannot easily meet. An attribute market is a second-order play on the same thesis. If the buyers driving copper demand are the same companies under the most pressure to report supply-chain emissions — utilities, automakers, hyperscalers — then the miner with the lowest-intensity tonnes has something to sell beyond the metal.
A credit stream is also structurally attractive compared with a physical premium. Premiums get negotiated away when the market loosens. A separately contracted attribute can be sold to a buyer who never takes delivery of the copper at all, which widens the pool of potential counterparties well beyond smelters and fabricators.
The catch is scale. One pilot between two companies does not create price discovery. A credit only has a market price when there are multiple sellers offering comparable instruments and multiple buyers bidding for them under a shared methodology. Until then, whatever changes hands here is a negotiated figure, not a market clearing level.
How the two companies' shares stand
Neither side is a pure-play on this. In the most recent session before publication, with markets closed as of 20:00 GMT on 25 September 2026, BHP last traded at 84.97, down 0.14% from a previous close of 85.09, within a day range of 84.20 to 85.28. Amazon last changed hands at 249.67, up 0.12% from 249.38, having traded between 247.18 and 250.88.
Both moves were smaller than the broader tape. The S&P 500 tracker closed at $771.35, up 0.54%; the Nasdaq 100 tracker finished at $744.50, up 0.46%; and the Dow tracker ended at $517.49, up 0.94%. A pilot programme of undisclosed size is not a share-price event, and the quotes should be read as context rather than as a reaction.
What would show this is working
Three things are worth watching. First, whether a recognised standards body or registry adopts a methodology for copper emissions attributes — that is the step that converts a bilateral deal into an instrument others can replicate. Second, whether other major copper producers announce comparable programmes, since a single-seller market is not a market. Third, whether buyers outside technology follow: automakers, cable manufacturers and grid operators consume copper in volumes that dwarf any individual data centre campus, and their participation would be the real test of demand.
If those pieces arrive, copper will have done what aluminium, steel and power have each attempted with varying success — built a parallel market for the carbon content of the material. If they do not, this remains a well-constructed bilateral arrangement between two companies that both had reasons to want it.
Key facts
- Participants: BHP and Amazon, in what both describe as an industry-first copper pilot
- Product covered: Credits for lower greenhouse gas emissions tied to copper concentrate and cathodes
- Source asset: Escondida, Chile — the world's largest copper mine, on renewable electricity and desalinated water
- Last close (25 Sep 2026, 20:00 GMT): BHP 84.97 (-0.14%); AMZN 249.67 (+0.12%)
Frequently asked questions
What exactly are BHP and Amazon trading in this pilot?
They are buying and selling credits that represent the lower greenhouse gas emissions associated with copper concentrate and cathodes produced at BHP's Escondida mine. The credit is an environmental attribute sold separately from the physical metal, so the copper itself still moves through the normal supply chain to whichever smelter or fabricator buys it.
Why is Escondida the mine chosen for this?
Escondida is the world's largest copper mine and it sources its electricity demand from renewable energy while running on desalinated water rather than freshwater. Those two features lower the emissions and resource footprint of its production relative to conventional operations, which is what creates a differentiated attribute BHP can certify and sell.
Why can't BHP just charge a premium for low-carbon copper instead?
Copper concentrate and cathode are fungible. Concentrate is blended en route to smelters and cathode is drawn into wire and tube by fabricators buying on grade and price. Provenance is effectively untraceable downstream, so the physical market offers no reliable mechanism to reward lower emissions intensity in the price of the metal.
Why would Amazon buy copper emissions credits?
Amazon consumes copper indirectly and at scale through data centres, substations, grid connections, batteries, electric vans and charging infrastructure. Those emissions sit several tiers upstream in its supply chain, where it has no direct operational control. Buying attributes from the mine itself is a way to address that category and signal demand to producers.
What determines whether these credits are credible?
The methodology. A credible instrument needs a defined emissions baseline, clear system boundaries covering mining through refining, registry infrastructure preventing the same reduction being claimed twice, and independent third-party verification against a published standard. Without those, a credit is a bilateral claim between two companies rather than a tradable market instrument.
How did BHP and Amazon shares close most recently?
As of the last trade at 20:00 GMT on 25 September 2026, BHP finished at 84.97, down 0.14% from a previous close of 85.09. Amazon closed at 249.67, up 0.12% from 249.38. Both moves trailed the broader market, where the S&P 500 tracker rose 0.54% to $771.35.
Sources
- BHP, Amazon establish ‘industry-first’ pilot for lower emissions copper market — International Mining
