Rio Tinto Approves US$1.9 Billion Kangwinan Bauxite Project
Rio Tinto has approved US$1.9 billion to develop the Kangwinan bauxite project on Queensland's Western Cape York Peninsula, committing the group to a new mine, a 15-kilometre conveyor and expanded export infrastructure. First output is targeted for 2029 and full production for 2032.
The 23 October 2026 investment decision gives Kangwinan annual capacity of 20 million tonnes. That headline is real, but it is not a 20-million-tonne net expansion of Rio Tinto's bauxite system. The project will replace output from the aging Andoom mine and help compensate for the planned end of mining at Gove in 2029.
The number that matters is six million tonnes
Rio Tinto says Kangwinan will lift Weipa Operations' annual capacity from about 38 million tonnes to about 44 million tonnes. The net increase is therefore roughly six million tonnes, or 15.8%, even though the new production centre itself is designed for 20 million tonnes a year.
The difference is replacement capacity. Andoom is approaching the end of its life this decade, while Gove in the Northern Territory is expected to mine its last bauxite in 2029. Kangwinan keeps ore flowing through a regional system whose existing mines are depleting, while adding some incremental output at Weipa. Treating all 20 million tonnes as new group supply would overstate the effect on the market.
The gross Kangwinan capacity is still large. It is equivalent to approximately one-third of Rio Tinto's 2026 global bauxite production guidance of 58 million to 61 million tonnes. Yet even this comparison is a scale marker rather than a forecast of extra global production because the project is explicitly designed to replace retiring tonnes.
Measure | Rio Tinto disclosure | Agrilinkage reading |
Project capital | US$1.9bn | Includes US$150m approved in 2025 |
Gross new capacity | 20 Mtpa | About US$95 capital per annual tonne |
Weipa capacity | 38 to 44 Mtpa | Net increase of about 6 Mtpa, or 15.8% |
Schedule | First output 2029; full 2032 | Roughly three-year production ramp |
Company-modelled return | Approx. 22% IRR | Based on Q4 2026 third-party market forecasts |
A brownfield strategy with new logistics
Kangwinan is 15 kilometres south-east of Amrun, Rio Tinto's newest Weipa mine. Ore will move by a new 15-kilometre conveyor to expanded stockyards and the existing Amrun wharf. The wharf itself will be enlarged to handle larger capesize vessels, which the company expects to reduce freight costs and improve supply-chain efficiency.
That design matters because the investment is not confined to a mine pit. Conveyor capacity, stockyard availability, shiploading performance and vessel access will determine whether 20 million tonnes of nameplate mining capacity can become saleable exports. Reusing Amrun's existing corridor can reduce duplication, but it also makes the project dependent on the timely expansion and integration of shared infrastructure.
Rio Tinto's current Weipa operations include mines, processing facilities, shiploaders, two ports, power stations, a rail network and ferry terminals. Kangwinan extends that operating system rather than creating an isolated greenfield export chain.
The capital has a replacement purpose
The US$1.9 billion estimate includes US$150 million of early works approved in 2025. That earlier commitment represents about 7.9% of the total, leaving approximately US$1.75 billion, or 92.1%, within the newly approved programme. Rio Tinto classifies the project as replacement capital and says it is already incorporated in its medium-term capital guidance.
Dividing the total estimate by 20 million tonnes gives an indicative capital intensity of US$95 per tonne of annual nameplate capacity. This is not an operating cost and should not be compared directly with the realised price of bauxite. It is a rough project-scale ratio that includes mine and logistics infrastructure and does not capture the value of shared assets already in place.
The company expects an internal rate of return of approximately 22%, based on third-party market forecasts available in the fourth quarter of 2026. That is a company-modelled investment case, not an assured return. Construction cost, ramp-up, ore quality, freight savings and future bauxite prices will determine the realised economics.
There is a useful historical comparison. Rio Tinto said in its 2025 early-works announcement that Amrun cost US$1.9 billion and opened in 2018. Kangwinan now carries the same nominal headline cost, but the scopes, price years and use of existing infrastructure differ, so the equality should not be read as evidence that the two projects have identical economics.
Rio Tinto is resetting its northern Australian bauxite network
The replacement logic becomes clearer at Gove, where Rio Tinto says the final bauxite is expected to be mined in 2029. Kangwinan's first-output target falls in the same year. This is a handover strategy: new Queensland tonnes must begin arriving as Northern Territory supply leaves the portfolio.
Andoom is also approaching closure, while East Weipa stopped producing in 2024. The emerging Weipa system will therefore be more concentrated around Amrun, Kangwinan and, potentially, the adjacent Aurukun deposit. Rio Tinto agreed in September to acquire Aurukun from Glencore and Mitsubishi, but that transaction remains subject to government approvals and the company has not presented Aurukun as part of Kangwinan's approved 20-million-tonne plan.
Keeping those projects separate is important. Kangwinan now has an approved investment, a construction sequence and production targets. Aurukun is an agreed acquisition with a future approvals pathway still to be assessed. Combining their resource potential or treating Aurukun output as committed would move beyond the evidence currently available.
Jobs are concentrated in construction, not a new permanent workforce
Construction is expected to ramp up in 2027 and support more than 1,300 jobs. That is a construction-period figure, not a promise of 1,300 permanent positions. Over the longer term, Rio Tinto says Kangwinan will help sustain its existing residential workforce of around 1,500 people in Western Cape York as older mines close.
The wording signals continuity more than a large permanent employment expansion. The economic value for the region will depend on how much construction procurement reaches local suppliers, how workers transition from Andoom and other operations, and how long the expanded mine and port system supports activity after commissioning.
Approvals reduce one risk but do not remove delivery risk
Rio Tinto says environmental permits for Kangwinan were previously secured through the state and Commonwealth-approved South of Embley environmental impact statement. Early works began in 2025, including an accommodation camp and access-road development, so the project is further advanced than a newly announced concept.
Cultural-heritage and community obligations continue. The project was named Kangwinan at the request of the Wik Waya Traditional Owners, and Rio Tinto says it will continue consultation on cultural heritage, waterways and access routes. Existing environmental approvals do not remove the need to manage those commitments during detailed construction and operation.
Why the project matters beyond bauxite
Bauxite is the ore from which alumina is refined and aluminium is smelted. Rio Tinto's Australian chain links Weipa mines with the Yarwun and Queensland Alumina refineries in Gladstone, domestic smelters and export customers. A disruption at the mine or port stage can therefore travel downstream into refinery utilisation, metal production and maritime trade.
Rio Tinto's third-quarter operations review reported 44.2 million tonnes of bauxite production in the first nine months of 2026 and maintained full-year guidance of 58 million to 61 million tonnes. Kangwinan is a long-dated investment relative to those current flows, but it is designed to prevent the supply base from shrinking when Andoom and Gove retire.
What to watch next
The first test is construction mobilisation in 2027. The 15-kilometre conveyor, stockyard expansion and wharf modifications must progress in sequence with mine development. A delay in any one element could constrain the system even if the other components are ready.
The second test is the handover around 2029. Kangwinan's initial output must overlap effectively with Gove's planned closure and Andoom's decline. The third is the ramp to full production by 2032, when ore quality, equipment availability, shipping performance and customer demand will determine whether the asset reaches its stated capacity and return assumptions.
For the wider bauxite market, the main question is not whether 20 million tonnes of mine capacity has been approved. It is how much net saleable supply remains after retiring mines are removed from the system. Rio Tinto's own figures point to a measured increase at Weipa, supported by a much larger replacement programme.
Methodology and sources
Agrilinkage based this analysis on Rio Tinto's 23 October 2026 investment announcement, its 22 May 2025 early-works disclosure, the company's 2026 third-quarter operations review and official Weipa and Gove operation pages. Calculations use the company's reported capital, capacity and production figures. Forward-looking costs, capacity, jobs, schedules and returns remain company estimates.
Comments