Rio Tinto's flagship iron ore division has confirmed an 8 per cent slump in exports, while the miner conceded its Mongolian copper project could suffer cost blowouts of up to $US1.9 billion ($2.69 billion).

Rio had expected an underground expansion of Mongolia's Oyu Tolgoi mine to cost $US5.3 billion when it was approved in 2015, but has known for the best part of a year that the project would be delayed.

Rio won't know the exact scale of cost and schedule blowouts until late 2020, but current estimates suggest the project will cost between $US6.5 billion and $US7.2 billion.

Those cost estimates do not include the cost of building a power station for the mine, which could cost Rio and its Oyu Tolgoi partners a further $US1 billion.

The cost blowouts are much bigger than analysts had expected, and have been partly caused by weaker than expected rock forcing changes to the mine plan.

Some close to the project believe shortcomings in project management have also contributed to the cost and schedule blow outs.

RBC had earlier this year modelled a $US500 million blow out at Oyu Tolgoi, while Deutsche had estimated a blow out of $US800 million.

Goldman Sachs analyst Paul Young had assumed a $US1.1 billion increase for a total cost of $US6.4 billion, but even that is below the bottom end of Rio's updated cost estimate range.

Sustainable first production of the mine could be up to 30 months late, although Rio said it had built eight months contingency into that estimate.

Originally scheduled for the early weeks of 2021, Rio said first production could now come between May 2022 and June 2023.

Rio flagged it may record an impairment against Oyu Tolgoi when it reports its half year financial results next month.

''Rio Tinto is reviewing the carrying value of its investment in the project and will announce if any changes are required in the half year results on August 1, 2019,'' the company said.

The cost and schedule blow outs will not be popular with the Mongolian government, which is a 34 per cent shareholder in the Oyu Tolgoi mine, and has previously expressed concern about the rising costs.

Mongolia is keen for the mine's copper, gold and silver to start bolstering its budgets as soon as possible, and the latest delays mean the nation's young democracy will have to wait longer than desired for the revenues to flow.

Even Rio's ability to give shareholders a ''definitive estimate" of the updated cost and schedule for the mine has been delayed.

Rio told shareholders last year it would publish such a definitive estimate by mid 2019.

By February that timeline for the definitive estimate had been pushed back to late 2019.

On Tuesday Rio said it would be the second half of 2020 before it would be able to provide a definitive estimate of costs and schedule, suggesting a major rethink of the mine design is underway.

''All options under consideration present a pathway to sustainable first production, and have different cost and schedule implications,'' said Rio in a statement on Tuesday.

''Significantly more work is required to complete the final assessment."

While Oyu Tolgoi is Rio's most important growth project, its Australian iron ore division remains its biggest money-spinner, generating more than 66 per cent of its underlying earnings in 2018.

Iron ore could deliver an even higher proportion of Rio's earnings in 2019, given prices for the steelmaking ingredient have been trading near five year highs in the past month.

The type of iron ore Rio typically sells from Western Australia, ore with 62 per cent iron, was fetching $US119.25 per tonne on Monday.

That price includes the cost of freight, which is currently about $US9.55 per tonne.

Rio said it had received an average price of $US85.30 (excluding the cost of freight) for its iron ore over the past six months.

Iron ore price strength has been largely driven by supply curtailments in Brazil, but Rio's exports have also been soft, with the company confirming on Tuesday it had shipped 85.4 million tonnes in the three months to June 30.

That was about 8 per cent less than in the first half of 2018.

Analysts at UBS had expected Rio to report iron ore shipments from Australia of 85.2 million tonnes in the three month period, including tonnes owned by joint venture partners like Gina Rinehart's Hancock Prospecting.

Rio's mines produced 79.7 million tonnes of iron ore in the three months to June 30; below the 81.2 million tonnes expected by UBS.

The weak performance at the mines appears to relate to the operational problems in the Pilbara that Rio disclosed to the ASX on June 19.

Rio began the year promising to ship between 338 million and 350 million tonnes of iron ore from Western Australia.

By April bad weather and a fire at a port had forced Rio to cut its iron ore export guidance to a range between 333 million and 343 million.

Guidance was cut a second time in June to between 320 million and 330 million tonnes, after mine sequencing issues emerged in the Brockman hub.

If Rio delivers at the bottom end of its latest guidance range, it will be the company's weakest year of iron ore exports since 2015.

Rio said major rail maintenance would occur on its iron ore railways in coming months, but that work was already factored into its export guidance.

The weaker exports have driven up Rio's unit costs from a range between $US13 and $US14 per tonne of iron ore, to between $US14 and $US15 per tonne.

UBS expects Rio to ship 324 million tonnes of iron ore from Australia in 2019.

Rio has also downgraded expectations for its Canadian iron ore business, which was expected to produce between 11.3 million and 12.3 million tonnes in 2019.

The Canadian division will now produce 10.7 million to 11.3 million tonnes in the year.

Meanwhile, Rio confirmed it had completed the sale of its stake in Namibia's Rossing uranium mine.

The Iranian government's ownership of a stake in Rossing had complicated life for Rio, which was forced to make regular, detailed assurances to US authorities about the nature of its involvement with Iran.

Published Date:2019-07-16