Shipping iron ore is easy. Knowing how much iron you have actually shipped is another matter - one that keeps both the exporters and the buyers awake at night because it is the iron content, not the raw tonnage, that determines the final price.
Tricky as that might sound, it all comes back to the fact that every iron ore mine is different and most mines also have different grades of iron
in the same pit.
In one corner the ore might grade as high as 60 per cent iron, whereas another part of the same orebody might grade 57 per cent.
This leads to that critical question: how much iron is actually being sold?
Welcome to the specialised world of measuring the iron content of iron ore.
It is the starting point of a process that has been confusing outsiders for decades as they attempt to put a value on a tonne of iron are.
The problem is that iron ore is traditionally sold in 'dry metric tonne units'.
A DMTU is one per cent of the iron contained in a tonne of ore, excluding moisture. Science enters the business in measuring how much iron is being shipped.
The conventional measurement system is a messy, wet assay carried out at both the export port and the import port.
The buyer and seller compare assays, haggle and finally agree on a price. Some countries prefer this tried-and-true route, and some accept the assay from the port.
Australia has been leading an effort to modernise the process, opting for an x-ray fluorescence (XRF) measuring technique several years ago under the system of international standards (ISO 9516), which rapidly measures twenty elements.
It provides great accuracy for 19, but not always for the 20th element, the one that counts most - iron.
That's why for more than 20 years researchers at CSIRO have been looking for a better XRF method to accurately measure the iron and one that will eventually become a new internationally accepted standard under the ISO system.
"XRF is a very quick way of measuring ore," says Sally Birch, who is leading the work for CSIRO 's Minerals Down Under Flagship.
"In a matter of minutes you can make up a sample (in the form of a glass-like bead) and get a very accurate XRF analysis report. The problem is that the iron measure is not regarded as good enough for a reference method (in pricing).
"The problem is caused by iron being the major element in iron ore, and iron being a heavy absorber of x-rays. It is also highly subject to curvature of the glass bead. If the bead is curved in the slightest way it massively affects the iron value."
The technique showing the most promise is to add terbium, a rare earth, to the glass bead as an internal reference to produce a sample that is unaffected by curvature.
The original trials used 11 per cent terbium oxide in the sample disk, but that has now been cut to a two per cent mix, which is a major cost saving.
"The next big challenge is to achieve ISO acceptance, and that is a process which can take several years," Birch explained.
"If we can achieve international acceptance we will have made a major improvement in measuring the iron content of iron ore, and will have improved the process of determining how much iron is actually shipped."
* This article originally appeared in CSIRO's Process Magazine.