With proven one-to-one direct scale up from laboratory testwork, installing IsaMill™ technology significantly reduces project risk.

### How Scale Up Is So Accurate

Accurate testwork and scale up is essential for project performance. So a key feature of IsaMill™ technology is that all full production IsaMills™ can be accurately scaled up from testwork in a 4-litre (M4) laboratory mill using as little as 12.5kg of sample. The excellent scale-up of the IsaMill™ is a result of a combination characteristics including:

- using media of the same specification and size in the full scale mills and the test units;
- in both full scale and test IsaMills™ media and slurry is agitated in a similar method; and
- the layout of the mill in the full scale and test units is the same, the mill is horizontal and has multiple grinding chambers.

IsaMill™ scale-up is accurate in terms of power consumption and product size distribution (PSD)

- Power scale up is accurate because power is measured as power consumption at the agitator shaft to determine the specific energy requirements and size the mill; and
- PSD scale-up is accurate because of the horizontal IsaMill™ layout, which eliminates any short circuiting, and the classifying action of the product separator. Both features are common to the laboratory and full scale units. The sharp IsaMill™ PSD is a big advantage for flotation and leaching operations. Alternative grinding technologies tend to produce wider size distributions at full scale, since large scale screens or cyclones are less efficient than in laboratory tests. Even worse laboratory tests may even be done in batch mills, which can not indicate size distribution for a continuous discharge mill.

This accurate scale up is different than scale up for ball mills or other stirred milling technologies.

In ball mill laboratory tests 25mm balls in a small laboratory ball mill have different trajectories and interact differently with the shell lifters and ore particles than in a large production ball mill. Additionally while techniques like the Bond Work Index and "scale up factors" are useful for coarse grinding, they can significantly underestimate grinding needs below 100 microns.

Similarly Tower Mill laboratory tests typically use much finer media than can be used in full scale operation and testwork is done in "batch" mode.

In contrast, an ore particle in an IsaMill™ encounters small media particles (say 2-3mm) stirred at high speed, irrespective of the shell size. Our experience has shown very accurate scale up from laboratory IsaMill™ tests, so long as the sample is representative. The mill scales up accurately on both a power and product size distribution (PSD) basis.

A test program must rigorously account for energy measurement, size measurement, media size, classification and batch testing versus continuous operation. While Glencore Technology uses rigorous scientific design methods, and confirms the scale up on every plant. IsaMill™ installations always meet design.

### Our Approved Laboratories Get Testing Right

While scaleup on the laboratory M4 is very accurate, care has to be taken with several factors, so we insist on using trained and approved laboratories. For instance:

**Power Measurement:**the only accurate way to measure mill power is to directly measure power or torque input to the shaft. Indirect measurements are inaccurate and will underestimate mill power (Nesset, J.E. et al, Assessing the Performance and Efficiency of Fine Grinding Technologies, 38th Annual Meeting of the CMP, Canada, January 2006). It is important to measure both "net power" – ie. the amount that achieves useful grinding, and "gross power", which includes losses in the motor, gearbox and transmission. Net power scales directly from IsaMill™ laboratory tests.**Particle Size measurement**is not a trivial matter, especially for ultrafine grinds because of the exponential increase in grinding power. Going from 9 micron to 7 micron can increase grinding power by 50% – so sizing must be measured accurately and consistently.**Classification, Continuous vs Batch, Particle Segregation:**Scale-up methods must accurately predict real-plant continuous operation and classification effects. The test procedure must ensure steady state and no "hold-up" of coarse particles in the test mill – a particular issue with vertical mills. IsaMill™ test procedures use the same continuous feed and classification as full scale, and specifically confirm that steady state has been reached.**The media size and type:**For any given ore, grinding energy efficiency is determined primarily by media size and type. Glencore Technology's procedures use the same size media as the full scale IsaMill™ – essential for accurate scale up. These procedures test first with a benchmark media and then with other media types if nominated by the client.

So while your plant or laboratory is the best place to prove the liberation and process advantages of an IsaMill™, once you have selected your target grind size the final mill scale up work should be done in an authorized IsaMill™ laboratory, using our standard mill and measuring equipment. This has been proven to give accurate scale up.

### Testwork Facilities

The best way to accurately size an IsaMill™ is by testwork on a representative feed sample. Glencore Technology has its own facilities and has also authorized a number of independent laboratories to conduct IsaMill™ tests. IsaMill™ test procedures consistently achieve one-to-one scale up, because the test rig configuration is the same as a full scale mill – continuous feed, internal classifier, same grinding action and same grinding media.

The size of sample we need depends on the SG of the solid in your feed. It varies between just 12.5 and 22.5 kilograms. For a low SG, it's 12.5kg. For an SG of 3 (e.g. Copper), it's 15kg. For an SG of 3.5 (Copper, Lead, Zinc) it's 17.5kg. For an SG of 4 (Pyritic Gold) it's 20. And for an SG of 4.5 (Magnetite) you could expect 22.5kg.

Initial testing uses an M4 IsaMill™ to accurately determine the relationship between specific power draw and product size – the "signature plot". Alternatively, fully self-contained M20 plant pilot rigs are available for on-site campaigns. M20 pilot rigs can determine power requirements while investigating media consumption and downstream processing performance on real plant streams.

See testwork facilities on a map.