In the past, the lack of an easy, non-disruptive, economical measuring technique has meant that, in most cases, rock fragmentation size has not been defined in quantitative terms. Now, the Split-Online® Rock Fragmentation Analysis system provides an economical alternative to manual sampling and screening and an objective measure rather than subjective qualitative estimates.

Technical decisions; blasting costs; operational efficiency and productivity; equipment performance-all these can be related to optimum rock fragmentation. Profitability can be improved by optimizing the fragmentation of the run of mine ore to maximize throughput and performance of the crushing and mill circuits.

The most convenient location for a permanent camera is above a moving conveyor belt. This includes apron feeders feeding primary crushers, belts carrying crusher feed and product, belts carrying AG or SAG mill feed, belts carrying ball or rod mill feed, or other locations.

Validation and calibration samples from recent installations in September 2001.

Schematic from typical Split-Online fragmentation measurement system.

Samples of material for screening are collected each time a Split-Online system is installed. The purpose of this is to calibrate the fines correction for each system and make sure that the hidden material is appropriately accounted for. This involves stopping the conveyor belt after a sample has been recorded on video, removing and screening the belt material and comparing the screened results with the image processing results.

It is important to realize that the systems employ fairly standard cameras in that they can only see the surface. It is assumed that the largest material is visible on the surface and that fines, which normally fall to the bottom, are the only hidden material that needs to be estimated. Split-Online Rock Fragmentation Analysis system employes a fines correction which uses the surface material to estimate amount of hidden fine material. Over the past six years Split Engineering has collected more than 80 sets of sieve results and corresponding images from operations worldwide and this data set shows that the fines correction approach is reasonable.