Cost savings through improved ore recovery using automated core logging

CIM Montreal 2007
Richard A Williams,
Abstract Cost savings through improved ore recovery using automated core logging and ore sorting spectrometers.

Dr Gavin Hunt and Richard Williams, Spectra-Map Ltd, UK

Summary
In order for mines to operate profitably throughout their life it is important to monitor and maximise ore recovery. A common problem facing many mines today is the underestimation of clay type and distribution in ore at the mine face, blast hole or in the ore stockpile. Too much clay entering the heap leach ore stream for example causes a dramatic reduction in heap leach efficacy and/or increased leaching times. The end result is poor leaching performance and much reduced metal production. This is a problem that faces the largest copper producers, such as Escondida, through to the smaller gold operations in China for example.

Poor clay identification and quantification is attributable to many factors, including costly and slow off-site mineral analysis techniques such as XRD, XRF and petrology, the visual limitations of the core logging geologist and a lack of suitable low cost, high throughput instrumentation to supplement the traditional methods.

In recent years field-portable infrared reflectance spectrometers have enabled mine geologists to get real-time mineral identification from chip samples, blast-hole core and stockpile ore samples. However these instruments are point sampling devices that are limited to only a few hundred readings per day. These limitations mean that although correct mineral identification can be made there are huge problems related to accurate quantification.

To get round these problems a new generation of high-speed, non-contact, imaging spectrometer was developed by Spectra-Map Ltd. - the SpecCam. This device, like the point instruments described above, operates in the Short Wave Infra Red (SWIR) region, but crucially can collect millions of data points per day through its novel imaging technology.

Each point, or pixel from the SpecCam image data, contains spectral information that allows a wide variety of minerals types, including the smectite and illite clays, and phyllosilicates such as talc and kaolin to be identified. Because all the ore can be imaged a much more reliable analysis can be made. Hence, changes in mineral composition - related to variations in phenocrysts, macrocrysts, veining or wall rock - is possible. More importantly, for the first time very accurate % abundances for the different minerals can be obtained using an image-based modal count technique.

The novel modular design of the camera, plus its extremely flexible high speed filter, mean that the SpecCam can be easily installed on different equipment, each one suited to different ore recovery problems. For example the SpecCam can be used on an automated portable logging frame to measure drill core, RC chips, hand samples or powder, or alternatively installed on a Geotek multi-sensor core logger for high-throughput, automated logging of drill or blast hole core using co-registered multi-sensor data.

An exciting recent development has been the installation of the SpecCam on CommoDas ore sorting equipment, used around the world for 24hour mill-feed monitoring and automatic ore sorting. Real-time ore sorting decisions can be made as material is monitored on a conveyor belt running at 3m s-1, from which the ore can be automatically sorted into high-grade, low-grade or waste streams for example.

The three installation options for the SpecCam mean that most mineral analysis and automated core logging or ore monitoring needs can be met at the mine. For the first time rapid and automatic mineral identification and accurate quantification can be made. All of which allow the mine geologist or metallurgist to make important decisions about the best way in maximising the ore recovery process.
Keywords: SpecCam, Ore sorting, Spectra-Map, Core logging, spectrometer
Full Access to Technical Paper
PDF version for $20.00
Other papers from CIM Montreal 2007