Metso Insights Blog Mining and metals blog Improving flotation control with rapid slurry measurement
Mining Metals refining
Dec 19, 2017

Improving flotation control with rapid slurry measurement

Reflectance spectrum measurement offers a relatively inexpensive and simple method to reduce the sampling interval of centralized X-ray fluorescence (XRF) on-line analyzers. This emerging technique can improve both process monitoring and automatic process control performance in flotation circuits.

Flotation circuits are typically controlled using on-line elemental assays obtained from the slurry lines. In base metal applications this analysis is typically performed using centralized XRF analyzers. Outotec solutions based on XRF technology include Courier 5 and Courier 6 analyzers.

Reflectance spectroscopy, on the other hand, is an emerging technique that is used to obtain more frequent measurements from slurries. It is especially beneficial in rougher and final concentrate streams where grades are on a suitable level to allow accurate measurement. With reflectance spectroscopy, the sampling interval of the elemental assays can be radically reduced. When this measurement option is included in the Courier line, the typical sampling interval of 10 to 25 minutes is reduced to just 10 seconds. This virtually continuous assay information reveals rapid changes in grades and improves the performance of automatic process control solutions.

Reflectance spectrum measurement

Reflectance spectrum measurement is based on the mineral-specific absorbance of light in the visible and near-infrared (VNIR) wavelength range. Because the system is relatively inexpensive and simple, it is possible to measure simultaneously in several slurry lines with a very short sampling interval. However, the information contained in the VNIR spectrum is difficult to interpret, and the spectrum is also affected by other slurry properties. As a result, the rapid reflectance spectrum analysis needs to be regularly calibrated using an XRF analyzer.

The rapid measurement device consists of measurement probes installed in the secondary sampler of the Courier multiplexer unit, and an analyzer unit attached to the side of the multiplexer. Due to the centralized structure of the Courier analyzer, all the important slurry lines are available for practically continuous reflectance spectrum measurement. The structure of a typical installation is shown in Figure 1.

Reflectance spectrum analyzer installed on a Courier multiplexer unit
Figure 1. Reflectance spectrum analyzer installed on a Courier multiplexer unit.

After the spectrum of the slurry is measured, a calibration model is used to translate the raw measurement data into elemental content estimates. These values are then available for the distributed control system (DCS) in the same way as the traditional Courier analyzer assays.

Ensuring measurement accuracy

The calibration of the reflectance spectrum measurement is based on the XRF samples of the Courier on-line slurry analyzer. A static calibration model can capture the relationship between reflectance spectra and the elemental contents present in the calibration data set. However, in a continuous process such models seldom remain valid for long. Several other factors affect the reflectance spectra besides the direct changes in the elemental contents of the slurry, including particle size distribution, changing mineralogy, and solids content. It has been shown that in real process conditions reflectance spectrum analyzers require frequent and regular calibration.

To address this issue, recursive updating of the calibration model is used. The idea is that every time an XRF assay is available, the reflectance spectrum measured at the same time is selected and the calibration model is updated using this new calibration sample pair. Older samples are gradually forgotten in order to keep the model up to date.

Continuous elemental content measurements

The main benefit of reflectance spectrum analysis is the drastic reduction in the sampling interval of the elemental assays. Figure 2 shows an example of grade oscillations detected in a copper rougher concentrate slurry line. In this case, because the period of the oscillations is in the range of three minutes they cannot be detected by XRF measurement. However, when the rapid measurement option is combined with the Courier analyzer, the periodical changes in the grade are clearly visible.

Fast oscillations detected by the rapid measurement option
Figure 2. Fast oscillations detected by the rapid measurement option.

Drastic drops in the concentrate grade often indicate problems in the upstream process. With rapid measurement these changes can be detected as they happen and corrective control actions can quickly be made to restore normal operation.

Rapid grade drops can be detected immediately
Figure 3. Rapid grade drops can be detected immediately.

Rapid measurement can be applied in advanced process control (APC) solutions. For example, in automatic grade control solutions rapid measurement improves control performance because the execution interval of the controller can be reduced accordingly. This means that the controller will react more quickly to changes in grades and the process will stay closer to the target values with smaller control signals. As an example, a copper rougher grade control improvement based on rapid measurement has been reported by Kejonen et al. (2017).

Conclusions

Reflectance spectroscopy provides a relatively inexpensive and simple method to reduce the sampling interval of a centralized XRF on-line analyzer. An automatic calibration scheme is utilized where the periodic on-line XRF assays are used as the calibration source for the reflectance spectrum measurement. This ensures that the calibration model is always up to date regardless of changes in the particle size, solids content, or mineralogy of the slurry. The resulting continuous grade measurements can be used to improve both process monitoring and automatic process control performance.

References

Kejonen I, Haavisto O, Martikainen J, Suontaka V and Musuku B: Improving grade control efficiency with rapid on-line elemental analysis. In Proceedings of Flotation ‘17, November 13-16, 2017, Cape Town, South Africa, 2017

Mining Metals refining