Small but Economically Significant Financial Gains in Recovery are Possible…

On a foundation of modern process mineralogy, there is a good basis for identifying flowsheet changes in a concentrator which would have a likelihood of achieving and sustaining grade or recovery improvements. The remaining problem is that concentrator data are autocorrelated in the short-term and have a high degree of variance. This makes the measurement and proof of small recovery increments difficult. The next challenge in concentrator optimisation is in proving that the performance gain from a flowsheet change was attained and is sustainable. The use of an undesigned plant trial format otherwise runs the risk of inconclusive measurements at the risk of project capital expenditure and team reputation (Napier-Munn, 1995; Lotter et al., 2010).  The other risk that must be managed is the status of the sampling, generally for metal accounting.  Unless this is at best practice, the baseline from which the operations staff are working to show the performance improvement is shaky, risking the plant trial outcome.  It is a definite recommendation that, prior to the design and implementation of a plant trial, a check is made on the existing sampling arrangements.

Use of an on-off block model to switch the plant operation from condition A to condition B (for example, collector A vs collector B in a flotation plant) in a statistically-designed test format with disqualification of the first few days of operation of the new condition breaks the short-term autocorrelation, and makes the differences in the test data clearer to see.  Significance testing with the Analysis of Variance (ANOVA) follows to determine the reality of the observed difference.  In a designed plant scale trial using this format at the Raglan concentrator, northern Québec, where sodium isobutyl xanthate was used instead of potassium amyl xanthate based on laboratory scale recommendations, these recovery gains were: Ni: 1.04%, Cu:1.01%, Pt: 2.67% and Pd: 1.79% in absolute terms.  These gains were proven to be significant at better than the 99% confidence level (Lotter et al., 2010).  This outcome validates the format of the designed plant trial.

Let Flowsheets assist you in designing and assessing your plant trial……

Lotter, N.O., Di Feo, A., Kormos, L.J., and Comeau, G., 2010.  Design and measurement of small recovery gains: A case study at Raglan concentrator, Minerals Engineering, 23 (2010), pp.  567–577.

Napier-Munn, T.J., 1995. Detecting performance improvements in trials with time-varying mineral processes – three case studies. Minerals Engineering 8 (8), pp. 843–858.

Lotter, N.O., and Napier-Munn, T.J., 2018. The Value of Incremental Performance Improvement in Concentrators – How to Secure and Quantify Small Gains, proc. Extraction 18, Ottawa, pp. 2847-2855.