I have done the same in several mines. Lean tools and Six Sigma work well in the mining environment.

I am very interested to know. What you mean by the concept of "recovery rate". In what units it was measured and in what way.

What keeps surprising me with this kind of article is that this is often presented as if this what breakthrough methodology/technology (i.e. Lean Six Sigma) while there has been many similar case studies published over the last quarter of a century.

While the name of the method and thought framework might have changed over the years, this is still using statistical analysis to know/understand the process and how the feed characteristics - plant - reagents interact.

By looking at their graph I make the interpretation that they mean the usual recovery meaning that we all know and it is in % in that graph. Look at the bar graph at the end of the page. They have the recovery in %some people like to use the ton/day (for example) and then multiply it by the grade and then by the recovery and they get ton metal/day or other unit, which is a rate, in this case a metal recovery rate.

I agree, this is not new, however I would say that is still relatively unknown in the industry. The process improvement process using lean tools or similar is a structured approach to optimization and people don't like that: structure. Minitab is only a tool. With no methodology Minitab doesn't do much for process improvement

I'm guessing that may be 90% of people in mining doesn't know about this methods and that probably only 5% has used it based on all the places where I have been in 23 years. I may be wrong and biased but for sure there are more people who don’t use these techniques than people who use it.

You made an excellent point. It’s not that the tools are necessarily new or breakthrough but it is scary how few people use statistical tools when looking at plant performance. Due to the complexity of the flotation process the number of variables that can potentially affect the outcome is huge, utilizing statistical techniques to sift through the noise is critical. I still visit many sites who undertake evaluations by two simple surveys (one in ‘on’ condition one in ‘off’ condition) and draw a conclusion based on this despite both surveys having major differences in a variable such as head grade or pulp density. Hopefully this work at Saucito will encourage some young metallurgists out there to looking into statistical techniques and see if it can help in their operations!

I've been using statistical methods for years now, and while Minitab is a great tool, not everyone has access to it and with no statistical training it can be dangerous. Most of the Minitab functions can be done in Excel with a bit of tweaking it makes the data more transferable and able to be audited. Personally, in the example shown, I would have used a Weibull distribution rather than a normal distribution, but still, great find!

I partially agree with you. Training is needed or Minitab will give garbage. However the package is powerful and doing the same in Excel will be a titanic fight. Minitab is not only for basic statistics. When the tools are properly applied you can develop Gage R&R with ease, graphs variables in a way that is close to impossible in excel. Also you can develop capability analysis in Minitab, you can find the effect of operators over your performance, you can develop DOEs and optimize the parameters, find transfer functions, test for normality, etc. 

The best way to get statistical methods applied in plants (which I agree is badly needed) is for it to be part of the corporate culture. I previously worked at an industrial minerals producer and one of the strengths of the company was that there was a corporate mandate for decisions to be data-driven, and the use of statistics was interwoven into the documented procedures. We had a corporate statistician and training programs on how to use and apply statistics. In my opinion, until chief technical officers, chief metallurgists, GMs and Superintendents etc, start demanding this approach of everyone in their chains and providing the training & tools to do so, we will still see incorrect decisions made on statistically insignificant on/off or A vs. B comparisons as observed. When operators start insisting on statistical approaches, decision making will improve. No junior metallurgist is going to wake up on their own and decide to use statistics; it will have to be trained into their mindset, using brute force or torture if necessary. 

Any technological improvement gives benefit in Recovery, Grade, and reagent consumption. I don't agree with Thiophosphate promoters. We have tried all promoters of CYNAMIDE Company but failed.

Finally we tried Pyrite flotation from plant tailing. We could recover silver associated with pyrite. Pyrite flotation gave advantages in economics in recovering S, Zn, Pb, Ag. Even Bulk flotation with column flotation gives added advantages in Profits. Try to simplify flow sheet as far as possible. 

Decisions in the mining industry are not supported by data, in general. Even in companies that promote data driven decision. The support needs to come from the top. I can help to optimize a reagent suite using DOE. I also provide flotation training. The reagents you mention and the Pyrite flotation, all of that can be optimized