Image credits: TecProMin
News Article

Solution to Uncertainty in Metallurgical Analysis: Part 1

By Alejandro Ehrenberg | Wed, 08/19/2020 - 16:41

Sampling systems experts Promimex and TecProMin, the Zacatecas' Mining Cluster and Mexico Business Publishing joined forces to organize a two-day e-course on applying the Theory of Sampling for achieving optimal metallurgical balances. The course was delivered by world-renowned expert Dr. Francis Pitard.

The e-course had its first session on Aug. 19. Dr. Pitard started off by surveying some basic elements of the theory of sampling. He said that when buying or installing a sampling system, one has to know the fundamentals. Dr. Pitard focused first on small scale variability. He pointed out that sampling technicians introduce all types of variability that will confuse them if it is not well understood.

To keep small scale variability under control, Dr. Pitard advised to start by optimizing the sampling protocol. This includes geologists orderly dealing with in-situ nugget effect. It also involves fundamental sampling errors that are negligible for base metals but are key for precious metals. Moreover, the sampling protocol must take grouping and segregation errors into account: different metals have different densities and properties.

After optimizing the protocol, one has to implement it. Important aspects of this include increment delimitation errors, increment extraction errors and increment weighing errors. Dr. Pitard went over the main ways to minimize these errors.

Lastly, it is important to preserve the sample’s integrity. To do this, Dr. Pitard mentioned the importance of increment preparation errors. He also noted the importance of contamination, loss, alteration, human, fraud and sabotage error. Moreover, he highlighted the analytical error, which includes aspects like scope versus principle, dissolution technique, contamination and losses and proportional interference. Dr. Pitard also explained variability on a large scale. He underscored the relevance of interpolation errors, periodic errors and increment weighting errors.

Once the main sources of bias had been explained, Dr. Pitard said that one has to prioritize which ones make the most sense to eliminate first. He advised to set up a matrix and establish priorities. The matrix should specify the effect of a given problem vs the cost of fixing it. If it is a small problem with a costly fix, then nothing has to be done and vice versa. He emphasized the importance of having the support of top management to undertake an improvement of an operation’s sampling methods and systems.  

Dr. Pitard continued his lecture by explaining how most sampling equipment currently in the market is biased by design. This is not due to fraudulent intentions, he said. The problem is that engineers are not trained in the Theory of Sampling. Therefore, they build machines that are flawed by design. He noted that there are a few good manufacturers in the market, including Santiago de Chile-based TecProMin.

When selecting a sampling system provider, Dr. Pitard said it is important to have the following points in mind. First, reliable sampling for metallurgical balance cannot be done on coarse material, so the most vulnerable area is the sampling feed that goes into the plant. Second, one must know the heterogeneity of the constituent of interest: mineralogical studies, heterogeneity tests and relevant geological information are key to this end. Third, one has to select equiprobabilistic sampling systems and correct mechanical sampling systems.

Dr. Pitard ended his lecture by cautioning that there are systems that will never work and cannot be fixed. For example, grab sampling is hopeless. It never works. Instream stationary cutters are a little more acceptable, Dr. Pitard said. But they are not ideal for carrying out metallurgical balances. Another enemy of sampling correctness are multi-cutters. In sum, there are many systems that may work for process control but do not work for metallurgical balance. Dr. Pitard noted that the way to avoid these suboptimal methods is to become educated on the Theory of Sampling and to trust the leaders in the sampling field, such as TecProMin.

Photo by:   TecProMin
Alejandro Ehrenberg Alejandro Ehrenberg Journalist and Industry Analyst