Laboratory Testing & General Mineral Processing Engineering

Laboratory Testing & General Mineral Processing Engineering

  • To participate in the 911Metallurgist Forums, be sure to JOIN & LOGIN
  • Use Add New Topic to ask a New Question/Discussion about Mineral Processing or Laboratory Work.
  • OR Select a Topic that Interests you.
  • Use Add Reply = to Reply/Participate in a Topic/Discussion (most frequent).
    Using Add Reply allows you to Attach Images or PDF files and provide a more complete input.
  • Use Add Comment = to comment on someone else’s Reply in an already active Topic/Discussion.

Determining Sample Size per Particle Size (PSD) (3 replies and 1 comment)

Y
Yvette314
5 years ago
Yvette314 5 years ago

Good day - I want to determine the sample size needed for a run-of-mine. I want to determine the sample mass needed by particle size. I've created a tool determining the variance between the size fractions for the RoM PSD, and then based on that, and with a 10% relative error, how many kg of sample is needed per size fraction for the sample to be representative. However, I don't think it's correct. Does anybody have advise/tool they use to perform a similar exercise?

Mike
5 years ago
Mike 5 years ago
Like

Normally it is worked the other way.  As particle distribution is a function of the ore characteristics; such as fracture, clevage, mining method and such, trying to determine even what particle distribution you will get before testing is nearly impossible.  Also the crushing and grinding used will impact the processed size distribution.

For this reason iterative testing is often used where bulk samples are taking and run through a pilot circuit to determine the best method of processing to maximize the grade and recovery of the desired components.

Y
Yvette314
5 years ago
Yvette314 5 years ago
Like

Hi - Thank you for your reply. It is very interesting. I have the RoM PSD -  I want to take a sample to perform additional test work on that sample.

For particulate materials, the primary criterion for establishing the required sample size is that all kinds (sizes, shapes, etc.) of particles should be adequately represented in the sample. Because particles are discrete entities, sample size is thus dictated by the number size distribution of the material being sampled. Consider, for example, a system of quartz particles (specific gravity 2.65) that contains 10% by weight of 1-cm particles. Each 1-cm particle would weigh about 1.4 g, so that a sample weight of 14 g would be necessary to ensure a reasonable likelihood of containing even one of these particles. Obviously, this weight would be quite inadequate because there would be a high probability of taking such a sample and finding none of those particles (0%) or two of them (20% by weight). If the sample size were increased tenfold, to 140 g, we would expect to find 10 of the 1-cm particles in any sample, and the errors introduced by the chance inclusion of one extra or one fewer would be correspondingly reduced; that is, a sample containing only 9 of them would analyze at 8.9% rather than 10%. Sample size, then, should be based on statistical criteria such that the errors introduced by random variations in the numbers of different particles included are acceptably small.

So I was using the PSD information, etc. and statistics (mean particle size, relative error, etc.) to set up a table to determine the sample size so I can use that sample to perform additional test work on.

Example attached!

untitled
https://www.911metallurgist.com/wp-content/uploads/2019/05/Untitled.png
Mike
5 years ago

On that basis you need to base your calculations on the size fraction that has the least material. Which means you will have plenty of material in the other fractions.
Trying to obtain the optimum amount in each fraction will skew your results as the size distribution will be different after crushing and grinding.

Y
Yvette314
5 years ago
Yvette314 5 years ago
Like

Thank you!

Please join and login to participate and leave a comment.