- To participate in the 911Metallurgist Forums, be sure to JOIN & LOGIN
- Use Add New Topic to ask a New Question/Discussion about Geology and Geometallurgy.
- 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.
Representative Sample Selection (15 replies)
The first thing that needs to be done in selecting a "representative sample" is to determine "representative of what?", and for what purpose will it be used? Without an understanding of that, you can't properly develop a sampling program.
For Metallurgical purposes you need to look at an ore body and determine how variable it might be. Often mineralogy changes in different portions of a deposit. There can also be changes in oxidation state, deleterious elements such as active carbon, or changes in amount of clays --which could impact on how a solutions in a heap leach pad might percolate. Changes in host lithology also need to be taken into account.
There needs to be some discussion with metallurgists in this process, and there may be a need for preliminary, and follow up sampling.
Often coarse rejects from drilling, or RC cuttings can be used, and they can be tailored, based on assay results and selected sample weights, to match an average for a particular area in a deposit. If the objective is to model flotation performance, then fresh samples are required, so that sulphides won't have oxidized on their surfaces.
Key is to understand the purpose of the sample and what it is supposed to represent, then to delimit an area to be represented, and to compare with some form of resource model or some statistics based on drilling or other sampling method, to determine what properties the representative sample should have.
Read Pierre Gy. No-one has described sampling theory better. A correct sample weight is determined by particle size and Heterogeneity. Once these have been determined, then a standard sample weight will ensure repeatability of assays for a given feed. Of course the actual sampling process is important as well to ensure no bias.
http://web.tecnico.ulisboa.pt/~mcasquilho/text/sampl_bulk/PierreGy.pdf
What will go into the mill is clearly a key factor in delimiting the volume that a sample must represent. You then need to look at the properties of the mineralization in the deposit and determine which ones will impact the processes in the plant. Different sampling sizes and protocols may be required depending on whether you are looking at a coarse free gold and GRG or other gravity recovery tests, or, perhaps in the same deposit, the impact of copper or zinc base metals on cyanide consumption in a leach circuit.
Pierre Gy's work is really important to understand proper sampling. When you begin selecting samples for metallurgical test-work you often have an idea from resource or exploration drilling what ranges of properties a deposit has, and working with a metallurgist you can determine how many samples are needed.
Samples with desired properties to represent specific portions of a deposit for specific purposes can be tailor made by selecting already sampled material such as drill cuttings, or development ore. Key is to ensure that there is a lot of thought going into selection of how many samples, from what areas of the deposit, and for what purposes.
A proper metallurgical test program is a cooperative effort between metallurgists and geologists. it requires a careful review of properties of the ore that will be of interest to the process people, and sampling to reflect that. Detail records need to be maintained of how each sample was selected and why. On any project I have worked on detailed record of where the metallurgical samples came from, how they were selected, and why are readily available.
If a geologist just cobbles together a sample for process testwork without working with a metallurgist, and without an understanding of what properties need to be tested, and how they may vary across the deposit, you won't have any guarantee that the samples tested are appropriate. The value of the samples tested depends on how much thought went into the design of the test program.
Thank you Alan. Yes "The value of the samples tested depends on how much thought went into the design of the test program." GIGO
I have seen so many "oh, results are bad = it must not have been a representative sample", while if results are good, the integrity of the sample is never put in question.
This forces a going back to the core shack for forms a piece-meal variability or now called geometallurgy program.
There have also been cases where new drilling programs started within what were called "ore reserves" because there was inappropriate or no metallurgical testing done and the drill core or RC cuttings were discarded. That gets really expensive, especially if investments have been made on assumed process recoveries.
That's a good and difficult question. I supposed it depends on how homogeneous the material is. So, in geology, that's impossible. On the other hand, it depends on what kind of study you do. If you think it's a representative sample, you only must demonstrate that properties is the same of all the rest of soil/rock.
Step by step geomet sampling needs to be done during the infill drilling stage as a resource is being confirmed and extended. Appropriate geochem sampling and mineralogy work needs to be done to flag potential issues during this time.
Are we talking in general or do you have a specific deposit in mind? Representative samples for met work should reflect what you expect to be milling. Keep in mind that It is quite possible that there is more than one 'representative' sample (ie, if your deposit has different alteration styles/intensities, grades, geochem characteristics etc.. ).
The practice of geology could use a good shake-up so thanks for starting a discussion to remind geos that the work we do, especially sampling has to be useful to other people and not just satisfy our compulsion to bring shiny rocks home.
We geos, most of all mineral collectors like myself, like samples that are most definitely not representative. Those samples have value too, both aesthetic and scientific for determining paragenesis etc. but they don't demonstrate economic viability of deposits.
I would, however suggest not calling anything a "truth machine." Geologists can log core very poorly, often because their bosses insist on nonsense. (That's topic that's been beat to death but not solved in several other discussions.) What good is all the lab work then? The model will be junk because you can't connect the dots when they aren't labeled correctly. Assays can be bad as well. I've seen tungsten values wrong by a factor of 100 that were ignored for a year despite being clearly at odds with the log that was correct that the veins in question were around 30% scheelite.
Best of luck to all of you who work at helping your professions be the best they can.
A wise Geo once told me "Selecting a representative sample is like having blind men trying to describe an elephant from the small portion that they can feel". What Does Selecting A Representative Sample Mean?
How can a few 100 kilos be "representative" of Millions or 100s of Millions of tonnes.
How do you Geos go about providing your Met with a solid sample?
Use the Social Share Bar on the Left. Tell everyone you can about https://www.911metallurgist.com/metallurgy/ It's FREE & GOOD.