Geology & GeoMetallurgy

Geology & GeoMetallurgy

  • To participate in the 911Metallurgist Forums, be sure to JOINLOGIN
  • 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.

Riffle VS Static Cone Splitter (4 replies)

(unknown)
8 years ago
(unknown) 8 years ago

Today I update the site's Hero image with two scatter plots of replicate samples from two drill rigs testing the same deposit. Rig A used a dump box and riffle splitter. Rig B used a static cone splitter. I have my interpretation of what is happening with replicate precision and bias here.

O
Obergruppenfuhrer
8 years ago
Obergruppenfuhrer 8 years ago

The fact that you're getting 1 g/t in one sample and 0 g/t in the duplicate for the cone splitter means either a dumb sample truck-up or waste material falling on a caked-up cone, with the caked-up material carrying grade and dropping off only on one side of the cone when triggered by the low-grade material falling on top of it, suggesting the cone isn't performing or being kept clean enough.

John Graindorge has a paper on cone vs. riffle comparisons as well I recall (from Sampling 2010). Catto& Church also did some work in 2003 and they found that the Cone was better.

I think it’s all about the setup. If the thing isn't level or clean then it stops pretty quickly, regardless of the Theory. TO me it shows that whoever was operating the cone wasn't doing such a good job.

(unknown)
8 years ago
(unknown) 8 years ago

How do we view the results, and do you have any photos of said equipment?

(unknown)
8 years ago
(unknown) 8 years ago

All apologies it seems this Hero Image thing is a bit random in appearance - here is a public link to the image I'm talking about - paste the following into a browser http://is.gd/UsYZhh
I have no photo's of the equipment only the graphs from the geologist who carried out this particular 'Pepsi' challenge.

(unknown)
8 years ago
(unknown) 8 years ago

It’s difficult to make a final judgement without more information. Could do with seeing Q-Q Plots to establish systemic bias and photos to establish set up - photos will probably explain most things. The riffle splitter's "apparent" repeatability can't be relied upon because the duplicate is taken from the last tier only (the final part of the split), not all the way through the individual tiered splits. In other words, any bias from tier to tier, of which there will be a lot, is not represented in the primary or duplicate sample. And then the duplicate is supplied from simply splitting the material from the final tier in half after bulk bias has occurred but the real joke is that and even with this flaw in its favour, it still can't provide a repeatable sample (tending to bias to the duplicate side on the graph.

As for the cone splitter, don't know without more information, other than to say, there are a lot of poor copies from the original Metal Craft Cone Splitter around now. And it was the Metal Craft Cone Splitter that was used in the paper mentioned above.

Perhaps a better way of explaining the above is that a cone splitter will take a primary and duplicate split from the bulk sample. A tiered riffle splitter, on the other hand, takes a primary sample ONLY and then splits it in half at the end for a duplicate. It is not therefore a "duplicate" sample at all and should not be used for any kind of QAQC. At best, it will just confuse you into a false sense of security. That is one of the many reasons why cone splitters exposed riffle splitters for what they are flawed, flawed, flawed. That said, cone splitters have now had their day as well, including the rotary/rotaport design that followed the original. We pioneered all of these products and continue to learn from them with new innovations, the latest one being the Progradex Sampling System. Now this really is a major step forward and, frankly, blows away anything that came before it. The results speak for themselves - see http://is.gd/Xux8qG for info.

Please join and login to participate and leave a comment.