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.

Proportioning Grade from an Ore Blend (3 replies)

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

Problem - Mine grade control is reconciled back to the mill at the end of month with a single factor, effectively on a pro-rata basis. One of the ore sources commonly returns assays up to 1000g/t. Current procedure is to cut these grades to 100g/t based on the favorable reconciliation back to the block model with like constraints. Grade control is based on grab sampling per 100 tonnes bogged. The mill sample is taken at the cyclone overflow (ie representative of the entire blend). If the grades of the problem source go uncut the ratio (factor) between grade control and mill ends up less than 1 and hence the other sources are downgraded. Our thinking now is that cutting is too final and we are searching other ways to proportion/weight/allocate grade in this situation.

Paul Morrow
8 years ago
Paul Morrow 8 years ago

I have just been involved in assisting a company to sort out this very issue in a mixed open cut and underground operation. The problem was found to be a lot more complex than just a single source of error. There were errors in the mill accounting, belt weightometers for different sources, truck weights, grade control sampling and the grade control models. They too were of the opinion that it was one set of numbers that was wrong. I do not recommend targeting one dataset without being 100% certain there are no other factors.

Let me know if you want some help and, good luck.

O
Obersturmbann
8 years ago
Obersturmbann 8 years ago

I suffered this problem a lot when working mine in the 1990s when the plant was being fed from multiple sources. The solution here was to convince the mine and process manager to run a few single source feed runs for a least a day so we could get a better idea of how the plant actually performed on different materials. It took little while to build up a test stockpiles from separate sources, but this gave clear indication of where the high grade was really coming from. You also have to accept that grab samples are going to be highly variable and likely biased. Have you done any testing to see whether rock chips from the larger rocks in the pile have significant different grades to the fines in the pile? Also, how does the grab sampling reconcile with face sampling in the longer run? This may give some insight as to where a top cut should be. Also with 1000 g/t results you must have some pretty coarse gold - what is the mass of the fire assay and perhaps you should be looking at screen-fires (or maybe a bottle roll leach) to dampen the outliers.

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

Cyclone overflows are also notoriously difficult to get a representative sample of. They contain the lowest fraction of solids (often 5-10%), so an enormous sample needs to be taken so that you can be confident you have not got a "nugget" effect. This can be compounded when you have multiple cyclones (and banks) that will operate differently because they are worn or with different pressures etc.

Validating the splits (with sampling and modelling over a range of operating conditions) around the cyclone and other sample points in the plant should help confirming the result. If the metallurgists mass balance gold bearing or other key indicator/trackable minerals would help relatively easily to get a handle on how certain the results are.

Often you can dig up historical sampling campaigns to give insight into the variability of different parts of the circuit. I would be surprised if there is not a good resource of sample campaigns (bread and butter for the gold metallurgists I know!).

Please join and login to participate and leave a comment.