• To participate in the 911Metallurgist Forums, be sure to be Logged-IN
  • Use Add New Topic to ask a New Question/Discussion about GeoMetallurgy or Geology.
  • 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.

Dump Truck and Stockpile Sampling (10 replies)

Tony Verdeschi
10 months ago
Tony Verdeschi 10 months ago

Today I updated to sampling site image to a thing still done on may mine sites - sampling a dumped load of ore anything from 50 to 100 t or more in a pile.The usual process is collect a number of fragments from the pile then assign the "maybe" material as ore or waste based on the result of these 'samples'. I have my preferred strategy for dealing with this type of 'could be ore' material - I am interested to hear what's yours?

Dizzy Flores
10 months ago
Dizzy Flores 10 months ago

I usually start with the premise - perfect sampling means anything has an equal chance of getting into the sample.

Some options (not in any order):

Option A: With large sample sizes of ROM material, crushing then splitting is the best option (if available).
Option B might be to "cone and quarter" with a loader / IT then crush and split further.
Option C: get the material on a belt and use a good sampling system
Option D: grab a few rocks and pretend you know the grade of the stockpile. [Even though you grabbed the most competent rocks that represent a specific geology and are unlikely to represent the geology / metallurgy / grade of the stockpile].
Option E: wait until the material is further down the processing route and mass balance.

Can't recommend option D (but see it often enough) and Option E often works out to be the most practical (particularly if you can validate / reconcile against one of the other good options A-C.

Unterstarm
10 months ago
Unterstarm 10 months ago

If the material is reasonably well broken, between 2-4 random grab samples by actually walking over the pile, and not just around its edges. Some ore like at the Hill 50 deeps you could get away with 1 grab sample due to homogeneity of the mineralisation. Nuggety qtz vein deposits needed 4 samples as a mine.

If it’s full of large oversize rocks, then fall back on "face sampling" of the largest rocks as well. The reality is, the material is a mixture of large, small and fines, and no sampling technique apart from crushing the whole pile is effective.
My solution is to send it to LG stockpile if you think its sub-marginal. The Mill will then give you a grade.

David Kano
10 months ago
David Kano 10 months ago

If these truck dumps are being created in the current mining operations, the question has to be asked why? Grade control practices should be of a sufficient standard to determine if the material is ore or waste. Also the mine geologist should be familiar enough with the mineralisation to make the call - that is what they are paid to do, make a decision!

If unsure, and as it suggests, send it to the low grade dump. This option however should not be used as an automatic default decision.

If they are historical stockpiles, depending upon total amount, how it is stockpiled (single dumps or combined dumps) etc, will determine the best course of action to try and determine grade.

Jean Rasczak
10 months ago
Jean Rasczak 10 months ago

This reminds me of the good old days of paddock or truck dump sampling commonly practised last century until a stockpile was re-allocated to the waste stockpile and re-sampled and re-allocated as ore. Spoil sport! Looks like old practices rediscovered!

Various methods were applied from throwing nets with different coloured squares and selecting a predetermined number of samples (smallest) to throwing cards, dancing around the stockpile to gather random specimens.

It is a belief system that you can reliably sample broken stockpiles; more related to witchcraft, the fairies and you will win the lotto and pay off the house mortgage.

As geologists we are all asked this question (instructed) and pay the consequences if we are wrong if sampling old stockpiles. If it is part of production I would say do your sampling correctly in the first place as it is cheaper.

I would prefer alternate shovelling or even fraction shovelling in preference to cone and quartering. Even if you finely crush the segregation in a stockpile can be considerable and sampling crushed material is not a trivial task. Processing plants are not good at reliably sampling small parcels of material and mass balancing suffers from compensating errors. How many processing plants comply with the AMIRA P754 code!
Back to the fairies!

Oberstorm
10 months ago
Oberstorm 10 months ago

Whilst I agree with both of you is it a case of knowing the commodity first. In a bulk commodity such as iron ore half a dozen grab samples per pile should give you an approximation of the piles precious metals worth require more and I agree with you on this approach. You have a good point on the homogeneity of the deposit even in a gold deposit like Hill 50.

Zander Barcalow
10 months ago
Zander Barcalow 10 months ago

For a pile of about the same size as you're talking about I had the pile mechanically flattened down to about 0.5m thickness as evenly as possible then sampled on 1m grid taking larger boulders and also dust down into the pile. Prepped all the samples (about 5kg each) then averaged the results.

Depending on the operation this amount of material is only a small fraction of the daily production and error/impact can be minimised by trickle-feeding into ROM. Although the extra handling and effort may not be worth the bother.

Agree with the above statements regarding style of mineralisation - clearly base metals at % levels going to be an easier job.

Also agree with comments about if it were done correctly we shouldn't have this problem...but who gets to follow the miners every second of the shift to track the tonnes properly? Minimising re-mucking is the idea but there is always a pile lying in an end somewhere that no-one seems to know about.

Hauptsturm
10 months ago
Hauptsturm 10 months ago

Sampling truck dumps depends on the ore body and economics. I've seen operations where it is a safety blanket and not really used for anything but checking everything's running ok in the stope or pit and operations where it second guesses all the good sampling and other work done before. The first is nice, but shows a lack of confidence in processes (or that the resource and grade control drilling/modelling/sampling or mining practices need some attention) while the latter is pretty scary and should be a significant alarm bell for management.

I've worked at some sites where a water truck and a trained eye is far cheaper and more accurate than grab sampling, while others (and in one case one stoping panel in an otherwise visual mine) where it is the worst thing you could do.

From an economic/operational view assigning destination once the material has been hauled to a temporary stockpile should be a last resort and done by exception, for example stopes with major dilution issues or rill mining in a pit, due to the added cost of double handling the ore or waste.

That said, as suggested earlier, walking over the pile is required if you are going to sample a truck dump, or alternatively, dozing it flat to -1m thick to ensure that the pile can be reasonably sampled. If possible keep it as dirty as you can (mud/dust etc.) and use less interested samplers (such as operators, technicians or surveyors (mainly because they are good at walking on rill slopes and uneven ground) as geologists tend to magpie out the shiny/interesting bits or oversample the waste to "balance" their grab sample, resulting in a poor sample.

Your suggestion of face sampling oversize can also lead to significant cost reductions. One gold operation I know of in the Yilgarn identified that most of the oversize was sub grade, resulting in it being removed from the ore feed and significant cost savings in rock breaking, trucking and processing for the operation.

Tony Verdeschi
10 months ago
Tony Verdeschi 10 months ago

All thanks for your comment. My preferred option is build a "maybe" stockpile from this material that geologists consider to be potential ore. This can be processed at the end of the mine life if somebody wishes to take that gamble.

Sturmbann
10 months ago
Sturmbann 10 months ago

We had an interesting case in the DRC with a 10,000t stockpile of ore (different to the usual mineralisation) that was due for processing a week later. The stockpile was already spread out and was sampled in a pattern across the surface with a TLB down to about 50cm. The material was put into a dump truck (about 20 tons) and that was then spread (about 20cm thick) and sampled in a pattern with a spade. About 200kg was then sent to the lab for further processing (crushing splitting etc) and analysis.
The information was required for setting rough operating parameters for when it was fed to the plant.

Carl Jenkins
10 months ago
Carl Jenkins 10 months ago

Some time ago I had to design a sampling exercise for a large old (more than 30 years since the mine closed) waste rock dump of material that contained disseminated monazite and sulphides in some of the fragments. The material was originally dumped on the side of a small hill by truck resulting in a kidney shaped wedge that was built over a period of at least 20 years. I have observed that the material on the dump range in size from clay size to very large boulders. The more sulphide rich the more weathered the clasts appeared and one may assume that at least some of these fragments disintegrated and contributed to the fine fraction. The larger fragments were relatively fresh and un-mineralized.

I was asked to provide a sampling programme that would enable me to issue a Mineral Resource Statement for the dump. I recommended an elaborate sampling programme, during which four large trenches had to be bulldozed radiating perpendicular to the curvature of the dump, with large samples that had to be collected from pits dug ant successive elevations going down at 1m sections in the dump and where all the material in the pit had to be collected in 200 litre drums and sealed. The intention was to truck this to a suitable facility where the material could be crushed then split and a representative aliquot be milled before analysis.

This would have been an expensive exercise. So the project geologist decided to first attempt sampling using a number of different drilling techniques, which failed one after the other. The costs were climbing and with the budget almost blown he then decided to sample the first four rows of pits dug following my design, and not to go deeper than the first meter. I inspected these pits and especially the discard left next to them which confirmed all my previous observations. The same geologist also decided that it was not necessary to collect large samples, but collected about 5 kg from each pit. On opening and emptying the sample bags onto plastic sheets, I found that only the smaller size fractions, up to about fist size, were sampled. When I then indicated that I will not be signing off on a Mineral Resource Statement for the dump, my services were no longer required and that of another consultant was acquired.

The moral of the story is that such dumps are some of the most difficult entities to sample and gain statistical confidence in the grades of minor to trace constituents. The volume of the dumps can normally be quantified with reasonable confidence, but the relative density may be as uncertain as the grades. I wonder if one should at all consider Mineral Resource Statements for such dumps, meeting best practice standards used for in situ resources and tailings dams. Collecting samples from fixed volumes on a regular grid on the surface, then crushing (with splitting only after secondary crushing), milling and analysis, may provide adequate information for a decision on re-processing them, if the grade obtained is sufficiently robust compared to the cost of re-mining and processing.