Each type of concentrating equipment such as jigs and centrifuges, have a range in which they operate best. As such classifying the feed will give improved performance.

1/32" or 0.08 mm is below the normal range for jigs, it is more suited to tables or spirals. And probably should be handled as a separate stream. Also do not expect very good recovery in this size.

Really I was thinking mm and used 2.5 which is cm.

1/32" is 0.8 mm not 0.08, Still would work better in on a table or spiral over a jig, and is in range for centrifuges.

As I have written before, in gravity separation (which is really separation based on particle mass), the more narrowly sized the feed to the gravity separating device, whether it be a jig, spirals, centrifugal concentrator, tables or whatever, the better. In an ideal world, the largest particle of gangue should have less mass than the smallest particle of gold. Given the vastly differing SG's between gold and most gangue minerals this gives some latitude in how narrowly the feed product is sized. However, as we all know, the ideal is rare in mining and processing.

Generally speaking, feed material which is not very narrowly sized will result in a higher mass pull, i.e., a dirtier concentrate. This is typically handled by staging gravity concentration in a series of steps to pull as much of the gold as possible into a high mass pull con in the primary stage followed by secondary and tertiary concentrations with further classification between concentrating steps to achieve high concentration ratios and a clean product. This is made easier, as the process stream to be dealt with by each succeeding classification and concentration step incrementally reduced.

If I understand your situation correctly, the maximum size of your gold is .25mm or 250 microns, and your gangue is typically about 1/32" or 800 microns. If this understanding is correct, this is ideal feed to a 2mm wet screen ahead of a semi-batch centrifugal concentrator. Of course you will have to screen ahead of the 2mm wet screen, as it will not like to receive oversize larger than about 1/4".

If properly sized and operated, the SB centrifugal unit(s) should pull a high percentage of GRG down to about 50 microns, or perhaps a little smaller. Cons from your centrifugal will need to be further sized into narrower ranges before secondary and tertiary processing.

Of course, there are a number of factors which will need to be considered regarding the practicality and the efficacy of the foregoing, including Corey shape factor of gold particles, character and composition of gangue, etc.

A jig does not precisely work on mass. That is why the centrifugal jig concentrators Kelsey works better than any centrifugal concentrator.

With diamonds, rubies, gold, COLTAN, tin etc. There is absolutely NO need to classify into different size fractions before jigging. I understand the theory, but in practice hindered settling for heavy minerals is the most important factor. Try jigging +8mm (let alone +12 or +16mm) near density material, the jigging action no longer lifts the material into a buoyant state to allow differential settling. Practice has shown me efficiency of jigging on unclassified (un-sized) material are excellent at small particle sizes. The increased effective viscosity of fines will lift larger size particles and allow differential settling. Rotary pans behave like jigs. With other density separation methods, particle size classification is essential.

In that if processing 10 mm x 0 material, the hindered settling rate of a 0.25 mm gold particle is less than the settling rate of a 5 mm rock. As such a jig would wash the gold right over the weir. As such a classification at around 1 to 3 mm would give better results.

Long time since I did these calculations, perhaps you can send it to me. Assuming you are right, gold seems to be an exception, but it's certainly not in practice. I see you have never observed or measured the efficiency of a jig working on minus 16mm. Jig theory is not only reliant on hindered settling. The larger particles form ragging, accumulating in the bed and are not moved so far down in the bed as finer ones, the rock particles wash over easily in waste. The main force to separate small particles is their interstitial trickling between larger particles, hence similar size particles separate - perhaps minus 2mm. Chinese & Thai jigs never size classify jig feed and recover to 95% plus @ 125 microns from tests I have done myself. The Chinese are the biggest producer of gravity gold in the world, their jigs are by far the best sellers in the world.

Actually I have operated, observed, and measured the efficiency of a jig. And when the particles are very fine, even a jig with good ragging, properly set up with good hutch water, and not over drawing on the hutch product has difficulty in processing minus 0.25 mm gold.

Does this imply you should perhaps classify a little less, feed primary jigs with bigger particle size and range, then use the jigs to classify to (what is the optimal classification before losing fine gold) and then take this concentrate, perhaps sacrificing optimal use of jigs, to avoid losing gold, and report that concentrate to a different secondary and tertiary concentration like centrifuges?

Have you experienced the efficiency of a circuit of jigs followed by centrifuges?

If one were to use jigs, what type of jig would it be? I thought the Chinese jigs were a variant of the Pan-American jigs, but not sure at all.

If not jigs, then what for primary concentration to balance volume of throughput and reduction of losses on the primary stage?

Gents, with a nod in partial agreement with all of the foregoing comments, and acknowledgement that hindered settling is indeed a factor in jigging, be it conventional or centrifugal, I still maintain the principal mechanism of gravity separation is a function of particle mass. This is, of course, mitigated to some degree by hindered settling, particle shape, i.e. Corey factor, and rheology of the pulp.

Based on both personal observation from having operated all manner of jigs for many years, and from a basic understanding of jigging principles that a conventional jig (1g), even when operated optimally will not be highly effective at sending -250 micron gold to the hutch while at the same time sending particles of gangue as large as suggested to overflow. This is particularly true if the gangue is relatively high SG.

With regard to the suggestion that centrifugal jigs work differently in this regard, I disagree. Centrifugal jigs, while incorporating centrifugal force to "artificially" increase the mass of the particles being subjected to the jigging process, the process increases the effective mass on all particles uniformly, large and small high SG and low SG, thus making each particle "heavier" and more likely to sink against the jig pulse and report to the hutch. The net effect of this is to minimize rheological effects on very fine particles, thereby allowing the fine heavies to sink, but it does not affect the relationship between fine heavies and coarse low SG particles with respect to their propensity to report to cons or be rejected to overflow.

Though I do not have any recent experience with centrifugal jigs, I did work with one of the original developers of the concept, back in the late 70's or early 80's. We tested the Campbell jig for several months at the Sunnyside Mine in Silverton, Colorado. The observation then was the same as the physics support, and as described above.

The bottom line is as has been well known for many years, excepting minor variations due to hindered settling as in jigging, gravity separation is indeed separation based on particle mass. Harking back to the original question in this thread. "Is there a point below which further classification would not add to the efficiency of recovery and might actually lead to losses at the classification stage?” in theory, the more narrowly sized the particles in the feed stream, the more effective the gravity concentration. In practice, some gravity concentration methods are more forgiving such as jigging, but I do not believe to the extent suggested by Mr. Canby, as supported by the observations of Mr. Albrecht and my own experience. In direct answer to the original query, in gravity concentration, as in all such things, there is a diminishing return for more and narrower feed stream sizing, each gravity concentration mechanism has its own "optimal" range, depending on a number of factors enumerated above. With regard to potential losses in the classification stage from "over classification", this is certainly a possibility at the extreme, but this is the subject of a whole different discussion. The best advice is to look at the classification specifications from the maker of whatever gravity concentrating device you intend to use, and err on the side of slightly tighter classification than the maximum range recommended to get the best result. Even better is to pilot test your feed material through the intended device to examine sensitivities to classification range in the context of your actual feed material.

The purpose here is not to fix a specific problem as such, but to get a general understanding of what may already be beaten paths and proven methods.

Once the level of detail gets too granular (excuse the pun) I agree the matter falls beyond the scope of fora such as this and more in need of examining specific data and actual operations.

Before that though, the intention is to get some general principles understood, so at the very least be able to understand how and what a specialist would be recommending and the logic behind such a recommendation.

As a non-geologist/metallurgist, it is rather easy to be convinced by one or other opinion/conclusion put forth by a relevant expert. The danger is that, if not properly challenged and substantiated, such a decision may well result in a project taking directions that may cost significantly more in terms of not addressing issues, or worse even, be convinced that they have done so, and apparently poor results may adversely decide the fate of a project that could otherwise succeed.

I was working on a project many years ago, that even though the ore was 0.3 oz./ton, and free gold, it was not gravity concentrate. The particle shape was the culprit, they were all flattened. Even using centrifugal jigs we could not get better than a 3:1 concentration, using two stages of jigging.

I wanted to try flotation, but ran out of money.

We also tested the Kelsey jig on gold tailings and the results were 3:1 about, but this is good enough to reduce the mass and process the pre concentrate by CN. This is the objective.

Jigs act as size classification devices in themselves as well as density separation devices and are complicated in terms of theory. Yes lithic particles being a pest in gravity concentration. In diamonds, if basalt which breaks in laths is present, they all report to sinks in DMS circuits!

I am assuming this is gold from alluvial?

IF this is correct, I see little benefit in sizing ahead of jigs. The jigs themselves will self-classify.

If the ore is crushed ahead of jigs, then pre-screening of fines will likely remove liberated value minerals. The jigs can then be optimised for the remaining coarser material.

In practice I have found repeatedly good results on different alluvial/eluvia materials on 0 - 25mm for diamonds, rubies and COLTAN. To 30mm if there is enough fines and clay to give the bed buoyancy. Throw ALL in a scrubber to liberate, discharge trommel at 25mm (or maybe more).

Then all in a jig with 2mm mesh to separate gems. (4mm for Coltan/tin etc., then jig through to more jigs at 2mm, then to spirals and tables) as would be expected.

Compartmented jigs screen bed if feeding all size feed is essential. Unless other heavy minerals present, nothing is cheaper and easier. All equipment must be Chinese! Buy a nice car or three with the difference. I kid you not. Issue is security in removing concentrate from jigs with scoops, and there is an ingenious, simple modification to accomplish this to eliminate security problems (at least at jig stage)

Agree with above comments. Classification is a method of concentration and it can be overkill at some point. Bench scale or pilot scale testing is required. The full range particle size analysis is required now - the percentage of gold and gangue in each category (extremely fine, very fine, moderately fine, fine, medium, coarse, etc.) and shape. Manufacturers and some consultants will try to sell you anything. The mantra I was taught was KISS - Keep It Simple Stupid. The operations that go broke are run by tinkerers which spend 85% of the time recovering 15% of the gold. The Little Squaw alluvial fan in Chandalar, Alaska is a case in point. The gold was coarse (70% above 40 mesh) but company directors demanded a fine gold recover plant which cost WAY too much, was too complex to build and maintain, and never recouped costs. A simple sluice box would have been much cheaper and would have gotten them into positive cash flow within weeks.