The measurement of residual xanthate has been used in bench scale test programs, pilot plant programs, and plant trials.

However, I do not think that its use has been extended beyond monitoring for investigative purposes.

A couple of reasons for this:

•How to reliably perform the measurement on-line to input into the process control system is not a trivial task

•A scheme performed the same outcome is a feed-forward control loop of the xanthate addition based on head assays (often converted to metal units per hour via multiplication by throughput) with a feed-back on the gain of that loop via the scavenger tailings assays. The availability of reliable on-stream analysers over the last few decades has been key in the success of this approach.

I don’t consider myself a flotation expert, but I’m brave enough to put a few comments here:

If the circuit is under dosed you can tell by the low recovery. So if you’re observing low recovery simply crank it up and monitor the recovery.

My understanding is that there’s always an overdose in the plants due to reasons below:

•Not all the xanthate added will interact with the mineral particles; I call it efficiency but I’m not sure if that’s the right term for it. So there’s a need to overdose compared to what the theory dictates.

•Operation people are conservative and they always add a big safety factor as we engineers say (I call it risk mitigation)

•Due to the change in the mill feed over the course of the life of mine, the need for xanthate might decrease and sometimes nobody bothers with adjusting the reagent addition according to the new feed (avoiding the risk of low recovery)

•The mindset in operations is to crank up everything to achieve high recovery. However, in this case there’s a big window to play around with xanthate addition before any signs of negative effects on the recovery starts to be observed.

The xanthate can be easily measured using UV-Spectroscopy: Jones, M.H & Woodcock, J.T: "Ultraviolet spectrometry of flotation reagents with special reference to the determination of xanthate in flotation liquors". London (Inst Min Met) 1973. 24 p.

And, there are some methods developed to online measurements of xanthate.

If you’re concerned about the environmental impact of the excess xanthate in the tailing ponds; I don’t think there’s too much to worry since xanthates degrade fairly quickly once in contact with water, especially in low pH. Most probably xanthate decomposes to CS2 and alcohols such as amyl, ethyl, isopropyl.

If you’re suspecting that there’s an overdose, simply crank it down in increments and monitor the recovery.

We have been operating with traditional methods as mentioned above and trying to stay conservative to optimize recovery. If we only target the Cu in the head grade and calculate the amount of PAX required collecting that amount of Cu, we may end up starving the circuit as there could be some "xanthate rubbing" gangue present.

However, by the time we figure out the recovery is suffering (depending on OSA availability and accuracy or Assay lab turn around on grab samples) it could be too late by the time we discover the recovery is suffering.

We are in a type of supergene zone ore that seems to be consuming extensive amounts of PAX when it shows up. To be proactive and adjust PAX addition based on presence of "xanthate rubbing" minerals or gangues, we need to monitor the free xanthate in the flotation liquor. We also want to test the ore samples prior to milling in order to figure out the level of collector consumption.

We have had a lab test done using UV-Vis (at 301 nm) and that seemed to provide reasonable results at a reasonable turn around.

We also are in the process of developing a lab titration for determining free xanthate.

Our goal is to optimize recovery(of course) and minimize PAX consumption ($$), but still maintain free xanthate at a certain level to ensure flotation is not starving for collector.

We are looking into buying a UV-Vis unit, DR6000 by HACH. Any opinions on this unit?

I'm just wondering if any operations are also trying this; if yes why and if no, why not?

You are providing good details about your situation and this helps providing more precise tips.

The supergene zone ore is likely consuming more xanthate than the regular ore due to release of copper ions in solution from the secondary (and possible oxide) copper minerals during grinding. These ions would then react with the xanthate effectively preventing it from doing its task. Also, if pyrite is present, the presence of copper ions would activate it and make it highly floatable with xanthate.

One way to better frame the issue would be to perform an analytical copper speciation (i.e. how much of the copper is present as chalcopyrite, secondary sulphides, oxides, and silicates) on a suite of samples covering a range of ore types fed to the concentrator. Between 10 and 25 samples should be sufficient to indicate if this would provide any hint(s) about the underlying issue(s). For example, xanthate alone is not sufficient to efficiently recover copper oxides and copper silicates; if those are present in the troublesome supergene zone ore, then this would explain the observed recovery drop with this feed. The observed symptoms in a plant would be higher than usual tailings assays even with extremely high xanthate dosages - i.e. similar to those you are mentioning.

As for the UV-Vis unit, any decent instrument would do well analytically. The key criteria are ease of use against the skills of whoever would be using it - say on the night shift or on weekends.

I do not know of any operations using residual xanthate determinations on a regular basis to diagnose or fine-tune their metallurgy.

I've got a simple suggestion. All you need is a lab batch cell and a means of getting a tail sample. I've done this at a number of operations and it’s a great troubleshooting tool.

Basically you want to get setup to do a float test where ever your batch machine is. Have some xanthate made up and ready. Then you go and get a sample of the rougher tail (or other stream you want to examine). Take this sample immediately to the lab and float it for a few minutes. If you have copper minerals floating then you likely have enough xanthate, keep floating until you see no mineralization in conc. At this point add some xanthate. If you get more copper minerals floating then you are under dosed. Get ops to adjust dosage (I really recommend adding down the bank if possible) wait for a few circuit residence times and repeat the test.

The beauty of these ‘hot’ floats is they are quick and give a pretty good indication. You can make them more complex if you want and start timing/analyzing concentrates but I think you'll find that its quiet a good troubleshooting tool without collecting this additional data.

What a great, practical, common-sense technique! Xanthate additions are well in excess of the required quantity, so there is always an excess in solution. The xanthate addition rates would have been selected based on test work - so what did your test work tell you based on this ore type?

People rarely measure residual xanthate, let alone for production control as points out. Where it has been done, it was for 'educational/training' purposes - e.g. xanthate levels in PGM flotation (which was found to be four times that required).

The fundamental issue here, in my opinion, is the cause of the xanthate consumption - which would be mainly attributable to precipitation by base metal ions as suggested above, depending upon the slurry pH. Supergene copper ores typically contain covellite, chalcocite, etc. and perhaps some copper 'oxides' as well iron 'oxides' etc. all of which can release lots of base metal ions into solution. Removing these base metal ions (higher pH, adding sulphidiser, etc.) prior to xanthate addition is required to substantially lower the xanthate addition requirement and making it more predictable.

The advantage of a sulphidiser is that you convert the base metals into fine sulphides which can be recovered rather than lost to the tailings (environmental impact).

Of course, if you are operating at low pH, xanthate half-life can be quite short (<6 hours), however this should have little impact in a rougher-scavenger flotation bank (<30 minutes).

Lots of good information and practical knowledge out there! As always I really appreciate all your input and I acknowledge the difficulty in conveying information via this method as it takes a bit of effort to get to the point, but never the less, priceless input and tool.

Without going too much into specific details and getting myself caught in some breach of company policy, I'll try to stay general and continue this discussion.

One of the things that is very surprising to me is that why doesn't any flotation circuit operator look into managing collector addition based on the needs of that circuit (i.e. Roughers, Scavengers, Cleaners). Why do we accept the fact that it's okay to overdose the circuit and not worry about it Why do we just rely on some lab tests that was done many years ago considering the ore type has changed since then and now we may be dealing with something slightly or quite different which requires less or more xanthate? I think this also applies if we run a bench test in more recent dates. Things can change rapidly.

If there is a tool to measure residual xanthate in the circuit, online, and in real time, why don't people jump on it and take advantage of it? The savings in terms of reduction in xanthate consumption and increased (or on-target) recovery would be quite significant.

The advantage in achievinghealthy recovery by monitoring residual xanthate in a copper flotation circuit becomes even more evident once you get into the supergene zones where there is basically a ore complexity depicting "dog's breakfast"; and like you guys mentioned, some of these secondary mineralization and/or gangue will readily consume xanthates and starve your good sulfides from having a chance.

If we wait until everything hits the tailings, it will be too late to realize there is still too much of good sulfides left to float. But if we check this at each stage of flotation, we can pin point the lack of collection power on the spot and correct it by increasing xanthate addition. I like what above suggests and I know many operations use that method as it's quick and relatively inexpensive. In terms of xanthate half-life I'm not too worried as mentioned earlier our residence time in flot circuit is quite short.

Like recommended, I agree that a mineralogy work needs to be done to identify what's constituting the majority of non-sulfide minerals and what are they specifically.

Sulfidization is a good choice for sure when it comes to floating oxide ores. However, I suggest sulfudizer addition after xanthate addition! Here is what I mean and how: You aim at floating your good sulfides at the front of the circuit and get them out as fast as possible, and then you hit the circuit with a sulfidizer and some morexanthate. The reason being is that if you add the suflidizer at the head end of your circuit it tends to depress sulfide ores and this will in turn have an impact on your recovery.

One reason why Free Xanthate may not get practised is the advent of OSA's. (On line Assayer's)

On line assays are able to indicate process disturbances that can be dealt with both manually and in many cases, through automated control, to optimise grade and recovery, thereby maximising the economic value of the ore body.

In this case you can use two collectors, one for float sulphides that are added to the head of Roughing (e.g. the betacol) and very little NaHS and PAX. Ahead of exhaustion is adding enough PAX and NaHS. At the scavengen added no reactifs. If you find that the content of Cu decreases in concentrate you can stop putting NaHS and PAX ahead of Roughing.

That is what people generally do with mixed copper 'oxide'-sulphide ores - float the sulphides first then chase the 'oxides'. If there are no acid consuming minerals present (e.g. limestone, dolomite), then acid leaching is employed rather than sulphidisation/flotation. Where there are acid consumers [or you don't have an SX/EW plant or don't want to recover the copper by cementation], then sulphidisation is used. Ideally, Controlled Potential Sulphidisation (CPS) rather than just adding sodium hydrosulphide at a fixed addition rate.

Anyway, you haven't offered much about what is going on in your operation - mineralogy, operational conditions (e.g. circuit pH, etc.), considered reasons for the increased xanthate requirements, any work that has been doneon the topic (test work, mineralogical analysis, EDTA analysis of slurries).

Perhaps you need to do some 'bucket' chemistry with a crisis on your hands but if lasts any longer (> week) then some structured test work and analysis is required.

None of this should have been a surprise if there was a good resource model that identified ore types, etc. on which the mine schedule was based (cf. geo-metallurgy). Perhaps this part of the orebody was not well logged (or insufficient assays taken) and thus not identified as an ore type. All ore types and blends should have been identified and tested in the various studies (Scoping, Pre-Feasibility and Feasibility) leading up to the commencement of the operation!

I would be having a discussion with the mine geologist to see how much of this transitional ore remains, and whether more ores like this or different to the normal ores exist and when they are likely to be presented to the plant.

In addition, I would be running a weekly programme on site to test future ore types, based on discussions with the geologists and mining engineers.

The golden rule in processing (or anything really) is no surprises and minimising surprises and preparing for changes is an important aspect of production.

Some answers to your first two questions:

«Why do we accept the fact that it's okay to overdose the circuit and not worry about it? »

Flotation operators, control operators, and plant metallurgists are the first to loudly hear from the mill/site manager when the recoveries are down. And most of the time, increasing collector dosages will increase recoveries - albeit at the expense of concentrate quality. Thus, to prevent to hear the question first thing in the morning, all involved will pre-empt it by increasing collector dosages. In the overall scheme of things, reagents are inexpensive compared to labour and power costs to operate a mill.

«Why do we just rely on some lab tests that was done many years ago considering the ore type has changed since then and now we may be dealing with something slightly or quite different which requires less or more xanthate?»

Changes and ore variability are concepts which are very difficult to fully appreciate from an operating point-of-view. There are so many instrumentation, material handling, mechanical, or process control issues to fire-fight in a day that variations in day-to-day quality of the feed cannot be directly observed beyond the benefit of the doubt.

The human mind tends to manage complexity of the information for decision making by assumptions, generalizations and simplifications.

With respect to assumptions, thinking within ones box only is very natural unless having been learned that there is value to do otherwise.

A simple geological feature like a fault dissecting an ore zone is considered from a rock mechanics view point for the safety of the underground workers and for maximization of the ore extraction. It is however rarely communicated to the metallurgist as an important feature of this ore zone which potentially could negatively impact the metallurgy - unless mine geologist and engineer are aware of the existence of such casualrelationship.

It comes down to communication and team work under an engaging leadership!

Performing laboratory flotation tests is too often perceived as a very costly exercise without direct influence on today`s issues in the plant.. And when it comes to cutting operating costs in a down turn, the people performing such laboratory flotation tests have historically been the ones first cut from the mill staff.

Perception of the values of a roles and tasks in a mine site organization easily gets disconnected from their real values. Notan easy challenge to overcome!

Finally, simple solutions which a floor operator can understand and apply during his/her shift have a greater probability of success than more complex ones requiringconstant refinement by the metallurgist.

Let me suggest monitoring liquid phase of rougher, scavenger and cleaners.

One thing we haven't done and I've been thinking about is the dual collector system.

The element of "surprise" is unfortunately part of the game as I've seen so far. I'm sure there are operations that don't get this, but Mother Nature can throw you curve balls when you least expect. At least this is what Geologists will claim (not taking a cheap shot at Geologists here :)).

It's not a chaos that I'm speaking of here, nor I'm expecting; but nevertheless throughout the troubleshooting, one can't help it but wonder why isn't there a big move in the industry towards measurement of free xanthate. If OSA's had 99% availability and were accurate more than 90% of the time, I would heavily rely on them by interlocking my reagent addition to the OSA. However, even if OSA is working to satisfaction, it only tells you the copper head grade and some other elements of interest, it will never tell you (correct me if I'm wrong) that the feed has changed and now we have some unwanted limonitic gangue present that will ultimately use up any xanthate available in the circuit. OSA will also give me the recovery data, which I still think is a bit too late! On a side note to keep OSA systems running optimal, you will need constant attention, but it seems like we are all okay with that and have accepted it as "norm".

I agree with you, that Communication between departments can ease some of these issues, but it's easily overlooked. If theMill knows what Geologists have seen in the latest blast, they can be proactive and prepare for the challenge.

Relying on "floor operator" can be a life saver only assuming if the operators are well seasoned. Nowadays, it's hard to find float operators who have more than couple of years of experience and it turns into a training program rather than optimization process.

My goal is to develop a test program where I can take blast samples and run a xanthate consumption test in the lab prior to feeding the ore to the Mill; and to also repeat this test during processing of this ore in order to stay at optimal xanthate addition.

That’s exactly what I have in mind to test the flotation liquid (or filtrate).

I'm still curious to know if some operations do measure free xanthate in their float circuit. If they are, then they must be quiet about it, if not, I'm wondering why not? I'm seeing much value in measuring free xanthate; unless I'm missing something here.

The OSA can tell you head grade, tail grade, grades for whatever float cell you pull a sample from. If you want a cell sampled you only need to direct the sample line to the OSA.

I guess the biggest indicator for the xanthate use would be finding the Tails grade higher that it should be by the OSA. This is an instant reading which would enable you to make Xanthate adjustments immediately. No waiting for lab results which can be up to 4 hours old.

It might sound like I am trying to sell them 🙂 Not the case but they do have a place in the float system and do save a lot of wasted reagent and do help run a float plant more efficiently. As you say they do require daily calibration, but that is a small price to pay for a good system.

As far as knowing your head grade, if you have any type of good management assisting you, you should be getting a daily heads up as to what the grades are going to the plant. The OSA won’t tell you that the head grade has changed but if you are sampling other sections of the float plant you would pick up on this, coupled with the headsup from the Geos etc that the grade has changed.

Your testing of the Xanthate usage on the blast samples sounds like it would take some time to qualify and though it would give you some kind of base line to work from seems time consuming and too late to address a feed rate that would be constantly changing.

As you say, the operator is the best controller of how the float cells are working by his visual inspection, and as you say they need to be experienced.

There is NO substitute for an experienced flotation operator.

You asked the question why people accept overdosing of any reagent, let alone collector.

Well there are several answers, not least of which is that test work and addition strategies are not that exhaustive nor developed. Typically just a fixed amount based on the sample head grade used in the test work! Ben Murphy described to me once how he used to make collector additions based on a per unit metal content which is far more sophisticated.

Plus one wants to make sure that there is enough collector present to ensure that flotation recovery is not affected given variations in feed grade and chemical conditions - so overdosing in the absence of data supportingan addition strategy is a given. Dee Bradshaw discovered that the collector levels in PGM flotation were four times higher than needed this had come about over many years with a little more being added routinely but who is going to risk their career over a loss in PGM recovery? After all, the cost of excess xanthate is miniscule compared to the value of lost PGM values.

The use of starvation levels of collector to effect separations is an interesting area which is performed without residual collector analysis (e.g. some forms of Pb/Cu).

In addition, one area where the topic of reagent residual levels (i.e. correct amount added) should be of financial (and to some extent performance) interest is the copper activation of sphalerite, since relatively large amounts are used.

And the pH at which the copper activation is conducted...much lower quantities at low pHs if the order of reagent additions for sphalerite activation is not critical.

Residual xanthate should be monitored at the same time of feed copper grade as well as the main metal to reject Fe?, in rougher step the overdoses of collector make safe result of recovery over 90% until starvation levels, but that means excess of reagents group used. Hear should define the selectivity of the concentrate values included besides copper among silver, bismuth by example. A solid liquid separation step could be added to recycle residual xanthate before regrinding and good washing in columns cleaning, open circuit of cleaner tailings. Activated carbon powder can assist to reduce the excess of reagents to get the best concentrate grade close to theoretical. Xanthogenate of copper and iron dixanthogeno of rejected metal don’t tell you much about proportions every time of the xanthate consumption. Automation control is recommended from my point of view.

Blue Cube Systems are busy developing a system to measure Xanthate. It is based on the MintekCynoprobe extraction system that will be connected to our Hydromet Analyser.

I would like to clear something up here, I do appreciate the usefulness of OSA's or ISA's and the data they generate is absolutely necessary to run efficient operations. I haven't been to a facility that does not utilize an OSA (of any brand). What I'm implying is that OSA's, just like any other tool, need constant attention to operate well and when they are not being looked after for any reason, then operations will be running the circuit with 4hrs-old data from the Assay Lab and this is not acceptable in my mind.

Looking at free xanthate is just another tool in operations' or Metallurgist's hands that can be very useful and it could give us that added benefit of knowing something else that's going on in the flot cells that we do not know. I'm referring to the free xanthate levels. Even if the OSA's run well and we get constant and reliable data, sometimes we are faced with a puzzle and before we figure it out, the conditions in the circuit change and we are off to a different battle.

Checking for free xanthate, without the use of new technologies that are out there as mentioned above, can be done quite easily and quickly. All one needs is a spectrophotometer unit (UV-VIS) set up in their lab, preferably inside the mill for easy/quick access. Then you just need to grab a sample from each cell that you suspect is starving for xanthate. Filter these pulp samples, and run the filtrate in the UV machine. The results are ready in less than a minute and the whole process will take about 15 minutes. Of courseone can install automated systems as well if budget allows.

This method can also apply to the blast-hole samples with exception of pulverizing and creating an slurry sample and then adding xanthates and allowing some time and then collecting pulp sample etc. Based on this, which is a proactive move, we can prepare xanthates at higher concentration in order to provide enough xanthate to the circuit. Another way to deal with this would be having more distribution pumps installed for xanthate. Typically there is more than one day of notice for a new ore being hauled to the mill, so this provides plenty of time to do a "free xanthate" test in the lab and figure out what type of beast heading to the mill.

We have noticed that assigning xanthate dosage based on the Cu content of the feed grade, can lead into xanthate starvation IF the feed contains xanthate-rubbing gangue minerals. So one needs to be careful when xanthate dosage recommendations are given based on test work at early stages of the mine life. The xanthate dosage should be adjusted based on Cu feed grade and other xanthate consuming minerals.

My two cents (and that is very cheap South African cents 🙂 - from a process control perspective. (By the way - I was also part of the development of the xanthoprobe some years ago).

The grade and recovery in flotation is of course affected by multiple variables, including the collector addition, the pull of the cells, the material itself, density etc.

So any test you do on adding collector and observing effect on recovery would have to be under conditions where you keep the other variables constant.

It is indeed a difficult thing to manage all these variables at the same time, but online expert systems that control the mass pull (using flotation cameras) and reagents (based on both feed grade, froth characteristics and tails and concentrated grades) in a multi-variable fashion is the way to go I believe. It aids the operator in consistently making the correct decisions. These can be coupled with online measurement of the residual xanthate if need be, but like others have said, usually best to overdose to some extent, but not so much as to poison the froth. 

Thank you for your input; in my world even cheap cents are valuable. Here in Canada we don't even have "cents" anymore! Our cents called "penny" is sadly out of circulation. I agree with the first paragraph in your comments. Any test that requires evaluation of effect of one variable requires steady state or solidification of all the other variables so the final effect is easily determined and conclusions can be made with confidence.

I'm surprised that you are not trying to tell me that Xanthoprobe is the way to go 🙂

My sense is that when dealing with supergene ore, there might be pockets of this ore that you do not want to send to the mill due to the fact that it will be high consumer of the collector. The grade of the target mineral might be there, but it would be costly to recover it due to the presence of high collector consumers; I call the collector thieves. Hence, we can do a free xanthate test in advance and make a decision prior to sending this ore to the mill.

We have found out, based on some lab test and some plant trials that over dosing are not always feasible without knowinghow high we need to go to get to the overdosing levels! The overdosing level is like a moving target and never the same dose. This is a hard concept to dig for those who have not faced a supergene ore. It was for me too. For example if I wanted to set a dosage that will be in overdose range for collector addition and cover the range of ore we receive, it would be in the 200 g/t range! Running at this dosage would kill good sulfide recovery and the cost would be enormous.

It's not possible to do a mineralogy test for each blast zone in a timely manner so that the mill can benefit from that data. There are tools available, but with 24 hour turnaround. However, a "free xanthate" test can be done in the lab as I mentioned in earlier posts above. Let's say a mineralogy test can be done in 2 hours (just hypothetically), then we could look at the gangue mineralogy and determine if we have the culprit gangue mineral(s) that we know will diminish the collector. Then a decision would be made to stockpile this ore on waste or the "healing" pile for a rainy day. Since we know this mineralogy test can't be done quickly, we're left with no solution but to wait until the ore hits the tails circuit so that we can get an assay (on OSA or from lab) and decide if our recovery is suffering or not.

The froth cameras have their benefits. Combined with OSA they can be effective tool, but will they indicate the circuit is suffering from lack of collector? I believe this is where a UV unit becomes an important tool to at least rule out this possibility.

I think you would get more value from flipping the description of the goal. Saving money on PAX might not have a big payoff. However, changes in the concentration of free xanthate could be indicative of composition. The supergene ore may be unpredictable on terms of copper species or xanthate robbing gangue species. If you monitor the free xanthate you should be able to react faster to the changes and maintain a consistent recovery of the sulphide portion. Rephrase the goal to 'reduce the loss of sulphides due to changes in composition by monitoring the free xanthate concentration.'

Also, if your circuit has a first rougher that bypasses the regrind or cleaner circuit, then you could monitor free xanthate in the appropriate tail to maintain collector starvation. This way you send only the good stuff through the bypass and protect the con grade.

It seems better to apply free xanthate measurements to circuit stability issues or selectivity issues. Perhaps it can also assist with the dosing of sulphidizers. Overdosing a sulfadizer will depress copper and defeat the purpose. If you monitor the free xanthate it should be in excess when there is not enough sulfadizer and then get consumed approaching the right dose then increase again when you have recovered the sulfadized species. If you could measure enough points with the on stream analyzer, assuming you could measure free xanthate, perhaps it could be better than measuring potential.

Perhaps it is not an exacting technique but could be an awesome tool for operation because it could be used to monitor chemistry which is typically unseen. 

Copper ores PAX + Thionocarbamateconsumption use to go between 4 g/t to 13g/t and it depends on if is Chalcopyrite specie or Bornite an Covellite these two last require SH- conditioning before Xanthate. Sericite gangue like Muscovite in Copper ions dissolved into grinding mill pulp produces an ORP reduction environment (- 270 mv), this silicates also can be activate to adsorb xanthate and frother as main consumers in rougher flotationporride froth texture (PAX 200 g /t ), the residual xanthate finally float the copper ores later in scavenger flotation step, which is not good way. Here think to deslime the argillic material before rougher step or change the mill ORP to (-90mv) by oxidant.

On/In Stream Analyzers have proven themselves in the flotation world for sure. I've worked with Outotec and Thermo Fisher OSA's. Without them in a circuit, it would be difficult to near impossible to run an efficient operation. I can't imagine how they did it in old days without these instruments.

By adjusting the collector addition based on low tails grade, it may still be possible that you will not increase the collector as much as you should to get the tails down. Then you may abandon that and move on to adjust another variable. Meanwhile, low collector in the system, or lack of "free xanthate" in the system, could be the root cause.

When dealing with "supergene" ore, I've noticed that just knowing the head grade is not sufficient. If the ore contains low levels of collector rubbing gangue, then your collector addition rate may be around 25-50 g/t. However, when you hit a zone with high collector rubbing gangue, you may have to go as high as 100 g/t collector to be able to float your target mineral. While the 100 g/t seems extremely excessive, it's what it takes to be able to lower tails grade. A good flot operator may spend hours to finally decide to double the collector addition as he will try to do this in small increments. We have seen some partially oxidized ore with 4% Cu that has produced very bad recovery. On the other hand, we have seen very low head grade partially oxidized ore that has produced great recovery. What I'm noticing is that there is gangue and then there is gangue! Not all gangue are created equally and behave the same. Some gangues are very active at consuming your entire collector before your target mineral has a chance to come in contact with xanthate. Again, I believe this situation pertains strictly to supergene ore zones as far as I can tell. Our idea of testing drill & blast samples prior to sending the ore to the mill is going to help us decide if we should even bother processing that ore. If that ore is a super high collector consumer, then we will put it in the waste dump or something like that.

There is typically a flotation test that we do to determine this, but most times we get a lot better result in the bench test than the mill. Using the UV it will give us another quantified decision making tool.

The testing itself will not take that much time to perform. When the sample is received from the Geology/Mine group, it will be pulverized, mixed with water to a specific pulp density, agitated and aerated, and then a sample will be taken to be mixed with specific amount of xanthate. Then a sample will be taken with a 45 micron filter syringe, and the filtrate will be set in UV unit for adsorption reading. The whole process is within acceptable time limit before the ore pocket has to be mucked out and transported.

Within the mill it's a lot simpler. We will cut a sample from a flot cell, filter the pulp using the 45 micron syringe, and run the UV.

If you had one more tool, a relatively easy one and a quick one, wouldn't you use it to optimize your circuit?

We are still looking for that good flot operator 🙂

You have a good point. Our main goal is not to save money on collector consumption, but to optimize recovery of oxides while ensuring the sulphides are not compromised.

I think you have a good understanding of the benefits of free xanthate measurement in the flotation circuit. We are getting pretty close to running this program.

We tried desliming in the lab, but it didn't show any benefit, but we didn't try any "oxidant" addition to adjust ORP. What type of oxidant would you recommend? Most of our gangue consists of limonite group minerals mainly goethite.

The desliming supposed to be at crushing classification stage and to reduce the potential difference of such galvanic coupling present, I use to apply interactively during grinding stage H2O2 of 10 V/V to get obvious results.