Tuesday, 21 August 2012

Flotation of chalcocite at very low pH

In my recent report on Zambia's Nchanga mine, I drew attention to the sulphides which were floating as a froth in the Tailings Leach Plant (TLP) leach pachucas (right).

Ian Pendry asked "How much copper is lost to this froth, what is its mineralogy, and why are the sulphides floating apparently readily here rather than in the conventional flotation circuit which precedes the TLP".

During my pre-visit briefing two weeks ago today, I was surprised to hear it mentioned that copper loss to final tailings is around 25%, the main loss being due to chalcocite. I do not know how much of this chalcocite is in the froth on the top of the air-agitated pachucas, or how much of the tailings loss is in this froth, but the fact that there is such an apparently rich froth is interesting.

The sulphide flotation plant is operated at neutral pH, but the pH in the pachucas is around 1.5. As xanthates decompose in acid media, is collectorless flotation taking place? Can anyone offer any possible explanations; does anyone have any experience of collectorless flotation of chalcocite and other copper sulphides at very low pH?

Having asked this question, I have a feeling that what is seen in the pachucas is likely to be only a very minor build-up of froth, and probably accounts for very small overall losses. It may be that the froth is merely entrained from the sulphide flotation circuit, and it is difficult to remove this from the bottom-pumped pachucas.

So a more relevant question might be- can anyone suggest how Nchanga might recover the high losses of chalcocite in the final tailings? Is chalcocite generally a poorly-floating mineral, and is anyone currently involved with research on its flotation?


  1. What's the final Cu concentration in the tailings? Chalcocite is easily leached.

    1. I will try to get more info from Nchanga

    2. Dr. Fuzail Siddiqui P. Geo.1 August 2013 at 03:10

      1) If my memory serves me right the Concentrator final tailings ranged from 0.5% on a 'good' day to 1+% total copper on a 'bad' day. Of course tailings were not sent to the tailings dam; these were/are pumped as feed to Tailings Leach Plant (TLP) where more copper is recovered through a leach/SX-EW circuit. The leach residue, ranging in copper content from <0.1% to a maximum of around 0.2%, is neutralized and pumped to residue pond/s.
      2) As to the statement that 'chalcocite is easily leached' testwork was done on different purified and milled copper sulphides, using the empirical conditions obtaining in Nchanga leach circuits, to quantify leaching characteristics. About 15% chalcocite leached compared to only about 5% chalcopyrite.

  2. The following comments have been received via the Minerals Engineers Group (http://tinyurl.com/cgdzbap):

    Jackson Sikamo • The effectiveness of Chalcocite flotation or any sulphide for that matter will depend heavily on the overall mineralogy, acid soluble copper to total coper ratio, the gangue matrix, and the stability of the ore characteristics in relation to these parameters. The Nchanga scenario needs to be evaluated as a special case with perculiar challenges in getting a stable feed material over time. The flotation taking place in the leach pachucas indicate inadequate flotation time at the concentrator. The Nchanga ore flotation is a long story and requires a full analysis at site with the plant metallurgists and operators.

    Gerald Cadwell • I agree with Jackson that the flotation taking place in the pachucas is due to short flotation time as it seems as if the chalcosite is very slow floating and as Jackson says a full analysis on site with the plant metallurgists and operators would be needed before a meaningful answer could be given. we would need to analyse the flow sheet and the reagent suite before making a recomendation.

    Zongfu Dai • Chalocite can realily float in a wide range pf pH (1 to 13), and it can easily be oxidized.

    If it does not float in the flotation circuit but floats in the leach circuit, there are two possible reasons: (1) too short flotation time in the flotation circuit; (2) over oxidation in the flotation circuit and removal of the oxidazation products from the surfaces under very acidic conditions of the leaching circuit.

    If it does not float in the flotation circuit and does not float in the leaching circuit, either, the reasons may be due to too fine particle size, poor liberation from non-floatable gangue minerals, or soemthing else.

    Credo Nguni • It would be interesting to establish the required flotation time at Nchanga concentrator.

    Dr. Agar has published credible methods for determining flotation time,

    The question is, will the copper mineral loss at Nchanga be significantly reduced under optimal flotation time?

    Ravi Sathe • Well, flotation at Nchanga has been a subject of intense interest and research for a very long time and I am pleasantly amused to see it still evokes same interest with same zeal.

    At Technical Services (TS) of the then Zambia Consolidated Copper Mines Ltd. we did lot of work on the subject of flotation at Nchanga in mid 80s and some of that was published in a paper at Copper 91 International Symposium in Ottawa, Canada. I was one of the authors.

    Chalcocite responds well to flotation. In Nchanga though it is lost mainly to very coarse and very fine size fractions of the concentrator tailings. This is because Nchanga flotation feed contains atypically high proportions of these fractions even at a moderate 55% passing 74 micron grind, a characteristic of Nachanga ore. To address short flotation time, large mechanical cells were installed, if I remember correctly, in mid 90s. That substantially addressed the issue of flotation time.

    Some of chalcocite lost to the tailings gets activated due to intense scrubbing and action of acid in the Tailings Leach Plant (TLP) environment and seen to float in pachucas. Lot of work was done at TS to recover sulphides lost to the TLP residue. After lots of lab work and a successful pilot run a flowsheet was proposed for plant scale incorporation but never got implemented if I am right.

    Jackson Sikamo • Ravi,

    Good to link up with you. The Nchanga leach residue flotation circuit was implemented with stainless steel flotation columns for both roughing and cleaning. Unfortunately the cencentrate generated was low in copper and extremely high in insols and the concentrate could not be treated in the Nkana or Mufulira smelters and the project was aborted. The capital spent was never recovered. The predicted pilot plant flotation results in terms of both recovery and concentrate quality could not be realised.

  3. More from Minerals Engineers:

    Bwalya Chishimba • One other challenge with the nchanga ore is the presence of cupreferous mica which renders most sulphides especially chalcocite difficult to float. Chalcocite mineral is mostly present in the open pit material and part of the percentage lost to the tailings is as a result of the mica's. However to talk on the residence time (flotation time) i will say tests are underway to optimize the residence time as Mets are working around the clock to address this problem and many other challenges. The flotation time is mostly dependnt on the through put the high the through put the lower the residence time, we will greatly appreciate all the help we can get from your great contributions especially those who are very familiar with the nchanga ore like Mr Rav and Mr. Sikamo

    Ravi Sathe • Jackson, Thanks for the update on TLP residue flotation. Sad, the projected results could not be realized.

    Ravi Sathe • Mr Chishimba, we found during our work that carbonacious shell which is found in abundance in Nchanga ores too can be a villain.

    Credo Nguni • Well I am back at Nchanga and operating a leach residue pilot flotation circuit. Unlike the column pilot work I carried out in 1992, todays pilot plant is using mechanical cells based on a lab. circuit a developed this april. With management,s approval I should be able to share the projects results in the near future.

    Ravi Sathe • Credo, That is interesting. Please keep us posted on the results if you can.

    Stephen Gay • A slight aside...
    When I did work for Olympic Dam the chalcocite floated at a different rate to chalcopyrite (I understood it was faster), and I understood this was well-known; however the modelling framework grouped the CuS minerals, rather than allow distinct minerals to maintain their integrity. This caused heaps of calculation issues.

    My main point being that the different mineral species to float at different rates and this needs to be included in the modelling software.

    Ravi Sathe • @Bwalya Chishimba: It is good that you are taking a fresh look at flotation time at Nchanga and I wish you and your mets a success. Just a reminder: Lots of work has been done on this count in mid to late 80s at Technical Services. If you can find reports of that time span it will help in your current endeavor.

    All said, ore characteristics are highly variable at Nchanga due to a variety of factors. It is therefore not easy to identify a flotation time as "the flotation time". Good luck.

    Bwalya Chishimba • Thank you Very much Ravi we will try and find the data so that we can compare. we are geting there though just fractuations in the through put has been hamping a complete success on the project but the calculated values are matching with experimental values...

  4. a) MMBO - You are probably already aware of this but………
    -Flotation with MethylMercaptoBenzoOxazole (Chelating Reagents for Flotation, Marabini et al…presented to Flotation ’09 in Cape Town) gives good selectivity in the presence of other sulphide minerals such as galena and sphalerite. MMBO type reagents exert a chelating action towards Cu (15) thru weak dative bonding of the Nitrogen and a saline bond of –SH. The oxygen atom of the oxazole ring favours chelation through an electron releasing conjugative effect. Three structural requisites appear essential to ensure collecting power; The position of the alkyl group on the benzene ring vis a vis the chelating functional group, with the optimal having the aliphatic group diametrically opposed to the electron-donor heteroatom. The structure of the alkyl group is best when it is a straight chain with an ether oxygen atom at the point of attachment to the aromatic ring. The length limit of the chain necessary to ensure adequate hydrophobicity ranges between three and six carbon atoms (all observed on the basis of steric factors and the effect of the aliphatic chain on the reactivity of the functional group. The 5 hexyloxy-2ATP molecule gave excellent collecting power.

    b) Thionocarbamates – We manufacture a range of Thionocarbamates at a plant in Qingdao China. Amongst other minerals, we have been testing a variety of new molecules on Chalcocite. So fare one of these has performed better than all of the others……however the degree of success is a case by case question. We are continuing the research and if will let you know how it goes in due course.
    Lex Tall, Tall Bennett, Australia

    1. Thanks Lex. The paper by Marabini et al that you refer to was actually presented at Reagents '06, and published in Volume 20 Number 10 (2007) of Minerals Engineering. It can be downloaded from ScienceDirect at http://tinyurl.com/9qmwdc7

  5. And more from Minerals Engineers at LinkedIn:

    Paul Bwalya • @ Bwalya Chishimba: In case you need to check on the archives for TS data, the Chamber of Mines Library in Kalulushi had a lot of historical work done during the ZCCM times. Am not sure if the place is still functional and open daily.
    @ Credo: who is running the cabalt business now that you are back at the concentrator?
    Jacques Bezuidenhout • i would be interested to find out what their reagent suite are; sounds to me like its a classical case of improper reagent selection
    Teresito (Terry) Malicse • Chalcocite floatation is pH and Eh sensitive.. and need to study a good modeling in the lab scale.. then upscale to plant level. One of the things that could work is a "good" attrition conditioning, and get red of the fouled up solution and replace substantially with new water addition.
    Peter Riccio • Fine (less than 10 micron) particles of chalcocite are historically difficult to float. The collector/frother suite may need to be reviewed to acheive better flotation kinetics but more importantly, at what size do the majority of the losses occur?
    Tebogo Sakaria • @ Teresito, As pH & Eh sensitive it is, i wonder if their instrumentation is uptodate. An optimized and automated reagent dosing interlinked to the changes in pH & Eh can be a good solution to their problems. Some of these problems are just made difficult because companiyes dont want to invest in morden technology and relies in operators who at times are not equiped to understand the fundamentals lof this minute parameters.


  6. George Widelski • I'm intrigued by the comments that chalcocite flotation is pH sensitive. In the literature and from personal experience and observation, chalcocite floats freely up to very high pH levels ~13. In a mixed sulphide circuit we had to cut back collector addition to starvation levels to prevent the chalcocite from floating too rapidly which would lead to the froth collapsing. In supergene porphyry ore projects, I have not come across anyone having problems floating the chalcocite/digenite components. I will be very interested to hear of any unusual conditions at Nchanga which may be causing this unusual behavior.
    Jacques Bezuidenhout • i agree; unless the mineral actually is tenorite and not chalcocite. They look similar.
    Stephen Grano • I agree with George above. A possible issue with chalcocite that I have noted is that it may develop a very high contact angle compared with other copper sulphide minerals. Depending on the chain length of the collector which may induce very high contact angle, the froth phase may become unstable, very shallow, and difficult to remove from large plant cells. This may cause low recovery from the cell/s and operators try to minimise the collector addition in response. In some plants the froth mobility changes down the bank of cells due to changes in mineralisation in the froth (% solids, contact angle) and the froth mobility may actually improve. A remedy is to use a short chain length collector (ethyl xanthate) prior to rougher flotation and then, if necessary, a longer chain collector in scavenging to recover coarse, composite particles as one is trying to maximise the contact angle on the value mineral surface. The chain length of the collector is important as well as pH and grind Eh. Of course, a short chain length collector will require more collector, but it may be easier to avoid froth collapse and loss in recovery. I know of one plant where that was the remedy. The Ian Wark Research Institute does research in this area.

  7. Fuzail Siddiqui27 July 2013 at 15:06

    Barry Wills, who authored the blog that led to this thread, left Nchanga in 1973. I arrived there in November 1973, especially hired as a full time Process Mineralogist. My co-worker was Karen Dalgleish who focused on estimating the mineralogical composition of thousands of drill core samples while I focussed on plant samples. I remained in NCCM/ZCCM till 1998 and although I changed many jobs I never lost touch with process mineralogy. I do not recall meeting Barry so we must have missed each other by a few months. I do remember talking to Ravi and recently re-established contact with Jackson Sikamo. Any way, let us move from this trip down memory lane to the subject of this interesting thread, namely, the flotation of 'chalcocite' in the leach pachucas. 1) first of all has anybody confirmed that what is floating is indeed chalcocite? I wonder if Jackson will be kind enough to arrange an XRD, to determine the mineralogical mix definitively, and share the results in this thread. I withhold comment till the mineralogy is established without doubt. 2) As to sulphide losses to tailings I recall pointing this out to successive concentrator and metallurgical supt's. Almost invariable these losses were due to sulphide particles that were too coarse to float. I even recommended that we introduce a step to regrind the coarsest fraction of the feed to bring the sulphide particles in floatable size range, but may be it was thought to be not worth the trouble. 3) One could see these coarse sulphide particles settled in many TLP launders; some chalcopyrite particles as coarse at 0.5 mm! 4) I remember the installation of the leach residue flotation columns, that didn't work but by then I was out of Nchanga to Kabwe, Luanshya and back to Nchanga years later. 5) With regard to sulphide losses from the concentrator, another factor came to light towards the end of my stay in Zambia. At that time I introduced the Site Daily Mineralogy program (I called it the SIDMIN program). This program resulted in delivery of a semi-quant. report on ore mineralogy of selected concentrator products daily and simultaneously with the daily assay report. Jackson remembers this well as he was the Concentrator Supt. at that time. 6) These daily reports demonstrated that even particles in floatable size range were escaping in concentrator tailings. Why? Jackson then put this question and demanded an explanation from the Mets. Someone came up with the idea that perhaps the reagent addition point was too close to floatation cells and may be the sulphide particles were not getting enough conditioning time to become hydrophobic. The reagent addition point was duly moved back and lo and behold there was instant improvement in recovery of sulphides. I presented a paper on the SIDMIN program at the Annual Canadian Mineral Processors Conference at Ottawa in 2000, a copy of which was provided to Jackson when we established contact again a few months ago. Those wishing to read a copy can send me an email at siddiquf@yahoo.com. The paper makes interesting reading for those interested in the dramatic positive impact that a combination of chemical and mineralogical information can make to plant recoveries. 7) For example I was alarmed at the amount of easily leachable oxides, such as malachite and azurite, in the TLP leach residue to which I drew daily attention of Jim Moore the then TLP Supt. He put the question to his Mets. It turned out that 3 out of 4 leach pachucas were non-operational resulting in quite inadequate residence time in the solitary operational pachuca. The rest of the pachucas were repaired fast fast and TLP recoveries improved by an unbelievable 4 to 5% plus! Sorry for this long contribution but this thread brought back memories and although the last post was over 9 months ago I venture to comment, just in case some of the past contributors are able to notice this contribution.

    1. Thanks very much for your very interesting comments Fuzail. I was particularly interested in your suggestion that the sulphide losses to tailings might be due to particles too coarse to float. During my recent visit to Metplant ’13 in Perth (posting of 20 July), I saw a demonstration of Eriez’s HydroFloat Separator, which can, it is claimed, float sulphide particles as coarse as 1mm. This might provide a solution to the Nchanga tailings problem? Eriez will be presenting a paper on the device at Flotation ’13 in Cape Town in November.

    2. Dr. Fuzail Siddiqui P. Geo.31 July 2013 at 13:00

      Barry, Nchanga's ore mineralogy mix is so unique. While it is a nightmare for metallurgists, it is a 'dream come true' for process mineralogists. Each shift brings its own challenges. One of the areas I tried to improve on was over- reliance on arbitrary assay limits as criteria of success of metallurgical operations. What worked best was a combination of 'daily composite assays' and 'daily composite ore mineralogy'. It is this combination that identified problems in real time for remedies in real time. Unless this approach is adopted, both at test work stage and at operations stage, Nchanga may often prove to be a graveyard of promising technological innovations; the failed TLP residue flotation columns are a case in point. As to whether Eriez's HydroFloat Seperator can be a solution would depend on results of very careful test work, on the above lines, on Nchanga concentrator feed samples.

  8. A good size-by-size database on mineralogy will inform the proper route to follow here. I love the comment regarding this ore being a Process Mineralogist's dream come true! Just be sure it is supported by good optical work and not 'cookie-cutter' automated mineralogy, as the nature of the chalcocite (sooty/microporous/coated) is key. Chris Hamilton


If you have difficulty posting a comment, please email the comment, and any photos that you might like to add, to bwills@min-eng.com and I will submit on your behalf