Comminution '16 conference diary

Comminution '16, the 10th in the MEI series, was held at the Vineyard Hotel, Cape Town, in consultation with Prof. Aubrey Mainza, of University of Cape Town, and was sponsored by Keramos, Starkey & Associates, Magotteaux, King's Ceramics & Chemicals, Crush Force, Grinding Solutions, Industrie Bitossi, TOMS, ZEISS, FLSmidth, Cenotec, CITIC Heavy Industries, Outotec, Russell Mineral Equipment and Loesche. Media sponsors were International Mining, Industrial Minerals and AT International, and the Coalition for Eco-Efficient Comminution (CEEC) was an industry advocate.

Monday April 11th

I welcomed the 177 delegates from 23 countries to Cape Town, noting that numbers this year are considerably lower than two years ago, for obvious reasons.

Prior to officially opening the event, I had the pleasure of presenting the 2015 MEI Young Person's Award to Hakan Dundar of Turkey's Hacetteppe University (posting of 11th April).

The search to improve comminution efficiency will undoubtedly include some novel new technologies, but significant benefits are also likely from evolving existing technologies, new ways of operating, and combining the old with the new. In his keynote lecture, which opened the conference this morning, Joe Pease, of Mineralurgy Pty Ltd, Australia, asked whether advances in drilling, blasting, crushing, HPGRs and stirred milling could relegate the SAG mill to a curiosity during our careers?

Joe is chairman of the Coalition for Energy Efficient Comminution, and we are pleased to have CEEC as an industry advocate. Following Joe's keynote, Grant Ballantyne of Australia's JKMRC discussed the use of the CEEC comminution energy curves to assess equipment performance. The database that supports the curves has more than tripled in size in the last year to include more than 1,600 Mt/y of rock throughput at a specific comminution energy input of 17.9 kWh/t. Presenting the data in the form of the energy curves allows operations the opportunity to measure their baseline performance and map relative improvements due to equipment selection and performance. One of the standard energy curves presents the size specific energy (energy required to produce new minus 75 µm material: kWh/t-75µm). The marker size of 75 µm is effective for assessing the performance of most SAG and ball mills, but on some occasions another size might be preferred. Grant described a methodology for converting results conducted with different marker sizes for comparison on the size specific energy curves.

Joe Pease (centre) and Grant Ballantyne (right) with CEEC Director
Mike Battersby of Maelgwyn Mineral Services

Prior to the coffee break, Dave Capstick, of FLSmidth, South Africa, showed how milling energy and CAPEX requirements in the processing of refractory gold ores can be substantially reduced by the use of the Krebs cyclone technology as a desliming cyclone prior to fine grinding in a VXP mill. The cyclone removes the portion of the ore that is already below the target milling size and doesn’t require any size reduction for liberation. This approach by combining two technologies as an integrated bundled design can significantly reduce the project CAPEX by reducing the number of mills required. Also by eliminating a portion of the ore overall milling energy costs are substantially reduced.

The long coffee break in the exhibition and poster area gave everyone their first opportunity of viewing the 15 poster displays and talking to the representatives from the 19 companies exhibiting.

Cobus Kotze, Chris Rule and Brian Loveday at the DMM booth

Following the coffee break, five papers, from Germany, Canada, Sweden and Australia examined ways of reducing comminution costs, by circuit optimisation, reduction of energy consumption and media wear. Holger Lieberwith of the TU Bergakademie Freiberg, discussed selective comminution, the exploitation of selected ore characteristics in order to enrich the valuable minerals in certain fractions while other fractions are left with uneconomic low amounts of valuable material only. A simple classification step will produce a pre-concentrate in this case, while other fractions can be discarded or processed separately. He introduced methods to quantify the degree of selectivity in comminution and presented successful examples with a number of different ores.

After an excellent Vineyard lunch, the afternoon session got underway with a paper from Michael Toifl of Montanuniversitaet Leoben, Austria, on the the influence of irradiation parameters on the microwave-induced damage of rocks for industrial applications. Lorena Guldris then presented the second paper from the large group representing Chalmers University, Sweden. She highlighted the increasing global demand for rare earth minerals and an increased need for development of more efficient extraction processes of such materials. Comminution models commonly predict size reduction with the focus on a single component, but few models integrate the mineral composition. Her paper focused on defining a multi component model, combining size reduction and mineral liberation.

Lorena Guldris (left) with Chalmers University colleagues

Taking us up to the coffee break, Lidia Auret of Stellenbosch University, South Africa, discussed the importance of early detection and root cause identification of abnormal events in industrial processes and in a joint paper from Stellenbosch and Curtin University, Australia, Chris Aldrich, formerly with Stellenbosch, now with Curtin, discussed the importance of the optimal control of mill circuits in the cost-effective operation of concentrator plants, and the neurocontrol of a ball mill grinding circuit by use of covariance matrix adaptation evolution strategies.

Three papers followed the coffee break, the first by Ari Rantala of conference sponsor Outotec. who discussed advanced grinding circuit control using on-line analyser systems.

Speed is a key factor in the efficiency of current multi-stage semi-autogenous and ball mill circuits and more recent latest generation High Pressure Grinding Rolls. A joint paper from the University of British Columbia, Canada, and Ingeteam Power Technology, Spain, presented by Fisher Wang, of UBC, showed how variable speed drives give the flexibility to maintain the process performance and product ore size between desired limits.

At SAG '11 in Vancouver Peter Rubie delivered a paper entitled “Simulation as a tool to enable world’s best mill relining practice” in which he described conference sponsor Russell Mineral Equipment's development of a Discrete Event Simulation-based technology. Managing Director John Russell ended the day's sessions by showing how the continued development, evolution and application of this work is now yielding expected and unexpected results. Its effectiveness has now been proven in all areas of mill relining including concentrator plant design, mill relining technology selection and application, liner design, liner exchange sequence and relining crew performance.

John Russell (centre) with his large team

After a very full day of presentations, it was good to relax with a few glasses of wine in the Vineyard gardens, for the first of our 'Happy Hours' (see posting of 11th April for more on Day 1).

Morning presenter Dave Capstick (2nd left) relaxing with Chris Martin, Alex Wang, Cobus Kotze,
Cathy He and Henry Deng at the first Happy Hour

Tuesday April 12th

As in most fields of endeavour, much of comminution science originates in close association with academic research. The successful development of new comminution science into real business practice is well exemplified by inventions such as high-pressure grinding. But beyond new plants and capital expansions, how is the science helping industry make the best use of the equipment at their immediate disposal? This was a question asked by Rob McIvor, of Metcom Technologies, Inc., USA, in his keynote lecture which introduced the second day sessions.

Plant operations present numerous rewards and challenges to the people who run them. Among their many tasks, metallurgical staff are responsible for the processing performance of the grinding circuits, both daily production and continuous improvement. Means to identify viable process improvement opportunities, and to engage management when needed to achieve implementation, are needed. These include the use of meaningful technical terminology, appropriate metrics, clear economic analysis, and means for follow-up validation that improvements were achieved. Rob presented practical tools for plant staff and company management that bridge the gap between grinding science and daily plant practice. As demonstrated by numerous case studies, these can affect meaningful, sometimes major, grinding process improvements in today’s existing industrial operations.

Following the keynote, Jianjun Tian, of CITIC SMCC Process Technology Pty Ltd, Australia discussed the company's comprehensive solutions, from flowsheet design and equipment selection to circuit optimization, in a paper co-authored with Steve Morrell. Steve is founder of CITIC SMCC Process Technology Pty Ltd which provides independent technical services to various mining projects, mainly in the area of comminution circuit design, technology and equipment selection, and optimization of mineral processing plants. It operates as an independent consulting company, and is ultimately owned by CITIC Heavy Industries, a market leader in China with equipment operating successfully in China and around the world including Asia, Europe, America, Africa and Australia, and it is good to have them involved in the conference as both a sponsor and exhibitor.

Steve Morrell (2nd left) with Spike Taylor of Multotec Rubber (centre) and CITIC's
Jianjun Tian, Richard Wood and Michael Rauscher

Prior to the coffee break, Johnny Kalala, of Hatch, South Africa, described the commissioning and start up of the MMG Las Bambas milling circuit, located in Cotabambas, Apurimac region of Peru. Las Bambas has significant mineral resources and ore reserves of copper, gold, silver and molybdenum. Once in production, it is estimated to have a life of mine in excess of 20 years and in its first full year of operation, it is expected to be one of the top three copper producing mines in the world.

Crushing was the topic of the 6 papers after the break, commencing with Hamid-Reza Manouchehri describing a new generation of Sandvik Mining's high capacity cone crushers, designed to intensify crushing performance in hard rock comminution.

Johannes Quist of Chalmers University, also of Sweden, showed how DEM simulations have been used to evaluate cone crusher performance, and Paul Cleary, of CSIRO, Australia, showed how DEM can be used to predict liner evolution for a cone crusher.

Paul Cleary with Claudio Schneider

Ryan Bracey, of the JKMRC, Australia, also described how DEM modelling has been used to evaluate the performance of the EDS multi-shaft mill, a novel device combining a series of rotating shafts with attached flingers which impact gravity fed material. Currently the mill is at pilot plant scale operating with a F80 below 40 mm and milling up to 80 tph, depending on ore type. A 1:4 scale model of the 10 shaft mill was on view on the Energy and Densification Systems (EDS) stand. South African company EDS was formed in 2005 and the Multishaft Mill is designed and produced in South Africa, EDS having strategically registered international patents in over 45 countries worldwide. The mill's footprint is approximately 2m x 1m and has a height of less than 2.5m and weighs less than 5 tons. The milling action combines a number of processes to provide an output product to match various specifications. These processes are crushing, milling, densifying, blending and attrition.

MEI's Jon at the EDS booth with Ryan Bracey and EDS's Jeff Wain, Boaz Friedland and Andrew Toy

Robbie Brown, of Stone Three Mining, South Africa, discussed various simulated control strategies to improve crusher performance and in a further Chalmers University paper Magnus Evertsson developed a previous theory for lifetime assessment of cone crushers which was used as a foundation for evaluation of full-scale operating cone crushers in minerals processing plants.

Robbie Brown, centre with Lidia Auret and Brandon Wakefield of Stellenbosch University

The afternoon session was devoted to High Pressure Grinding Rolls (HPGR), which are increasingly being adopted in modern plants as tertiary crushers, often as an alternative to SAG mills.

Gerard Van Wyk of ThyssenKrupp Industrial Solutions, Germany, asked if dry finish grinding with HPGRs could be the next step ahead in mineral comminution? Historically, HPGRs have been used mainly as tertiary crushers in mineral applications for the production of ball mill feed. In the cement industry, however, HPGR systems have been successfully applied for grinding limestone, clinker and slag to a final product fineness (P80) of between 30 and 90 µm without the need for downstream ball milling. The total energy consumption of HPGR finish grinding systems in the cement industry has been found to be 30 to 50 % lower than in ball mill systems. This leads to the question of whether the same methodology can be adopted in the mineral industry. Such a step would require the use of dry rather than wet grinding systems. Traditionally, dry grinding systems have been only an exception in mineral processing, since most post grinding processes are wet. However, there are successful, mainly ball mill based, dry grinding examples in gold, nickel and iron ore.

Persio Rosario, of Hatch, Canada, presented a technical and economic assessment of a non-conventional HPGR circuit and Holger Lieberwirth, of TU Bergakademie Freiberg, Germany, examined modelling work with HPGR to predict throughput, specific energy and fineness of the product for edge recycling circuits. Holger also discussed the dynamics in double roll crushers in a further presentation.

Multiple trade-off studies have been estimated assessing the performance of HPGR versus SAG milling resulting in energy savings of between 11 and 32 per cent. But as observed by Grant Ballantyne of the JKMRC one factor that is often misrepresented in these studies is the change in the Bond Work Index. There is a perceived tendency for the Bond test to overestimate the reduction in ball milling energy requirements for HPGR products because of the increased production of fines. Grant's presentation rigorously assessed the actual change in BWi achieved through the use of HPGR technology; outlining the effect of ore mineralogy and the relationship between the laboratory tests and the achieved site performance.

In the final paper of the day Hernan Anticoi, of Universitat Politècnica de Catalunya Barcelona Tech, Spain, discussed the use of population balance models in determining energy consumption applied to high pressure grinding rolls, after which during the coffee break delegates were invited to hear Aubrey Mainza and Indresan Govender present a short introduction to the potential use of Positron Emission Particle Tracking in developing comminution and classification models for design, optimisation and control (see posting of 21st March for full details).

And then an hour later coaches took us to the nearby Kirstenbosch Botanical Gardens for an informal evening of food, wine and entertainment in an idyllic setting under Table Mountain (see posting of April 12th for more photos of the dinner and of Day 2).

Dinner at Kirstenbosch

A very informal evening

Wednesday April 13th

Day 3 began with another interesting keynote, from Malcolm Powell, of JKMRC, Australia. He began by saying that predicting the performance of the whole mining process is surely the holy grail of predictive simulation, which he calls ‘integrated process prediction’. The objective is to be able to predict the response of the processing chain, from mining to concentration, to changes in the ore type. This can apply to greenfields design - selecting processing options to optimise the overall mine performance within the constraints and aspirations of the greater community and environment; expansion and optimisation of existing plant to deal with changes in economic drivers and ore types; and on-line control - predictive reaction to ongoing changes in ore processing types in an operation.

In order to achieve this there are several broad key components that are required: ore characteristics based on in-situ rock properties; prediction of how the rock fragments; the ability to predict and track progressive mineral liberation of the rock through the reduction process; the response of rock to separation processes as a function of the rock mineralogical properties. Malcolm looked at the value of this integrated approach; the challenges it poses; the technology and capability we require; and how as a comminution community we can address the challenges.

Malcolm Powell (right) with Arno Kwade and Indresan Govender

Okay Altun of Hacettepe University, Turkey, discussed closed circuiting HPGRs via air classification, showing that closing the circuit was more energy efficient than open circuit operation.

Loesche Vertical Roller Mills (VRM) have been used since the 1930s to grind cement raw material. They operate with two, three, four or six rollers and combine drying, grinding and classification operations in one unit. Hacettepe University is one of the world leaders in cement grinding and Hakan Benzer discussed evaluation of the performance of the mill in mineral ore processing. Following on from this, Warren Little, of the University of Cape Town, discussed the potential of the Loesche VRM operated in closed circuit with an air classifier as an alternative high compression grinding mill for the platinum industry.

Hakan Benzer (right) with Magnus Evertsson and Aubrey Mainza

Carsten Gerold, of Loesche, Germany, then showed how a VRM has been used succesfully in a hard rock application at Foskor Pty (Ltd) in Phalaborwa, South Africa for the past 16 years. In 1999, Foskor upgraded plant production capacity from 2.1 million tons to 2.7 million tons making use of a Loesche dry VRM process with a Loesche Mill type LM 50.4 over the wet rod and ball milling process used elsewhere at Foskor.

SAG milling was next on the agenda, with two papers from Australia and South Africa taking us to the lunch break.

Paul Cleary, of CSIRO, showed how applying DEM to prediction of tumbling mill performance is challenging because several different modes of breakage are active in the process. His presentation discussed the prediction of the evolution of rock size distribution, throughput and product size in AG and SAG mills by incremental damage, chipping, rounding and abrasion mechanisms.

Adrian Hinde, of AH Consulting, South Africa provided a critical review of the quantification of breakage kinetics for AG/SAG mills in terms of a single function, the so-called cumulative rates function, which has received attention in the literature for more than twenty years. His review emphasised predicting both the steady state and dynamic performance of single-stage and primary AG/SAG circuits. One of the challenges of modelling such circuits is predicting the behaviour of the mill at throughputs leading to ball and total charge loads above those corresponding to the maximum mill throughput capacity or power draw. This region of operation, which is inherently unstable unless appropriate control action is taken, leads to the generation of grind curves needed to guide the optimisation of AG/SAG circuits.

SAG milling also dominated the afternoon session, which started with a further paper by Paul Cleary showing how a coupled DEM-breakage-SPH model could be used to predict coupled rock breakage and slurry transport in a 3D pilot SAG mill.

Engineering feasibility studies have become a significant cost in the mining industry. Starkey & Associates Inc., of Canada, has completed a feasibility level grinding study in five weeks on the Borden Gold Project, now owned by Goldcorp. John Starkey described the testwork campaign and the steps taken to facilitate a rapid analysis turnaround. The program included a site visit, sample selection, SAGDesign testing on ten samples and a final mill sizing analysis. It was estimated that a comparable comminution program of this detail would require three to six months to complete.

John Starkey (3rd left) and Spencer Reeves of Starkey & Associates, with
Peter Radziszewski and Brian Loveday

Feed preparation plays a key role in the performance of circuits involving SAG milling. Variations in the feed size distribution affect the throughput and circuit stability which leads to fluctuations in the product size distribution. The inconsistencies in the product size lead to loss of recovery downstream, especially in situations where coarse unliberated particles are produced due to unstable operation in the comminution circuit. Aubrey Mainza, of University of Cape Town described surveys carried out to assess the influence of feed size on the performance of the Tarkwa Gold Mine CIL circuit in Ghana, conducted when the plant was treating ore with different feed size distributions. He also discussed the importance of treating a consistent feed size distribution on the stability of the circuit and the benefits of processing feed with a reasonably good balance between coarse rocks that act as media and fines in the feed on both throughput and product size.

Aubrey's co-author Paul Bepswa presented an assessment of different ore responses to changes in SAG mill operating conditions at Kansanshi Mining in Zambia. The Kansanshi copper ore body consists of distinct sulphide and oxide zones as well as a transitional zone called the mixed ore zone. The three zones exhibit distinctly different mineralogical characteristics. As a result, the mine employs different metal extraction routes resulting in the operation of three parallel grinding circuits each dedicated to treating one of the three ore types. The installed power for each circuit reflects the distinctly different strength characteristics of the three ore types. A series of optimisation surveys performed recently at the operation found that the throughput capabilities of each circuit depended primarily on the operating conditions imposed on the respective SAG mills in each circuit. The paper assessed the responses of each ore type to changing SAG mill operating conditions using the SAG Model in JKSimMet as a diagnostic.

Elvis and Johnny

The Freda Rebecca Gold Mine (FRGM) is situated near the town of Bindura, north-east of Harare, Zimbabwe. FRGM comprises a shallow underground operation and processes ore through a single facility utilising a combination of crushing, conventional SAG milling, a combined gravity and CIL process, electro-winning and bullion smelting. Secondary crushing of the ore before the two SAG mills has been integrated in the circuit in order to increase the throughput by 10% while maintaining the fineness of grind. Johnny Kalala, of Hatch, South Africa, and Elvis Mudoti of FRGM gave a joint presentation, describing the optimisation of the secondary crushing stage before SAG milling.

And then to finish the day, a complete change of topic, with Elizma Ford, of Mintek, South Africa evaluating the potential throughput benefit of adopting Derrick fine screening technology in a PGM slag ball mill circuit. She discussed how a PGM producer was interested in expanding the capacity of one of their conventional ball mill circuits closed with hydrocyclones. Mintek was contracted to quantify the effect on circuit throughput associated with changing from the current hydrocyclones to Derrick fine screening technology. The paper described the modelling approach followed as well as the results from simulation work to quantify the potential for throughput improvement. It is becoming well known that the ability of these machines to accurately classify by size only at efficiencies in the mid 90% range, as fine as 45 micron, has resulted in a paradigm shift in milling circuits, replacing hydrocyclones in the closing of secondary and tertiary circuits. Talking earlier to the Derrick representatives, I was interested to learn that the company, based in New York State, is a third generation family owned business.

Elizma with Derrick's Jean Makola of South Africa and Baojie Zhang of USA

Following questions, it was back into the Vineyard gardens to enjoy the second of our Happy Hours (more photos of Day 3 on the April 13th posting).

  

Thursday April 14th

Sandile with co-author Brian Loveday

The final day got off to a good start with a presentation by Sandile Nkwanyana of the University of KwaZulu-Natal who showed that adding pebbles to a ball mill can be a very attractive option for improving grinding efficiency. Marcelo Perrucci of ABB, Switzerland then described grinding milll optimization through SmartMill technical and commercial analysis, an embedded control for variable-speed drives, that utilizes online data for continuous, standalone control of mills.

Claudio Schneider of CETEM, Brazil discussed coarse particle milling in what he called "the abnormal breakage region". When particles are relatively large with respect to the grinding ball diameters that constitute the mill charge they are not properly nipped by the balls and tend to remain unbroken in the holdup. The particle sizes that can be fed to a ball mill depend on the ball size distribution of the charge and more specifically, the diameter of the make-up ball. This is also a function of the tenacity of the particles as this controls the size at which the selection function is maximum for a given ball charge. Particles that are larger than the size at which the selection function is maximum are said to be in the abnormal breakage region. There are a number of situations in which it is desired to feed the mill with coarse particles, particularly in primary milling stages. This allows for a reduction of the number of crushing stages and crushers before the milling stages. Also, smaller diameter ball charges are usually more efficient. In some cases a knowledge about the behaviour in the abnormal breakage region may allow for substituting conventional ball mills with the more efficient Vertimills, loaded with ball charges of smaller diameters. Characterizing for the abnormal breakage region is also important for understanding ROM mills ad AG/SAG mills as in these cases a considerable number of impacts will be in the abnormal region.

And then the rest of the day was devoted to stirred milling, reflecting the increasing importance of these mills in the minerals industry. Two papers from the Technical University Braunschweig, Germany examined axial grinding media distribution and rheological effects in wet stirred media milling.

The Outotec HIGmill is a relatively new technology and concept for fine grinding applications and is based on existing, well proven stirred milling technology. Following the launch of this vertical fine grinding technology in the metalliferrous mineral processing industry, Outotec embarked on an intensive testing campaign and three papers from Outotec discussed the development of the mill.

In February 2015, First Quantum Minerals (FQM) commissioned a new fine grinding circuit targeted at boosting the metallurgical performance of its Kevitsa copper/nickel concentrator operation in Finland. At the heart of this circuit is a 700 kW HIGmill producing a 20 micron regrind product for enhanced mineral particle liberation with improved downstream flotation performance. The start-up of this fine grinding circuit provided a unique opportunity for scale-up evaluation. Harri Lehto, of Outotec Finland, discussed the main outcomes of the recent developments of test validations and benchmarking against an industrial size unit, the overall performance of HIGmill in Kevitsa concentrator plant and observed plant performance improvements.

Alex Heath of Outotec Australia then described the development of a grinding power model to predict the power draw of the HIGMill and Hanspeter Erb, of Swiss Tower Minerals Ltd, Switzerland, reviewed the metallurgical performance of the mill in a primary milling application in the hard rock mining industry.

Alex Heath and Angie Voges of Outotec demonstrating the virtual reality system
for training operators in mill relining

Taking us up to the lunch break were two papers describing the performance of another type of stirred mill, Metso's Stirred Mill Detritor (SMD). This is a fine and ultrafine grinding mill that is used with feed sizes up to 200 microns and product sizes below 10 microns. It is currently being employed in many tertiary, regrind, and ultrafine grinding applications around the world. and Adam Moore, of Metso, USA explored the best practices for implementing this unit in mineral processing applications.

As discussed by Simon Bailey, of Grinding Solutions Ltd, UK, one of the key differences between SMDs and other types of stirred mill (IsaMill, for example) is the shaft orientation and stirrer type. This, along with the geometry of the mill, results in the formation of a vortex in the mill charge. The shape of this vortex is determined by mill and rotor design, impeller speed and charge (media and slurry) properties. He described an investigation, in collaboration with Magotteaux, into the effects of mill charge on power draw and grind size showing that a spike in power draw is obtained under operating conditions that results in an unstable vortex. Operation of the mill in unstable vortex conditions is shown to yield poorer grinding performance than operating in more stable conditions. The study demonstrates that industrial SMD operation can be optimised and energy consumption reduced by understanding and controlling better the mill speed and charge, in order to maintain stable vortex conditions.

Felicity and Nick Wilshaw and Simon Bailey of Grinding Solutions, with MEI's Barbara

The final session continued with the theme of stirred mills. Peter Radziszewski, of Metso, Canada, discussed the effect of pressure profile on stirred mill impeller wear, an area which has received little attention in the literature.

In dry grinding processes, which are usually designed as mill-classifier circuits, fine powders are not easy to produce and to handle. Particle-particle interactions which increase with decreasing particle size lead to agglomeration and material adherence and, thus, to a lower energy efficiency. The different challenges concerning grinding performance and material transportation can be solved by an appropriate application of grinding aids, and Paul Prziwara of Technische Universität Braunschweig, Germany showed how a new laboratory vertical stirred media mill for dry fine grinding has been developed in order to investigate grinding aids in a more representative way than using common laboratory tumbling and planetary ball mills.

Minas-Rio is an Anglo American property located in Brazil that started its operation at the end of 2014. The Minas-Rio regrind circuit is currently the largest Vertimill installation in the world with sixteen VTM-1500 Vertimills installed in closed circuit with hydrocyclones. Claudio Schneider, of CETEM, Brazil, described how the grinding efficiency of this circuit was verified through a sampling campaign and the results showed that the measured specific energy consumption was similar to the predicted specific energy consumption estimated at the engineering design stages of the project. He discussed the results obtained from the Vertimill model for scale-up and simulation, which has recently been developed based on the population balance modeling technique and validated with extensive data from Vertimill pilot scale tests.

There are many companies worldwide, some exhibiting at the conference, who produce ceramic beads for stirred mills so there is variation in formulations and processes which affect the cost of the media and impact on competence and performance. Paul Hassall, of SEPR Saint-Gobain ZirPro, France, discussed the behaviour of ceramic beads in ultra fine grinding mills, reviewing the developing morphology of ceramic beads during extended operations in high energy stirred mills. Mills operate with a ‘working mix’ or ‘seasoned charge’, invariably meaning that worn beads of various sizes are present in the mill. These beads can effect mill efficiency, overall media wear and particularly mill wear and Paul reported on preliminary findings of wear profiles of typical ceramic media and some of the potential effects of extended use.

It was good to see three delegates representing regular sponsor, Chinese company King’s Ceramics & Chemicals Co. Ltd, at the conference again. Alex Wang presented a paper showing how grinding media affects wet grinding in stirred mills, reviewing the relationship between stress intensity of grinding media and grinding efficiency. He showed that grinding efficiency can be optimized by choosing suitable graded grinding media with optimum density.

Jon and me with King's Alex Wang, Cathy He and Henry Deng

Chris Greet is a regular attendee at not only the comminution series but also the flotation conferences, and his company, Magotteaux Australia sponsors both series, so it was appropriate that Chris should end the conference with a very thought provoking question: does stirred milling have an impact on the pulp chemistry and subsequent flotation response? In recent times the application of stirred mills to regrinding duties within base metal sulphide flotation plants has increased in popularity, particularly as the target liberation size for cleaner flotation has decreased to below a P80 of 38 microns. However, while these machines provide the means to achieve the desired size reduction efficiently to obtain liberation very little attention has been paid to the impact on the pulp chemistry of the system and how this will influence the subsequent separation process. Chris examined the changes in the pulp chemistry (the pH, Eh, dissolved oxygen, temperature and oxygen demand) across a stirred mill grinding a copper rougher concentrate prior to cleaner flotation. It was apparent that while the regrind circuit product has the right particle size distribution to produce a high degree of liberation, the pulp chemistry is not necessarily optimal for efficient separation via flotation. That is, the stirred mill discharge has a reducing Eh, low dissolved oxygen concentration and a high oxygen demand. These pulp chemical properties can be conducive for pyrite flotation, which leads to dilution of the copper concentrate. An interesting conclusion to end the conference, and maybe provide the inspiration for papers at Flotation '17?

Chris Greet (right) with Joe Pease and Steve Morrell

In his excellent summary of the conference, MEI Consultant Aubrey Mainza noted that energy was highlighted as one of the major challenges but there were no papers presented that addressed water utilisation, which is becoming just as important as energy, and may be more important in the future. He remarked on the fact that, although there had been excellent papers presented, and many innovations on display at the exhibition, the people who would have benefitted a great deal, the operators, were to a large extent noticeably absent. He felt that, although times are hard at present, operators would have benefitted a great deal from what was on offer and they should seriously consider being present at Comminution '18.

In closing the conference, Amanda thanked everyone, sponsors, exhibitors, presenters, chairmen, delegates and the wonderful staff of the Vineyard Hotel for a very memorable week, and announced that Comminution '18 would be held at the Vineyard again in April 2018.

More on day 4 on the April 14th posting.

All the technical papers, and most of the poster papers are available in draft form from MEI Online, and authors have been invited to submit final papers for peer-review for possible inclusion in a special comminution issue of Minerals Engineering.

 

Twitter @barrywills