Mineral processing case studies, 2025–2026: what plants reported at recent conferences

By /

Between March 2025 and October 2026, operators and suppliers presented plant data at three conferences: Comminution ’25 and Flotation ’25 in Cape Town and the AusIMM Mill Operators Conference in Brisbane.[14][16][17] We read the published papers and checked them against company disclosures. Nine case studies follow, covering fine and coarse particle flotation, high-intensity cells, variable-speed SAG milling, ore sorting and plant diagnostics. Several explain how the operator measured the gain, which is often the most useful part. Every figure below is taken from the numbered source listed at the end of the article. Our October 2026 news roundup listed the Mill Operators program; this article covers the results.

Recent mineral processing case studies at a glance

Operation Technology Reported result Conference Source
Kevitsa, Finland (Boliden) Concorde Cell in nickel cleaning 80% Ni recovery at an enrichment ratio above 1.6 Flotation ’25 [1]
Oyu Tolgoi, Mongolia Magnetic conditioning in the cleaner circuit Higher copper recovery and grade, driven by the minus 25 µm fraction Flotation ’25 [2]
Northparkes, Australia Sodium silicate dispersant, 100 g/t Gold recovery from 73.1% to 74.4%; adopted permanently Flotation ’25 [3]
Mantos Blancos, Chile (Capstone Copper) NovaCell pilot on rougher tailings Up to 56% copper recovery from rougher tailings Flotation ’25 [4]
Carrapateena, Australia (BHP) HydroFloat coarse particle flotation Commissioned in 2025; no performance data published Mill Operators 2026 [5][6][14]
Fruta del Norte, Ecuador; Caraíba, Brazil Jameson Cells in rougher duty Gold recovery from 87.1% to 90.4%; copper throughput from 400 to 600 t/h Flotation ’25 [9]
Ban Houayxai, Laos (PanAust) Variable-speed drive retrofit on a SAG mill Specific energy from 9.4 to 8.8 kWh/t; media use down 16% Comminution ’25 [10]
Tallebung, Australia (Sky Metals) Two-stage XRT ore sorting Tin from 0.17% to 2.32% at 94.8% recovery in stage 1 Mill Operators 2026 [11][14]
Constancia, Peru (Hudbay) Flotation bank survey with mineralogy Entrainment of 15% to 40% in the first rougher cells Flotation ’25 [12]

Fine particle recovery: Kevitsa, Oyu Tolgoi and Northparkes

Kevitsa: a single Concorde Cell compared with a cleaner bank

Boliden commissioned a Metso Concorde Cell in the nickel cleaner circuit at its Kevitsa copper-nickel-PGM concentrator in Finnish Lapland in January 2025. The Concorde Cell is a forced-air pneumatic machine. It mixes slurry and air in a high-shear blast tube, which raises the chance that ultrafine particles attach to bubbles. Kevitsa ore grades about 0.3% copper and 0.2% sulphide nickel, and pentlandite is the main nickel mineral.[1]

The circuit can feed the cell either nickel rougher concentrate or third-cleaner tailings. The Flotation ’25 paper by Metso and Boliden metallurgists reports these results:[1]

  • On rougher concentrate, the cell recovered 80% of the nickel at an enrichment ratio above 1.6. When the operators favoured selectivity, the enrichment ratio rose above 2.5 at over 65% recovery.
  • Most concentrate samples from rougher concentrate feed graded above 10% nickel, and some exceeded 14%.
  • On third-cleaner tailings, the cell reached an enrichment ratio above 1.5 at 60% nickel recovery.

The authors compared the cell with the third and fourth mechanical cleaner stages, which have six and four cells. They concluded that the single unit gave the best nickel recovery and enrichment curve. Metso led the paper. The mechanical-cell data came from separate sampling campaigns with the Concorde Cell offline, and the authors note that the feeds were similar but not identical.[1]

Oyu Tolgoi: a magnetic conditioning trial, and a question about sampling

Oyu Tolgoi treats about 120,000 t/d through rougher, cleaner-scavenger and column cleaner stages. Its feed is shifting from open-pit chalcopyrite ore to higher-grade underground ore rich in bornite. The site added flotation capacity for this change even though planned throughput stayed the same. It also trialled magnetic conditioning in the cleaner circuit to lift fine copper recovery.[2]

The paper by Oyu Tolgoi and Ausmetec used a paired t-test. It found that magnetic conditioning raised selective copper recovery and final concentrate grade with high confidence. Most of the gain came from particles finer than 25 µm, and the authors report a short payback but do not give a figure.[2]

The trial also exposed a measurement problem. Each automatic plant sample was split in two. One half went through routine assaying, and the other was sized and assayed as a project sample. Over four months, feed and concentrate assays from the two methods agreed closely. The average cleaner tailings assay differed by 10% between the methods, and that stream drives the recovery calculation. Both methods showed lower tailings grades with magnetic conditioning switched on. The authors found contamination an unlikely cause and left the question open. They also found that the circuit needed one full shift to reach equilibrium after each switch between on and off.[2]

Northparkes: 100 g/t of sodium silicate

Some liberated gold at the Northparkes copper-gold operation in New South Wales, 80% owned by Evolution Mining, was being lost to flotation tailings. The site linked the losses to clay slime coatings, after a rise in kaolinite and chlorite in the mill feed. The paper by Sumitomo Metal Mining and Northparkes reports that gold recovery in rougher tests improved with sodium silicate dosage up to 200 g/t, but copper recovery fell slightly at the highest dosage. The team chose 100 g/t so that copper recovery would stay at or above baseline.[3]

A three-month plant trial, evaluated with paired t-tests and regression analysis, gave these averages:[3]

  • Gold recovery rose from 73.1% to 74.4%.
  • Copper recovery rose from 86.4% to 86.8%.

Northparkes has added sodium silicate permanently to its reagent scheme.[3] The result depends on clay being the cause of the gold losses. Our complete guide to froth flotation covers dispersants, depressants and slime coating.

Coarse particle flotation moves from pilot to plant

Mantos Blancos: NovaCell pilot on rougher tailings

Capstone Copper’s Mantos Blancos concentrator in northern Chile grinds 7.3 Mt/y to a P80 of 250 µm. About 58% of the copper lost to rougher tailings is in particles coarser than 212 µm, and over 70% of those particles have less than 10% surface liberation. Another 15% or so of the losses are in liberated particles finer than 45 µm.[4]

Jord’s NovaCell combines high-shear downcomers for fine particles with a low-shear fluidised bed for coarse particles. A containerised pilot plant designed for about 1.7 dry t/h started on rougher tailings in September 2024. The paper by Capstone, Jord and OrloMET reports up to 56% copper recovery from rougher tailings, an upgrade ratio of 6.1 and a mass pull of 8%. Mineralogy confirmed that the cell recovered coarse composite particles with less than 30% surface exposure. These are the particles that conventional cells lose.[4]

Capstone used the pilot data to build a kinetic model and simulate full-scale units in the plant circuit. The authors state that the results justified detailed design of a full-scale NovaCell installation in rougher-scavenger duty.[4] We found no later public confirmation of a construction decision.

Carrapateena: HydroFloat in operation

BHP, Eriez, VOLPRO and Paradocs Metallurgy announced in the second half of 2025 that they had commissioned a HydroFloat coarse particle flotation plant at Carrapateena in South Australia.[5] Eriez says the plant allows a coarser grind and supports throughput of about 20,000 t/d.[6] BHP’s Rhys Jones presented “Commissioning coarse flotation at BHP Carrapateena” at Mill Operators 2026.[14] BHP and Eriez describe the plant as meeting its targets but have not published recovery or energy figures.[5][6] The Haib copper PEA also designs HydroFloat coarse gangue rejection into its flowsheet.

Metso presented an unnamed coarse particle flotation cell at Flotation ’25. That paper described a pneumatic cell that feeds slurry on top of a deep froth layer and a 25 t/h demonstration plant for a South American copper concentrator. It said the technology would be named at its launch in the first half of 2026.[7] In September 2026, Metso launched the Sampo Cell, which it says feeds slurry directly into the froth phase. Metso reports that a Sampo Cell has run in scavenger duty at a Chilean copper mine since March 2026.[8] Neither document states that the two are the same machine. We covered the Sampo Cell launch in our October news roundup.

Jameson Cells in rougher duty: Fruta del Norte and Caraíba

Jameson Cells were historically favoured in cleaner duty. Glencore Technology’s Flotation ’25 extended abstract reports results from rougher and rougher-scalper installations, which it links to larger 500 mm downcomer designs. The figures are supplier-reported.[9]

  • Fruta del Norte, Ecuador (Lundin Gold): One cell was installed as a rougher scalper at the head of the circuit and another as a scavenger at the end. Gold recovery rose from 87.1% to 90.4% and final concentrate reached 199 g/t gold. Fine particle losses below 20 µm fell by 13.1%. Sulphur recovery rose by 8.2%, which lowered the sulphur grade of the CIL feed and cut cyanide consumption from 54 ppm to 37 ppm.[9]
  • Caraíba, Brazil (Ero Copper): A B6500/24 cell scalps a final copper product. It recovers 67% to 81% of the copper at grades up to 43% copper, against a 36% target. Feed throughput rose by 50%, from 400 t/h to 600 t/h, without expanding the plant footprint.[9]

The abstract also cites Northern Star’s Pogo gold mine in Alaska. There, a rougher scalper was installed on bench-scale data alone because time was short, and flotation throughput rose by more than 60% to 5,000 t/d.[9] See our guide to modern developments in froth flotation for how pneumatic and mechanical cells compare.

Comminution: variable speed at Ban Houayxai

PanAust’s Ban Houayxai gold-silver mine in Laos ran a fixed-speed SAG mill on feed of variable hardness. The site blended soft oxide ore with harder primary rock. In the wet season, limited pit access reduced the hard ore available for blending, and the mill ran on poor feed mixes that wore mill components. In June 2021, a containerised slip energy recovery drive was installed on the mill’s wound-rotor motor. The drive allows speeds from 85% to 105% of synchronous speed. It went in during a scheduled reline without extending the shutdown. The paper says a rental option, paid from operating budgets, let the site make the change without a large capital commitment.[10]

The Comminution ’25 paper by drive supplier Electrodrives and Phu Bia Mining reports:[10]

  • Specific energy fell from 9.4 to 8.8 kWh/t, while throughput held at about 572 dry t/h.
  • Average SAG liner life rose from 4.5 to 5.2 months.
  • Grinding media consumption fell by 16%.
  • Estimated savings exceeded US$1 million a year.

Our grinding mills guide explains how mill speed affects charge motion and breakage.

Ore sorting: rejecting waste before the mill at Tallebung

Sky Metals’ Oliver Davies presented sensor-based sorting at the company’s Tallebung tin project in New South Wales at Mill Operators 2026.[14] The company’s 2026 annual report describes the bulk sample program, first announced to the ASX on 18 August 2025. Near-surface material was crushed to minus 40 mm and screened at 7 mm. The 7 to 40 mm fraction then went through a full-scale TOMRA XRT sorter in two stages. Averaged across four bulk samples:[11]

  • Stage 1 upgraded tin from 0.17% to 2.32% at more than 94.8% recovery and rejected over 93% of the mass.
  • Stage 2 upgraded the product to 10.8% tin at more than 70% recovery.
  • Silver rose from 7.44 g/t to 75.9 g/t in stage 1, at more than 80% recovery.

These recoveries apply to the sorter feed, not to the whole sample. The minus 7 mm fines bypass the sorter, and a large proportion of the tungsten reported to them. The fines are going to dense medium separation tests at ALS Perth and then to pilot gravity testing at ALS Burnie. Sky Metals says the weathered surface material was probably the least favourable zone of the deposit for sorting.[11] The report does not give a whole-flowsheet tin recovery. Our sensor-based ore sorting guide explains how XRT sorters work.

Plant diagnostics: Constancia and Kakula

Two Flotation ’25 papers measured where copper is lost before proposing new equipment.

  • Constancia, Peru (Hudbay): Universidad Técnica Federico Santa María and Hudbay characterised three industrial banks, using diaphragm pumps and full concentrate samples from each cell. They estimated entrainment of 15% to 40% of non-valuable minerals in the first rougher cells. Copper recovery rose linearly with liberation. The larger loss of ultrafine particles in the scavenger bank led the authors to suggest that fine-particle technologies could improve plant performance.[12]
  • Kakula, DRC (Kamoa Copper): A 2021 site survey at a nominal P80 of 50 µm found that 35.3% of the chalcocite was finer than 6 µm, which the authors describe as strong evidence of overgrinding. Site laboratory tests of flash flotation on secondary mill cyclone underflow, at about 60% solids, recovered 21.9% to 33.5% of the copper in a concentrate grading 56.2% to 64.6% copper. Stage milling with stage flotation raised overall laboratory copper recovery by 1.1 to 1.7 points over the baseline.[13] This is testwork, not a plant result. Kamoa-Kakula’s tailings regrind project is covered in our April to August 2026 roundup and in its plant profile.

What the case studies have in common

Small gains, measured carefully. Northparkes gained 1.3 points of gold recovery, and Oyu Tolgoi’s improvement was concentrated in one size fraction.[2][3] Both operators evaluated their trials with paired t-tests, and Northparkes ran its trial for three months.[2][3] Oyu Tolgoi’s sampling discrepancy and its one-shift equilibration time are practical guidance for anyone planning a plant trial.[2]

Losses at both ends of the size range. Mantos Blancos targets coarse composites and ultrafines in the same stream. At Constancia and Kakula, the papers point to ultrafine losses. Kevitsa’s cell targets fine particles and Carrapateena’s targets coarse ones.[1][4][5][12][13] In each case, the operators used size-by-size and liberation data to decide where to act. Our guide to ore mineralogy and liberation covers these measurements.

Upgrades that avoid major construction. The Ban Houayxai drive went in during a scheduled reline, the Caraíba scalper raised throughput without expanding the plant footprint, and the Northparkes change was a reagent addition.[3][9][10]

Read supplier papers with care. Equipment suppliers wrote or co-wrote the Kevitsa, Jameson and Ban Houayxai papers.[1][9][10] Operator co-authors and statistical trials make the data more credible, but the authors still chose which results to publish. Some figures need checking. The Ban Houayxai paper describes the liner-life change as over 25%, but its own figures of 4.5 to 5.2 months give an increase of about 16%.[10] Carrapateena’s commissioning has no published metallurgy so far.[5][6]

IMPC 2026 and what comes next

The XXXII International Mineral Processing Congress met in Cape Town from 18 to 22 October 2026.[15] Its speaker list included Newmont’s Ronel Kappes on “The Frontiers of Flotation” and Vale’s Pedro Porto Silva Cavalcanti on “Reimagining Mineral Processing: From Recovery to Value Creation”. Valterra Platinum’s Agit Singh was listed to speak on South African PGM profitability and the need for a step change. Other keynotes included Imperial College’s Stephen Neethling on modelling and simulation, JKMRC’s Mohsen Yahyaei on process automation and TU Bergakademie Freiberg’s Holger Lieberwirth on grade engineering.[15] IMPC papers were not publicly available when we wrote this article. We will report plant results from the congress once the proceedings are published.

When the Mill Operators 2026 papers are published, we will look for those on fine particle recovery at Cadia, flash flotation in the Ernest Henry regrind circuit, HPGR ramp-up success factors from Orway Mineral Consultants and PGNAA-based ore blending at Kanmantoo.[14] More processing news is in the MillMatters processing news archive.

Sources

Conference papers are cited from the draft or published versions hosted at the URLs given. All links were checked on 1 November 2026.

  1. Mattsson, T., Kela, H., Musuku, B., Kasala, M. and Yáñez, A. “Industrial scale performance of Concorde Cell technology at Kevitsa Cu-Ni-PGM concentrator.” Flotation ’25, MEI, Cape Town, 17–20 November 2025. min-eng.com
  2. Byambasuren, B.-E., Dagva, T. and Lumsden, B. “Cleaner flotation circuit optimization at Oyu Tolgoi: a plant trial of magnetic conditioning.” Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  3. Yamane, M., Curtis, M., Thompson, G. and Garside, M. “Improvement on gold flotation recovery at Northparkes with the addition of sodium silicate.” Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  4. Flores, A., Jaques, E., Amelunxen, P., Akerstrom, B., Orlich, J., Parraguez, D., Saavedra, M., Cooper, L. and Morgan, S. “NovaCell pilot performance at Mantos Blancos: advancing toward full-scale implementation.” Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  5. SME Mining Engineering. “BHP commissions HydroFloat coarse particle flotation plant at Carrapateena Mine.” 2025. me.smenet.org
  6. Eriez. “Carrapateena Mine.” Eriez news. eriez.com
  7. Kupka, N., Emer, Ç., Carvalho, M., Yáñez, A., Concha, V. and Rinne, A. “First look at Metso’s groundbreaking coarse particle flotation technology at industrial scale.” Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  8. Metso. “Metso launches Sampo Cell, a next-generation coarse particle flotation technology to extract more value from every ton of ore.” News release, 10 September 2026. metso.com
  9. Rogers, C., Crane, S., Morin, E., Price, A., Jones, R. and Anderson, C. “Jameson Cells in industrial rougher applications.” Extended abstract, Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  10. Calvo, M., Pascual, S., Romero, J. and Arena, A. “Optimizing comminution at Ban Houayxai mine: upgrading to variable speed in a fixed-speed SAG mill.” Comminution ’25, MEI, Cape Town, 31 March–3 April 2025. min-eng.com
  11. Sky Metals Limited. Annual Report for the financial year ended 30 June 2026. ASX release, 17 September 2026. asx.com.au
  12. Vallejos, P., Yianatos, J., Medina, M. and Sotomayor, G. “Comprehensive assessment of flotation banks through mineralogy, liberation, and particle size at Constancia concentrator.” Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  13. Mulanga, J., Bongo, J., Shungu, G. and He, S. “Optimization of copper recovery at the Kakula Concentrator, DRC.” Extended abstract, Flotation ’25, MEI, Cape Town, November 2025. min-eng.com
  14. AusIMM. Mill Operators Conference 2026 program, Brisbane, 19–21 October 2026. ausimm.eventsair.com
  15. IMPC 2026. XXXII International Mineral Processing Congress, Cape Town, 18–22 October 2026: plenary and keynote speakers. impc2026.com
  16. MEI. Flotation ’25 technical programme and draft papers, Cape Town, 17–20 November 2025. min-eng.com
  17. MEI. Comminution ’25 technical programme and draft papers, Cape Town, 31 March–3 April 2025. min-eng.com

Explore the global processing-circuit library

Browse all mineral processing jobs

Related Posts

Scroll to Top