MINTEK 75 TECHNICAL CONFERENCE

“Pyrometallurgy – It Seemed Like a Good Idea at the Time”

by

Lloyd NelsonHead of Smelting and Refining Technology

Anglo Platinum

4 June 2009

Theme

• Through a suite of pyrometallurgical projects, with which Mintek has been involved historically, aim to illustrate the benefits & criticality of:

- Appropriate up-front process & technology identification- Sound understanding of pyrometallurgical engineering principles- Role of Process, in dictating equipment design, versus alternatively

“force-fitting a technology/design to a process”- Rigorous piloting- Continued ramp-up, optimisation & development in the industrial

operating environment

to

… SUCCESSFULLY realise a FULLY COMMERCIAL Technology! …

FeSiCr Reduction of Lime/Ore Meltto Produce LC FeCr

• Commercial Perrin Process:- Ladle “cocktailing”- Ar(g)-stirred

• Process dependencies:- Liquid phase mass transfer- Adequate slag superheat- Adequate low lime-ore melt

liquidus temperature (1000C)

• Mintek involvement includes:- Novel “Hi-Lo” Chrome Process- Middelburg Technochrome- Zimalloys

Slag AtlasSlag Atlas

• Opportunity:- Pre-reduce Lime-ore melt- Save on electrical energy/cost to

produce Si (in FeSiCr)

• Process dependencies:- Reduce Cr ore to lower CrO

(Cr2+)- High lime-ore melt liquidus

temperature (>1600C)- Inadequate slag superheat

• Mintek involvement includes:- Fastest 1-day project – late ’80s- Records showed idea already

rejected in ’70s!

Slag AtlasSlag Atlas

Emulate Commercial LC/MC FeMnPre-reduce Lime/Ore Melt?

Slag AtlasSlag Atlas

“Read Across” SiO2-CrO/Cr2O3 System?

• Cr2O3 now raises solidus/ liquidus temperature

• Silica/”CrO” now lowers liquidus temperature to yield “fluxed reduction”

Diversity of Chromite Smelting Options

• Classical Benchmark - HC FeCr (> 65% Cr):

• Conventional Lump - ChCr (50% Cr) only with advent of AOD Process:

• Abundant Chromite Fines:

• Yet, SA = World’s largest installed smelting base for ChCr!

• Mintek involvement:- World Class!- All Aspects, Fundamental Industrial

Product Pre-treatment Technology Flux B Cr Rec. SECSource Form % MWh/t alloy

ChCr Bushveld Lump SAF Open Basic 1.3 82 3.8Lump SAF Open/Closed Acid 0.7 80 3.6

Ore(CaO+MgO)/SiO2 w/w%

Product Pre-treatment Technology Flux B Cr Rec. SECSource Form % MWh/t alloy

ChCr Bushveld Fines (Lump) Open SAF Acid 0.8 70 > 4Fines Pellet Sinter/Preheat Outotec+SAF Acid <1 85Fines Pellet Pre-Red/Preheat SDK/Premus+SAF Basic 1.3 88-90 < 2.5Fines Acid flux Pre-Red/Preheat (CDR)/DC-arc Basic 1.5 90 < 2.0Fines DC-arc Patent Acid 0.8 90 4Fines DC-arc Basic 1.6 90 4.4-5

Ore(CaO+MgO)/SiO2 w/w%

Product Pre-treatment Technology Flux B Cr Rec. SECSource Form % MWh/t alloy

HC FeCr Non-SA Lump (Sinter if Bushveld Fines) SAF Open/Closed Basic 1.3 92-95 3.7

Ore(CaO+MgO)/SiO2 w/w%

DC Plasma-arc Chromite Smelting

• Feed Characteristics:- Direct fines treatment – ore, flux &

reductant

• Slag Metallurgy:

- Prone to lining refractory failure- Adopt “sidewall bank protection”

• Mintek involvement includes:- World First!- Novel Patent with MS&A- 16 40 MVA Upgrade Furnace

PalmietPalmiet FeCrFeCr

Totals 134.7%?

Patent Commercial

“Acid” “Basic”

0.8 B 1.6

> 1600°C Liquidus > 1700°C

Irony/s of the DC-arc Furnace?

• Patent:- “Acid” not operated Industrially!- “35% CaO” inadvertently

afforded 20 years Protection?

• Did the Patent Really Advantage Samancor?

- 13 years to commission M3- 20 years to “Freeboard plate

coolers” optimisation- 24 years (after Chambishi) from

“hollow electrode, centre-feed”to “side-feeding”: Realising SEC savings Despite Mintek piloting in ’80s!

- Fully 26 years to M4 Furnace!

• DC-arc ilmenite smelting commercialisation (Mintekpatent) = 10-15 years?

• SA Patent on Mintek Website

Another Another SiteSite

Middelburg Middelburg FeCrFeCr (M3) (M3) Freeboard Freeboard

Plate CoolersPlate Coolers

DC-arc Furnace FeCo Production

• “Complex” Process Requirements:- Difficult alloy to tap & then atomise (Alloy liquidus 1370C)– Slag/refractory issues:

• highly superheated (T = 400°C)• corrosive siliceous slag ( 50% SiO2)

– Classical metal-slag interface “tidal zone” corrosion

• Mintek involvement included:- Mintek Patent- Piloting- Chambishi (Zambia) 40 MW furnace - Avmin

• Commissioning & Ramp-up:- Despite extensive piloting, 3 commercial campaign failures in first

18 months!

Hi-Intensity Bath Coolers

• Sidewall bath containment:- Copper Waffle coolers:

• Slag = 100 kW/m2

• Alloy = 500 kW/m2

(FeCr/FeMn 10-20kW/m2)- “Freeze Lining” essential!

• Ramp-up:- Missed M3 “learning” – no

upper sidewall plate coolers, leading to refractory failure!

- Enforced side feeding to form “protective sidewall banks”

- World First “DC Immersed-arc”

- 4 year campaign coincided with Co commodity boom!

Missing sidewall refractory above

waffle cooler

ChambishiChambishi

ChambishiChambishi

PGM Pyrometallurgy– Classical Merensky Smelting

• Process Considerations:- Flux silica to lower viscosity

& slag liquidus temperature using:Lime/Limestone additionFayalite converter slag

return- Tslag = 1300-1400C- Tmatte = 1200-1250C

• Design Considerations:- Basic magnesia-lined, low

intensity smelting furnace- Commensurate with classical

copper & nickel smelting!

TL, Add MgO

TL, add CaO

Eric & Eric & HejjaHejja19941994

SiO2 Cr2O3 MgO FeOMerensky 41 0.4 18 23

PGM Concentrate

PGM Pyrometallurgy– UG2 Smelting Issues

• Classical “Chrome” Problem:- Furnace spinel accumulation &

“buildup”- Lime flux adversely stabilises

chromite spinel, raising slag:Liquidus temperatures >

1700C“Effective” viscosity

• Terminate converter slag return – need Cr2O3 “bleed”

• JCI & Mintek patents with reductant addition:

- Matte + “Alloy” PGM collection- Increase slag CrO solubility in acid

slag (“fluxed reduction”)

SiO2 Cr2O3 MgO FeOMerensky 41 0.4 18 23UG2 47 2.8 21 15

PGM Concentrate Composition, mass %

PGM Pyrometallurgy– UG2 Smelting & Hi-Intensity Furnaces

• Design Considerations:- High-intensity water-

cooled copper in the furnace linings

• Industry Impact:- From Mortimer = 18MW

toWaterval/Impala = 34MW

toPolokwane = 68MW

- “Doubled” … even “Re-doubled” smelting throughput in a decade!PolokwanePolokwane

slag tappingslag tapping

at 1659at 1659°°CC

PGM Pyrometallurgy– Hi-Intensity Furnaces & Corrosion

• Corrosion Issues:- Both Copper & Magnesia-based

refractories corrode

• Corrosion Mechanism:- “Chloride-accelerated Sulfidation”- Contributions even at:

< 2% labile S< 0.1% Cl40C copper tip temperatures

- At best marginal acid dew-point corrosion contribution

Yellow Yellow Elemental Elemental

sulphursulphur

Minor green Minor green CuCu22Cl(OH)Cl(OH)33

Abundant Abundant black black CuSCuSxx

Copper Copper waffle knobwaffle knob

Waffle coolerWaffle cooler Magnesium Magnesium hydroxyhydroxy--chlorideschlorides

Concluding Remarks

• Despite some obvious exceptions, the evidence for increasing process smelter intensification is compelling:

- Double productivity with “same” furnaces on FeNi smelters > 80MW- Ilmenite smelters > 70MW furnaces- PGM smelting industry at least doubled – 34-68MW furnaces- > 1.2 Mt/a copper concentrate single vessel smelters/converters

• Without exception ALL use water-cooled copper technology

• Can any meaningful local advances in pyrometallurgy hope to progress in isolation of this? Potential opportunities:

- Acid practice, DC-arc furnace chromite smelting (old Mintek patent)?- PGM ConRoast to render it more robust (“evolving”, Mintek patent)?

• The Challenge for Mintek … Industry =Courage & Vision to lead technology development, to ensure:“Pyrometallurgy Indeed Remains a Good Idea at the Time!”