The implications of ore hardness variability on comminution circuit energy efficiency (and some other thoughts)

Transcription

1 The implications of ore hardness variability on comminution circuit energy efficiency (and some other thoughts) Peter Amelunxen Thursday 6 December 2012

2 Capital (Millions) Current Cu/Mo Mill Design (a green perspective) HPGR Selected Mill Projects Capital SAG NPV optimization subject to constraints Scarcity of capital, water, human resources, land, energy, an others It s a competition It s not about reducing the carbon footprint, it s about cost and risk optimization Social and ecologic sustainability are evaluated in this context. Project Economics and Project Risk define the limits of corporate citizenship. This is where is geometallurgy is useful

3 Topics What is being done with geomet? Some examples What could be done with geomet? Some examples

4 What is being done with geomet? Example 1 HPGR Trade-off Study

5 HPGR Tradeoff Studies Usually grinding circuits are sized for a homogenous ore Constant A*b, SPI, Wi, etc., Usually evaluated for a fixed tonnage No variability is incorporated Problems: SAG mill circuits are directly coupled to ball mill circuits, so changing ore properties (feed size, hardness, etc.) lead to changing process bottlenecks (SAG-limited vs. ball mill-limited) For any given unit operation, an upstream or downstream bottleneck creates a loss of efficiency Sizing a SAG-based circuit for a single ore hardness ignores this important concept.

6 Trade-off NPV (thousands) Cumulative Distribution Not Another HPGR Tradeoff Study! Global Distribution of Bond Ball Mill Work Index Hard Ore, BWI = 18.5 kwh/tonne Medium Ore, BWI = 14.4 kwh/tonne Soft Ore, BWI = 11.8 kwh/tonne $150,000 Work Index (Metric) Item Cost $ Units Crusher Liners $2,900 $/tonne HPGR Rolls $1,700 K$ / set SAG Liners $3,500 $/tonne SAG balls $1,000 $/tonne BM Liners $3,500 $/tonne BM balls $1,250 $/tonne Energy $0.09 $/kwh $200,000 $100,000 $50,000 $0 ($50,000) ($100,000) Hardness (SPI, min) SABC HPGR Crush kwh/t kwh/t kwh/t Hard Medium Soft HPGR HPGR Cone Crusher Depends Amelunxen, P., & Meadows, D. (2011). Not another HPGR trade-off study! Minerals & Metallurgical Processing, 28(1), 1-7. ($150,000) ($200,000) SAG

7 Variability: Effects on HPGR Economics Simple Trade-off study 96K TPD Perfectly homogenous ore body Equipment sized for average hardness Incorporate revenue stream using annual mine plans Yearly SPI & Bond Wi P80 vs. Recovery Incorporate revenue stream using daily hardness variability Simulated values

8 Case 1 Perfectly Homogenous Both circuits sized for the median ore hardness SPI = 104 minutes, BWI = 14.0 kwh/t $158 Million additional CAPEX for HPGR circuit $ 0.49/t lower operating cost SAG circuit wins by $33 million NPV

9 Trade-off NPV (thousands) Comparison $200,000 $150,000 $100,000 Case 1 (median ore hardness) HPGR $50,000 $0 ($50,000) Hardness (SPI, min) Cone Crusher ($100,000) ($150,000) ($200,000)

10 SPI (min) Bond Work Index (kwh/mt) Case 2: Annual Variability Yearly Throughput Assumptions T80 limit = 5mm P80 limit = 225 microns Upstream limit = 125K TPD Downstream limit = 1.15*Nominal Ramp up limit (9 months) SPI & BWI vs Year SPI 11.0 Bond Work Index Year No.

11 TPH P80 (microns) TPH P80 (micrones) Economics Including Mine Plan 140,000 TPH & P80 - SABC-A 96K TPD 40' x 26' inside shell x EGL SAG Mill, 26' x 40' Ball Mills , ,000 98,220 99, , , ,493 97, , Higher tonnage most years 80,000 83,214 87,500 92,425 92,425 92,067 91,323 88, , TPH & P80 - HPGR 96K TPD 3 x 2.4m x 1.65m HPGRs, 2 x 26' x 41' Ball Mills , ,000 Plant Capacity (TPD) Plant P80 (microns) Year Small difference in recovery Due to coarser P Throughput Capacity (TPD) 160 Cyclone Overflow P80 (microns) Year HPGR circuit wins by $50 million NPV

12 Trade-off NPV (thousands) Comparison $200,000 $150,000 $100,000 Case 2 (Annual Variability) HPGR $50,000 $0 ($50,000) Hardness (SPI, min) Cone Crusher ($100,000) ($150,000) ($200,000)

13 Bond Wi SPI Normalized Gamma Case 3: Daily Variability SPI & Wi - First 5 Years SPI Work Index Day No Time Semi-Variagram (Normalized) Cerro Verde Mill Feed Bond Wi y = ln(x) R² = Days Mill performance simulated for each day and results summed over the entire mine plan HPGR circuit wins by $84 million NPV

14 Trade-off NPV (thousands) Comparison $200,000 Case 3 (Daily Variability) $150,000 $100,000 HPGR $50,000 $0 ($50,000) Hardness (SPI, min) Cone Crusher ($100,000) ($150,000) ($200,000)

15 Conclusions Case 1 Variability should be an integral part of any HPGR/SAG tradeoff study Particularly those near the equilibrium ore hardness point Study highlights the importance of geometallurgical profiling, mine plan, and production forecasts early in the development cycle

16 What is being done with geomet? Example 2 Feasiblity Level engineering study for a Chilean SAG mill concentrator (recently constructed)

17 Design & Engineering Performed by a well-known international engineering firm Grinding circuit designed for the hardest of several ore types About SPI tests performed but were not used. JK DWT on composites representing each ore type Some McPherson tests, some pilot plant tests Scale-up was performed principally from pilot plant

18 Result? Plant started up and reached significantly less than design capacity Ore was significantly harder than the composite samples SAG mill was the bottleneck The good news? Recovery was a bit higher because of the higher retention time and finer grind Expansion project implemented to increase the tonnage to design levels Total Cost: $1 billion in losses ($3.00 copper, approximate) $850 million lost revenue $150 million aprox. for the expansion $150 thousand for the SPI tests that weren t used

19 The problem? Risk is fuzzy, costs are not What is risk? Is SPI risky? Is JK DWT or JK SimMet risky? Is a pilot plant risky? Is geology or metallurgy risky? What s a fatal flaw? Torremolinos diamond mine? Escondida moly plant?

20 Summary of Current Practice Good design approach Cost of geometallurgy: $100K 300K Benefit: $117 million Challenged design approach Cost of geometallurgy: 0, (not including unused SPI tests) Losses: approx. $1.0 billion (we ll never know for sure) Question: Why is the downside (in these examples) so much greater than the upside?

21 US$, thousands What else could we do? Grind size vs. recovery optimization What if higher head grade ore is softer? What if harder ore is deeper? NPV vs Grind 2,060,000 2,040,000 2,020,000 2,000,000 1,980,000 1,960,000 1,940,000 1,920, P80 (microns)

22 Yield Stress (Pa) What else could we do? Tailings Impoundment and Water Balance 600 HD/P Thickener Design P80 = 220 um 100 P80 = 180 um P80 = 140 um Solids Concentration (% m/m) Mill Underflow density affects evaporation losses For HDTT, % solids, % fines, clay content affect deposition angle Seepage losses are affected clays, % fines

23 US$, thousands Grind Recovery Tailings Optimization This was from cost-benefit around grinding & flotation, using geometallurgical methods 2,060,000 NPV vs Grind 2,040,000 2,020,000 2,000,000 1,980,000 1,960,000 1,940,000 1,920, P80 (microns) Incorporate the cost of tailings and water handling in the NPV calculation, using geometallurgical methods

24 What else can we do? Mill Percent of Opex Percent of Capex Desalination Plant Desalination Pumping Everything Else Desalination & Pumping OC Power Plant Everything Else Grind-recovery-tailings-water-power optimization?

25 Components Block model with distributed geometallurgical parameters Mine plan Integrated process models (validated) Capital cost models Operating cost models Fast computers

26 Summary (my opinions) The positive economic impacts of geometallurgy are significantly understated in our field The scope of geometallurgical programs are driven by risk avoidance rather than economics Adoption rate for geometallurgical methods is increasing

27 Capital (Millions) Current Cu/Mo Mill Design (a green perspective) HPGR Selected Mill Projects Capital SAG NPV optimization subject to constraints Scarcity of capital, water, human resources, land, energy, an others It s a competition It s not about reducing the carbon footprint, it s about cost and risk optimization Social and ecologic sustainability are evaluated in this context. Project Economics and Project Risk define the limits of corporate citizenship. Geometallurgy is part of good corporate citizenship

28 Thanks Co-authors M.A. Mular, J. Vanderbeek, L. Hill, and E. Herrera (Freeport-McMoRan Copper & Gold) D. Meadows (FLSmidth) Organizing Committees for Geomet, Gecamine CEEC (Coalition for Eco-Efficient Comminution) You the audience, for your patience