The implications of ore hardness variability on comminution circuit energy efficiency

(and some other thoughts)

Peter Amelunxen Thursday 6 December 2012

Current Cu/Mo Mill Design (a green perspective)

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

0

1000

2000

3000

4000

5000

6000

7000

8000

Cap

ital

(M

illio

ns)

Selected Mill Projects Capital

HPGR SAG

Topics

What is being done with geomet?

Some examples

What could be done with geomet?

Some examples

What is being done with geomet?

Example 1 – HPGR Trade-off Study

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.

Not Another HPGR Tradeoff Study!

0

10

20

30

40

50

60

70

80

90

100

0 5 10 15 20 25 30

Cu

mu

lati

ve D

istr

ibu

tio

n

Work Index (Metric)

Global Distribution of Bond Ball Mill Work Index

Soft Ore, BWI = 11.8 kWh/tonne

Medium Ore, BWI = 14.4 kWh/tonne

Hard Ore, BWI = 18.5 kWh/tonne

($200,000)

($150,000)

($100,000)

($50,000)

$0

$50,000

$100,000

$150,000

$200,000

0 20 40 60 80 100 120 140 160 180 200

Tra

de-o

ff N

PV

(th

ou

san

ds)

Hardness (SPI, min)

HPGR

Cone Crusher

Amelunxen, P., & Meadows, D. (2011). “Not another HPGR trade-off study!” Minerals & Metallurgical Processing, 28(1), 1-7.

HPGR

SAG

Depends

SABC HPGR Crush

kWh/t kWh/t kWh/t

Hard 19.2 16.5 17.0

Medium 14.4 13.5 13.6

Soft 10.9 11.4 11.7

$ 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

Item Cost

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

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

Comparison

($200,000)

($150,000)

($100,000)

($50,000)

$0

$50,000

$100,000

$150,000

$200,000

0 20 40 60 80 100 120 140 160 180 200

Tra

de-o

ff N

PV

(th

ou

san

ds)

Hardness (SPI, min)

HPGR

Cone Crusher

Case 1 (median ore hardness)

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)

90

106

99

133

101

122

87

97 98

109

99

115

13.74

14.78

13.81

14.97

13.7713.99

13.36 13.3813.68

14.33

13.86

14.61

10.0

11.0

12.0

13.0

14.0

15.0

16.0

17.0

18.0

19.0

20.0

0

20

40

60

80

100

120

140

160

180

200

0 1 2 3 4 5 6 7 8 9 10 11 12 13

Bo

nd

Wo

rk In

dex

(kW

h/m

t)

SPI (

min

)

Year No.

SPI & BWI vs Year

SPI

Bond Work Index

83,214

87,500

92,425 92,425

98,220

92,067

99,320

110,400 109,602

105,493

91,32388,858

97,709

109,601

206

196

201 201

207

215

208

215 215212

203207

212

218

140

160

180

200

220

240

260

280

300

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2014 2016 2018 2020 2022 2024 2026 2028 2030

P8

0 (

mic

ron

es)

TPH

Year

TPH & P80 - SABC-A 96K TPD40' x 26' inside shell x EGL SAG Mill, 26' x 40' Ball Mills

Plant Capacity (TPD)

Plant P80 (microns)

Economics Including Mine Plan

140

160

180

200

220

240

260

280

300

0

20000

40000

60000

80000

100000

120000

140000

2014 2016 2018 2020 2022 2024 2026 2028 2030

P8

0 (

mic

ron

s)

TPH

Year

TPH & P80 - HPGR 96K TPD3 x 2.4m x 1.65m HPGRs, 2 x 26' x 41' Ball Mills

Throughput Capacity (TPD)

Cyclone Overflow P80 (microns)

Higher tonnage most years

Small difference in recovery Due to coarser P80

…HPGR circuit wins by $50 million NPV

Comparison

($200,000)

($150,000)

($100,000)

($50,000)

$0

$50,000

$100,000

$150,000

$200,000

0 20 40 60 80 100 120 140 160 180 200

Tra

de-o

ff N

PV

(th

ou

san

ds)

Hardness (SPI, min)

HPGR

Cone Crusher

Case 2 (Annual Variability)

Case 3: Daily Variability

0

20

40

60

80

100

120

140

160

180

12

13

14

15

16

17

0 500 1000 1500

SPI

Bo

nd

Wi

Day No.

SPI & Wi - First 5 Years

SPI

Work Index

y = 0.1232ln(x) + 0.2886R² = 0.9104

.0

.1

.2

.3

.4

.5

.6

.7

.8

.9

0 10 20 30 40 50 60 70

No

rmal

ize

d G

amm

a

Days

Time Semi-Variagram (Normalized)Cerro Verde Mill Feed Bond Wi

Mill performance simulated for each day and results summed over the entire mine plan

…HPGR circuit wins by $84 million NPV

Comparison

($200,000)

($150,000)

($100,000)

($50,000)

$0

$50,000

$100,000

$150,000

$200,000

0 20 40 60 80 100 120 140 160 180 200

Tra

de-o

ff N

PV

(th

ou

san

ds)

Hardness (SPI, min)

HPGR

Cone Crusher

Case 3 (Daily Variability)

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

What is being done with geomet?

Example 2 – Feasiblity Level engineering

study for a Chilean SAG mill concentrator

(recently constructed)

Design & Engineering

Performed by a well-known international

engineering firm

Grinding circuit designed for the hardest of

several ore types

About 100 – 150 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

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

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?

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?

What else could we do?

Grind size vs. recovery optimization

What if higher head grade ore is softer?

What if harder ore is deeper?

1,920,000

1,940,000

1,960,000

1,980,000

2,000,000

2,020,000

2,040,000

2,060,000

100 120 140 160 180 200 220 240

P80 (microns)

NPV vs Grind

US$, th

ousands

What else could we do?

Tailings Impoundment and Water Balance

0

100

200

300

400

500

600

50 55 60 65 70 75 80

Yie

ld S

tres

s (P

a)

Solids Concentration (% m/m)

HD/P Thickener Design

P80 = 220 um

P80 = 180 um

P80 = 140 um

Mill

For HDTT, % solids, % fines, clay content affect deposition angle

Seepage losses are affected clays, % fines

Underflow density affects evaporation losses

Grind – Recovery – Tailings Optimization

1,920,000

1,940,000

1,960,000

1,980,000

2,000,000

2,020,000

2,040,000

2,060,000

100 120 140 160 180 200 220 240

P80 (microns)

NPV vs Grind

US$, th

ousands

Incorporate the cost of tailings and water handling in the NPV calculation, using geometallurgical methods

This was from cost-benefit around grinding & flotation, using geometallurgical methods

What else can we do?

Mill

Desalination Plant

Desalination

Pumping

Everything Else

Percent of Opex

Desalination & Pumping

Everything Else

Percent of Capex

OC Power Plant

Grind-recovery-tailings-water-power optimization?

Components

Block model with distributed

geometallurgical parameters

Mine plan

Integrated process models (validated)

Capital cost models

Operating cost models

Fast computers

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

Current Cu/Mo Mill Design (a green perspective)

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

0

1000

2000

3000

4000

5000

6000

7000

8000

Cap

ital

(M

illio

ns)

Selected Mill Projects Capital

HPGR SAG

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