the implications of ore hardness variability on ......tph & p80 - sabc -a 96k tpd 40' x...
TRANSCRIPT
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