014 Alges Sampling 2011 Cmcc v2 Comp

Embed Size (px)

Citation preview

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    1/81

    Advanced reverse circulation drilling as a replacement

    to blast hole sampling: increasing short term planning

    profitability at Cerro Colorado copper mine

    Eduardo J. Magri, Julin M. Ortiz (Universidad de Chile)

    Rodrigo Moya, Andrs Salazar (Compaa Minera Cerro Colorado)

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    2/81

    Outline

    Introduction Cerro Colorado Operation

    Cerro Colorado Geology

    Methodology

    Results Conclusions

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    3/81

    Introduction

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    4/81

    Motivation

    Cerro Colorado is a producing open pit mine (~55 ktpd of ore)

    This presentation is an attempt to evaluate the economic impact ofreplacing blast hole sampling by advanced RC drilling for shortterm planning, this includes:

    Impact of reducing the sampling errors (precision and bias)

    Definition of the spacing of the advanced drilling grid

    Impact of the grade control methodology and estimation parameters

    Impact of the misclassification of geological units (clays)

    The approach is based on building conditional simulations ofgrade distributions, simulating the sampling and grade controlprocedures, and evaluating the resulting profit.

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    5/81

    Cerro Colorado Operation

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    6/81

    Cerro Colorado is a porphyry copper deposit located in the Tarapaca Region of Chile

    (lat. 2002'S, long. 6915'W), 120 km East of the city of Iquique, at an elevation of2600 m

    It belongs to a belt of porphyry copper deposits in the Central Andes

    The operation was started by Ro Algom Ltd. in 1994.

    Cerro Colorado is an operation of BHP Billiton Pampa Norte, a business unit that alsoincludes Minera Spence located in the Antofagasta Region.

    Mining is done by open pit, extracting oxides and sulphides

    Background information

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    7/81

    Mining includes oxides and secondary enrichment Ore throughput: 52-55 Ktpd & 19.2 Mtpa

    Rock extraction / Stripping ratio: 90 Mtpa / 1: 3.7

    Ore Process: Heap leaching and SX-EW Ore type: oxide, secondary sulphide and MSH (mixed of secondary sulphide

    & hypogene copper). Ore mineralogy: chrysocolla, brochantite, chalcocite and covellite.

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    8/81

    Metallurgical recovery: 69% Geometallurgical short term planning model considers blending of different clays

    Production FY11: 90.5 Kt of Cu Cathodes

    Plant capacity (design): 130 Ktpy of Cu Cathodes

    Reserves (Probable + Proven): 167.9 Mt @ 0.65% TCu & 0.36% SCu. (c.o.g: 0.3%TCu)

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    9/81

    Background information

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    10/81

    Cerro Colorado Geology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    11/81

    Cerro Colorado 53 Ma

    MetallogenicsBelts

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    12/81

    Regional Geology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    13/81

    Cerro Colorado Geology

    Cerro Colorado is a porphyry copper deposit located in the north-south metallogenic belt, dated from the paleocene-inferior eocene.

    The regional geology consists of a andesitic volcanoclasticsequence, from the Cretaceous (Cerro Empexa formation), intrudedby tonalitic and quartz-monzonitic fases from the lower tertiary(porphyries and breccias). These units are covered by volcanic andsedimentary deposits from the Pliocene (ignimbrites and gravel fromthe Altos de Pica formation).

    The copper mineralization is related to intrusive events through a

    trend NE to EW, in a series of sub horizontal bodies (West sector)and more complex units (East sector) of copper oxides, supergenecopper sulphides and primary copper sulphides in depth. Theextension is approximately 2.3 km in length in the EW direction and1.5 km in the NS direction.

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    14/81

    Cerro Colorado Geology

    MineralizationAltos de Pica

    60m av-

    Lch: 25-75m thick.

    CuS/CuT > 0.5 25-125 m thick.

    Cu leached >=80 % 10-100 m thick.

    Cu leached: 50 79 %

    Cu leached< 50 %

    (Br-Cr)

    (Cs-Cv)

    (Cs-Cv-Cpy)

    (Cpy>>Cs-Cv)

    Gravel - Ignimbrites

    Leach cap

    Cu Oxides

    Supergene sulphides

    Mixt sulphides

    Primary sulphides

    Waste

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    15/81

    Cerro Colorado Geology

    Lithological units

    Porphyry

    Breccia

    Andesite

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    16/81

    Cerro Colorado Geology

    Lithological units

    Porphyry

    Breccia

    Andesite

    Gravel

    Ignimbrite

    CerroColorado

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    17/81

    Cerro Colorado Geology

    Mineralization units

    Cerro ColoradoSulphidesOxides

    Hypogene MSH

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    18/81

    Cerro Colorado Geology

    Alteration units

    Cerro Colorado

    Potassic Alt.North sector

    PhyllicAlt.

    Potassic Alt.Phase 5 low

    grade

    Topography Phase 7Extreme

    north sector

    Phyllic chloriticAlt.In deeper levels

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    19/81

    Cerro Colorado Geology

    SE view of current pit Blue line A-A: Orientation of sections

    A

    A

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    20/81

    Cerro Colorado Geology

    Ore type units

    2000 msnm

    2600 msnm

    CURRENTPIT

    CURRENT

    PIT

    HYPOGENE

    OXIDES

    LEACHED

    CAP

    SECONDARYSULPHIDES

    TRANSITIONALSULPHIDES

    GRAVEL

    A A

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    21/81

    Cerro Colorado Geology

    Lithological units

    2600 msnm

    2000 msnm

    CURRENT

    PITCURRENTPIT

    PORPHYRY

    IGNEOUSBRECCIA

    IGNEOUS

    BRECCIA

    PORPHYRYIGNIMBRITE

    IGNIMBRIT

    EGRAVEL

    A A

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    22/81

    Cerro Colorado Geology

    Clay units (ARC)

    CURRENT PIT

    CURRENT PIT

    2000 msnm

    2600 msnm

    ARC9

    ARC1

    ARC7

    ARC7

    ARC7

    ARC4

    ARC

    2

    ARC1

    ARC

    6

    ARC8

    ARC8

    ARC

    1

    ARC7

    ARC8

    ARC1

    ARC Description

    1-2-3-9 Detrimental clays

    4-5 Clays with moderate effect

    6-7-8 Clays that do not affect the process

    A A

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    23/81

    Cerro Colorado Geology

    Alteration units

    CURRENT PIT

    CURRENT PIT

    2600 msnm

    2000 msnm

    PHYLLIC

    PHYLLIC

    PHYLLIC

    POTASSIC

    POTASSIC

    CHLORITIC

    ARGILLIC

    GRAVELS

    CHLORITIC

    A A

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    24/81

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    25/81

    Goals

    Evaluate the effect of: Information quality:

    blast hole sampling error (current) vs advanced RC drilling and sampling(proposed)

    Effect of systematic bias

    Effect of improving the accuracy of geologcial interpretation

    Information quantity:

    Blast hole spacing (current) vs different advanced drilling grid spacings(proposed)

    Estimation method:

    Inverse distance squared weighting (current) vs Ordinary Kriging (proposed)

    Estimation parameters

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    26/81

    Methodology

    Inputs: Long term planning model UGCUT, Density, ARC (clays)

    Drill hole 10.0 m composite database Total copper (CuT), soluble copper (CuS) Solubility ratio (RSol)

    Duplicate blast hole data

    Total copper (CuT), soluble copper (CuS) Economic and operational parameters (costs, commodity price, W/O ratio,

    recovery functions, acid consumption, etc.)

    Considerations: Long term geological attributes are fixed (UGCUT, Density, ARC)

    Drill hole data are used to infer statistical distribution of the variables and

    variograms to be used in simulation Sampling errors for CuT, CuS are inferred from duplicate blast hole data

    Output: Simulated realizations at point support of CuT, RSol CuS (dense

    spacing)

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    27/81

    Methodology

    Definition of the study volume East 71700-73900 2200 m

    North 82900-84100 1200 m

    Elevation 2270-2530 260 m

    Geological units and clay units:

    UGCUT Description

    2 Oxides

    3 Oxides

    4 Sulphides

    5 Sulphides

    ARC Description

    1-2-3-9 Detrimental clays

    4-5 Clays with moderate effect

    6-7-8 Clays that do not affect the process

    Long Term

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    28/81

    MethodologyLong Term

    Model

    Drill holedatabase

    Blast holedatabase

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    29/81

    Methodology

    Sampling errors: Calculated from duplicated blast hole data

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    30/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    31/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    32/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    33/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    34/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    35/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    36/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    37/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    38/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    39/81

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    40/81

    Duplicate blast hole data - CuS

    Error = 24.3%

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    41/81

    Duplicate blast hole data CuS > 0.1%

    Error = 15.0%

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    42/81

    Duplicate blast hole data - CuT

    Error = 19.9%

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    43/81

    Duplicate blast hole data CuT > 0.1%

    Error = 14.0%

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    44/81

    Duplicate data by range

    Sampling errors by range of values

    Range of grades

    Mean Relative Error

    (%)

    CuS CuT

    Average of duplicates > = 0.1% 15.02 13.97Average of duplicates > = 0.0 y < 0.1% 29.04 31.39

    Average of duplicates > = 0.1 y < 0.2% 26.71 21.53

    Average of duplicates > = 0.2 y < 0.3% 23.74 20.68

    Average of duplicates > = 0.3 y < 0.5% 21.53 16.96

    Average of duplicates > = 0.5 y < 0.8% 18.23 17.90Average of duplicates > = 0.8 y < 1.2% 22.43 18.54

    All 24.34 19.94

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    45/81

    Methodology

    1. Ten dense simulation models were builtCoordenate Number

    of points

    Size (m) Min Coord

    (m)

    Max Coord

    (m)

    Extension

    (m)

    X (East) 1100 2.0 71700.0 73900.0 2200.0

    Y (North) 600 2.0 82900.0 84100.0 1200.0

    Z (Elevation) 26 10.0 2270.0 2530.0 260.0

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    46/81

    Methodology

    2. Block average represent the truthCoordenate Number

    of points

    Size (m) Min Coord

    (m)

    Max Coord

    (m)

    Extension

    (m)

    X (East) 1100 2.0 71700.0 73900.0 2200.0

    Y (North) 600 2.0 82900.0 84100.0 1200.0

    Z (Elevation) 26 10.0 2270.0 2530.0 260.0

    Coordenate Number of

    blocks

    Size (m) Min Coord

    (m)

    Max Coord

    (m)

    Extension

    (m)

    X (East) 220 10.0 71700.0 73900.0 2200.0Y (North) 120 10.0 82900.0 84100.0 1200.0

    Z (Elevation) 26 10.0 2270.0 2530.0 260.0

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    47/81

    Methodology

    Correlations for pairs of variables

    CuT vs CuS

    = 0.4 0.9

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    48/81

    Methodology

    Correlations for pairs of variables

    CuT vs RSol

    = -0.3 0.3

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    49/81

    Methodology

    Correlations for pairs of variables

    CuS vs RSol

    = 0.4 0.5

    The variableswith lowercorrelation are

    CuT / RSol

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    50/81

    Methodology

    Implementation details: Point simulation of CuT and RSol for each unit (UGCUT) with

    hard boundaries. Models are combined by geological unit.

    CuS can be deduced at point support from the simulated valuesof CuT and Rsol:

    CuS = RSol x CuT

    Block average over grades to block support simulations of

    grades at block support No ratio is ever block averaged

    Reblocked models represent the truth (10 cases), which areused for comparison purposes

    M h d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    51/81

    Methodology

    Point simulation is done by geological unit: Declustering / Transformation

    Variogram modeling

    Point conditional simulation

    Validation: histogram / variogram reproduction

    Declustereddistributions

    Gaussiandistributions

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    52/81

    Methodology

    Variable Unit ModelSill

    contributionRange 1 Range 2 Range 3

    NS_CuT 2 +3

    Rotation N45E/0 N45W/0 N0/-90

    Nugget 0.15

    Shperical 0.55 25.0 30.0 25.0

    Shperical 0.30 350.0 230.0 115.0

    NS_CuT 4

    Rotation N45E/0 N45W/0 N0/-90

    Nugget 0.15

    Shperical 0.38 90.0 85.0 65.0

    Shperical 0.47 680.0 530.0 95.0

    NS_CuT 5

    Rotation N0/0 N90E/0 N0/-90

    Nugget 0.10

    Shperical 0.40 95.0 90.0 100.0

    Shperical 0.50 450.0 330.0 185.0

    NS_RSol 2 +3

    Rotation N0/0 N90E/0 N0/-90

    Nugget 0.20

    Shperical 0.35 40.0 40.0 35.0

    Shperical 0.45 450.0 450.0 140.0

    NS_RSol 4

    Rotation N0/0 N90E/0 N0/-90

    Nugget 0.20

    Shperical 0.20 85.0 85.0 70.0

    Shperical 0.60 630.0 630.0 95.0

    NS_RSol 5

    Rotation N0/0 N90E/0 N0/-90

    Nugget 0.15

    Shperical 0.35 90.0 90.0 130.0

    Shperical 0.50 450.0 450.0 190.0

    Variogrammodels

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    53/81

    Methodology

    Dense simulation CuT grades

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    54/81

    Methodology

    Block support simulation CuT grades

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    55/81

    Methodology

    3. Dense point simulations are sampled at different grids: 6 x 6, 8 x 8 (current blash hole grid), 10 x 10, 12 x 12, 14 x 14, 16 x 16,

    18 x 18, and 20 x 20m

    The exact simulated value represents a sample without error (or withnegligible RC error)

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    56/81

    Methodology

    Dense simulation CuT grades

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    57/81

    Methodology

    Samples over a 20 x 20m grid, without error added CuT grades

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    58/81

    Methodology

    4. A Gaussian random error is simulated and added (or subtracted) tothe actual value:

    14% error for CuT

    15% error for CuS

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    59/81

    Methodology

    Dense simulation CuT grades

    M th d l

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    60/81

    Methodology

    Samples over a 20 x 20m grid, without error added CuT grades

    Methodolog

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    61/81

    Methodology

    Samples over a 20 x 20m grid, with error added CuT grades

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    62/81

    Methodology

    5. Grades are estimated over 10 x 10 x 10 m blocks: Input:

    CuT and CuS samples without error for a given grid spacing

    CuT and CuS samples with error added for a given grid

    Estimation method:

    Current short term estimation plan ID2 with search radii 12.5m horizontaland 5.0m vertical, 1/6 samples

    Improved estimation plan ordinary kriging with search radii 30.0mhorizontal and 5.0m vertical, 4/16 samples

    Requires variograms of blast hole grades for kriging

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    63/81

    Methodology

    Estimated block grades by ID2, samples withouterror added CuT grades

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    64/81

    Methodology

    Estimated block grades by kriging, sampleswithout error added CuT grades

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    65/81

    Methodology

    Estimated block grades by ID2, samples with erroradded CuT grades

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    66/81

    Methodology

    Estimated block grades by kriging, samples witherror added CuT grades

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    67/81

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    68/81

    Methodology

    Methodology

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    69/81

    Methodology

    Parameters required:

    Recovery function

    Acid consumption: function of CuS and recovery

    Drilling, sample preparation and analysis costs Sample preparation and analysis = 10.0 (US$/sample)

    RC drilling cost = 20.0 (US$/m)

    Blast hole drilling cost = 14 (US$/m) incluides 1 m subdrill

    UG CuT UG = Ox / Sulf ARC Recov (%)

    2 o 3 Ox 1, 2, 3, 9 69

    2 o 3 Ox 4, 5 72

    2 o 3 Ox 6, 7, 8 80

    4 o 5 Sulf 1, 2, 3, 9 64

    4 o 5 Sulf 4, 5 67

    4 o 5 Sulf 6, 7, 8 72

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    70/81

    Results

    Results

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    71/81

    Results

    Results

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    72/81

    Results

    Main result

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    73/81

    Current situation:BH with error + ID2 1/6

    Proposed situation:RC with low error + KR 4/16

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    74/81

    Blast hole gridNo extra drilling cost

    Any other grid spacingrequires additional drilling cost

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    75/81

    Reducing the error requiresRC drilling cost

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    76/81

    Difference due to reduction ofgeological misclassification

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    77/81

    Sampling error reduction andreduction in geological

    misclassification pays foradditional drilling cost

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    78/81

    Current situation:BH with error + ID2 1/6

    Proposed situation:RC with low error + KR 4/16

    Results

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    79/81

    Results

    Sensitivities showed that: 5% systematic bias results in: 8-12 million dollars when inverse distance is used

    0-1 million dollars when ordinary kriging is used

    Copper price does not change the recommendation and improving thegeological knowledge (from 2 to 5% misclassification) is more relevant

    RC drilling cost is an important parameter, as the additional drillingrequired when an advanced grid is used, may become very costly andovercome the profit made by better knowledge (either due to bettersamples or to better interpretation)

    Fluctuations due to spatial variability (between realizations) are small

    compared to those due to geological interpretation: Up to 4 million dollars when inverse distance is used

    Up to 1 million dollars when kriging is used

    Results

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    80/81

    Results

    Losses are more significant when using inverse distance

    Implementing ordinary kriging is simple for short term plan andreduces the conditional bias

    Using more samples in the estimation improves the results

    RC cost is significant and should be lowered as much as possible toincrease the profit

    The use of an advanced drilling grid benefits the final result in twoaspects: A more careful sampling and sample preparation procedures may be

    implemented

    A more detailed and accurate interpretation of the geological attributes

    of the blocks may result and input in advance in the short term plan The study showed that implementing an advanced drilling grid with a

    spacing of 18x18m may impact improve the profit in 130 milliondollars over the five years period evaluated.

    Acknowledgements

  • 7/28/2019 014 Alges Sampling 2011 Cmcc v2 Comp

    81/81

    Acknowledgements

    ALGES Lab Advanced Mining Technology Center

    Universidad de Chile

    Department of Mining Engineering

    Universidad de Chile

    Compaa Minera Cerro Colorado