Whittle_TakingFourXToLimits.pdf

8/9/2019 Whittle_TakingFourXToLimits.pdf

http://slidepdf.com/reader/full/whittletakingfourxtolimitspdf 1/7

Strategic Mine Planning Conference Perth, WA, 26 - 28 March 2001 163

the outline by 50 ft increments generated a similar set.E-mail: [email protected] Street, Bristol BS99 M7Y ER D, U SK ys. tem (MEDS) general mining package (GMP). Shrinking

1. Rio Tinto Technical Services, PO Box 50, Cast lemead, Lower pit outline by 50 ft increments using tools available in the2). First, a set of pit shells was generated by expanding the A04 previously designed ultimate pit outline, the A04 pit (see Figureconcentrator 20 km away, delivered 1.137 Mt of 25.9 per cent

Sector analysis, in this exercise, took as its starting point thetrators, Copperton, 8 km from the mine, and North, the older results.Total ore production in 2000 was 58.662 Mt. Two concen-analysis was run in parallel with Four-X in order to verify themolybdenum.analysis to determine the ultimate pit limits. For this study sector the world. It also yields significant amounts of gold, silver and Prior to applying Four-X, KUCC used a cone miner and sector one of the most modern and cost efficient producers of copper in

pit is nearly four kilometres wide and 800 metres deep, and is KUCC SECTOR ANALYSIS METHODthe Oquirrh mountains near Salt Lake City (Figure 1). The openKUCC operates the century-old Bingham Canyon copper mine in

The model was built using MEDSystem.THE BINGHAM CANYON MINE and costs for the block.

exposed blocks. Net value is the difference between the revenues provides an adjustment factor when calculating the tonnage of compared.the percentage of the block below the topographic surface,Bingham Canyon mine. The results from the two techniques are

techniques used in Four-X to model the complexities of the the block to the metallurgical model. The topo per cent field, ie2The paper describes KUCC s sector analysis method and the silver (oz/t) and moly (MoS ) (per cent). An ore type code links

Attributes in the model include copper (per cent), gold (oz/t),might be further enhanced.sector analysis process and to indicate areas where the designs Each block is 100 ft 100 ft 50 ft (30 m 30 m 15 m).Four-X software in order to verify the limits obtained by the columns by 160 rows by 135 levels, 3 672 000 blocks in total.generate optimum pit outlines using Whittle Programming s The dimensions of the Bingham Canyon block model are 170

KUCC also invited Rio Tinto Technical Services (RTTS) toTHE BINGHAM CANYON BLOCK MODELrefinements to the previous pit design.

and used a sector analysis technique to make incrementalmolybdenum.ultimate pit limits. For this exercise, however, KUCC devised

Historically KUCC has used a cone miner to determine the 18 400 t of molybdenum concentrate containing 10 100 t of 3.218 Moz of silver. The Copperton concentrator also delivered

revised economics (costs and prices).refinery produced 268 800 t of copper, 413 000 oz of gold and

a planned relocation of the in-pit crusher; and The smelter treated 1.072 Mt of copper concentrate and the

revisions to the geotechnical slope model; the north of the mine.copper concentrate to the Garfield smelter, also located 20 km to

of hard vs soft ore;a new metallurgical model with refinements to the treatment

FIG 1 - View of Bingham Canyon pit looking NW. block model;new drilling and consequential revisions to the geological

mine following:a review of the planned ultimate pit limits at its Bingham CanyonIn 2000 Kennecott Utah Copper Corporation (KUCC) undertook

INTRODUCTION

The results from the two techniques are compared.used in Four-X to model the complexities of the Bingham Canyon mine.The paper describes KUCC s sector analysis method and the techniques

the designs might be refined.to verify the limits obtained by this process and to indicate areas whereKUCC used sector analysis to determine the pit limits. Four-X was used

revised economics (costs and prices).

a planned relocation of the in-pit crusher; and

revisions to the geotechnical slope model;

vs soft ore;a new metallurgical model with refinements to the treatment of hard

model;new drilling and consequential revisions to the geological block

This review was occasioned by:review of the planned ultimate pit limits at its Bingham Canyon mine.

In 2000 Kennecott Utah Copper Corporation (KUCC) undertook a

ABSTRACT

R Wooller 1

Taking Four-X to the Limits



164 Perth, WA, 26 - 28 March 2001 Strategic Mine Planning Conference

Figure 4.easily adapted if necessary.elevation between the exit points for ore and waste can be seen in(Wooller, 1995). Also, being an in-house system, the code can be

Ore is hauled in the pit to an in-pit crusher. The difference infor building Four-X model files using complex parameter inputwas chosen because of the availability of tools within the system provide the lowest cost choice.

the pit (see Figure 3). The old dumps are still used where theyfrom the MEDSystem model and import them into OPD. OPDadvantage resulting from the proximity of the canyon duT mh pe sf ti orst task in the pit optimisation was to export these data being dumped in Bingham Canyon. There is a distinct cost

EXPORT TO OPD Most of the waste mined from the Bingham Canyon pit is now

MINING COSTSmacros to write the whole file.those files. This method could be extended with the creation of

grade attributes.files. This scheme was adopted to minimise errors when buildingThis was imported into an OPD model alongside the other several sheets. These were copied and pasted into the parameter

appropriate format for the parameter files were generated within write it to an ASCII file in a similar format to the input file.Structure Arcs and the Four-X model parameter file. Lines in the

calculate the concentrate grade; andRTTS s in-house Open Pit Design (OPD) software package, theinput parameter files for the OPD2WP4X program - part of exported from MEDSystem;

The optimisation parameters workbook was used in building read the ASCII file with the copper grade and ore type code

build the Four-X model and the pit optimisation. to: parameter files both for the utility programs that were used to A visual basic macro was written in optimisation parameterscontained several working sheets used to create the input

the metallurgical model.that were to be used in the optimisation. The workbook alsoGS and GI and a function of the ore type and are derived from parameters, such as costs, prices, metallurgical recoveries, etc

workbook was created as a central repository for all the GI = bulk concentrate grade intercept; andAt the start of the Four-X optimisation project an Excel

GS = bulk concentrate grade slope;OPTIMISATION PARAMETERS WORKBOOK CG = concentrate grade;

whereunprofitable material.where practical, reducing the width of pushbacks to eliminate CG = Cu per cent GS + GI

eliminating loss making pushbacks; and the formula:material; The concentrate grade is calculated from the copper grade by

enlarging current pushback designs to bring in profitable used the copper recovery must also be applied to the concentrate.of the block s tonnage, assuming 100 per cent recovery. Whenallow;concentrate grade is the quantity of concentrate as a percentageadding pushbacks to the pit where space and conditionshandling a concentrate grade block model was built. The

of the ultimate pit by: In order to calculate the costs associated with concentrateThe final stage of the process is the modification of the design

Concentrate modelments.for each sector in Figure 2 are for the +50 ft and -50 ft incre-

workbook. profitably pushed back or even shrunk. The two numbers shownThe worksheet was copied into the optimisation parametersand used as a guide in investigating where the pit can be

average revenue per ton for the sector. These were then inspected copper recovery (per cent).Canyon model. This was divided by the total tonnage to give the

SAG rate 2000 (h/kt); andshell evaluated using the net value for each block in the Binghamshells and the material between the sectors and each 50 ft pit used in the calculation of concentrate production;

A set of radiating sector lines was then overlaid on the pitcent/per cent); bulk concentrate grade intercept (per cent) and slope (per

in building the Four-X model were:FIG 2 - Sector analysis.The relevant attributes for each ore type from this model used

concentrator, assumed to be the more efficient of the two.optimisation was based on sending all material to the Copperton parameters for both concentrators were provided, theeach ore type were provided in an Excel spreadsheet. Although

-.45-.67 -.84 Canyon. The mill throughput, recoveries and processing costs for

-.55-.80 Up to 30 different ore types have b- e.5 e2 n modelled-.9 a3 t Bingham

-.41 -.76-.69-.80 1.04-.551.27-.13 -.52 Metallurgical model-.80

1.901.03.55

1.00 .23 .13 calculating the tonnage of each block.4.063.04 1.41 12.28 cu ft/ton gave an adjusted density model that was used in

-.79 3.321.48 created. Dividing the topo per cent by 100 and by the density of

.53 -.251.46.46 of 1 (rock) and all others a code of 0 (air) a rock model was

.46 .57 giving every block with a topo per cent greater than zero a code

Outline Two OPD models were built from the topo per cent data. ByA04 Pit

copied with values unchanged from the original models.moly, together with the ore type (metallurgical) code, were

The metal grade attribute models, copper, gold, silver and

R WOOLLER



7

6

9

0

7

4

9

0

7

2

9

0

7

0

9

0

6

8

9

0

6

6

9

0

6

4

9

0

6

2

9

0

6

0

9

0

5

8

9

0

5

6

9

0

5

4

9

0

5

2

9

0

5

0

9

0

4

8

9

0

4

6

9

0

4

4

9

0

4

2

9

0

$

/

t

o

n


was done through the PCAF for each block. As the processingsuite.mining cost was added as a credit to the processing cost. Thiswas built using the utility program (OPD2WP4X) from the OPDThe difference between the waste mining cost and the oreWith all the models and parameters in place the Four-X model

$1.00 per ton was used in the optimisation.waste mining cost for the bench. A reference mining cost of FOUR-X MODEL

The MCAF for each block in the Four-X model was set to the

factor (PCAF) were used. file.adjustment factor (MCAF) and the processing cost adjustment and is performed in the program that builds the Four-X model

To model these costs in Four-X both the mining costPCAF = (processing cost - ore credit)/processing cost program.

incorporate in the parameter file for the Four-X model-building Ore credit = waste mining cost - ore mining costspreadsheet. Expressions were then used to create the lines to

The calculation is:chart were incorporated in the optimisation parameterselevation is illustrated in Figure 5. The parameters that built this available when calculating the PCAF for each block.

The difference in mining costs for ore and waste by bench cost is a function of ore type it was necessary to have this cost

FIG 5 - Mining costs vs bench elevation.

Bench Elevation (ft)

0

0.2

0.4

0.6Ore

0.8

1

1.2Waste

1.4

1.6

and exit to dumps.

FIG 3 - Location of waste dumps. FIG 4 - View of Bingham Canyon pit showing location of in-pit crusher

In-pit Crusher

Canyon dumpsBingham Canyon

dumps Pit exit to Binghamdumping in

Old wasteCurrent waste

TAKING FOUR-X TO THE LIMITS




FIG 6 - Slope sectors plan.

the input parameter records for OPD2WPPN. program.number to be used for each block. It is used in the Structure Arcs in the optimisation parameters workbook, which also generatedThe profile number file is a text file that lists the slope profile conversion from pit numbers to slope profile numbers was stored

from the pit number and write the profile number file. TheProfile number file blocks from the block model, look up the profile number to use

Specialised software (OPD2WPPN) was written to read thewere used. parameters workbook.modelled within rectangular sub regions. Instead slope profiles Figure 9). These calculations were added to the optimisation

This complex slope scenario does not lend itself to being considered when creating structure arcs from slope profiles (seeslope sector detail (Figure 7) makes it clearer. effect over three benches, the minimum number of benches to beapplied with each sector, with slopes varying by elevation. The wall the slopes at these elevations were flattened to allow for itsdiagram (Figure 6). The diagram shows the slope angles to be sector diagram. To allow for the width of the ramp in the finalthe KUCC geotechnical department in the form of a slope sector wall is fairly well known. This has been included in the slopeSlope angles provided for the ultimate pit walls were provided by Being a mature pit the position of the ramps in the ultimate pit

applicable to the zone.SLOPE MODELLINGtranslated to a slope profile number according to the slope anglegiven a code unique to the shell. This was subsequently

from records generated in the optimisation parameters workbook. slope angle (see Figure 8). The blocks in each pit shell wereThe input parameter file for this program was created in part generating a volume that enclosed all blocks with a common

concentrate. zone. This was projected to surface with vertical pit walls, thusFinally parcels were created for copper, gold, silver, moly and base of an open pit shell, starting at the base elevation of a slope

the block. in OPDP. The outline of each slope sector was digitised as thecost for the bench and the ore processing cost for the ore type in AutoCAD as a DXF file and used as a guide to creating volumesadjustment factor (PCAF) was calculated from the ore mining drawing file (Figure 6). The drawing was exported fromthe waste mining cost for the bench. The processing cost The slope sector diagram was provided in an AutoCAD

The mining cost adjustment factor (MCAF) was taken from Design Program (OPDP), part of the OPD suite.density model. was done using the pit generation features of the Open Pitmodel. For rock the block tonnage was calculated from the file was to create a block model with all the slope zones. This

Whether a block was rock or air was identified from the rock The first stage in the process of building the profile number

R WOOLLER




parameters workbook. They were merged to form the final modelfrom the profile number file and the slope profile file.

Many records for this file were built up in the optimisation The Structure Arcs program was run to generate structure arcs

MODEL PARAMETERS OPTIMISATION AND RESULTS

optimisation parameters workbook spreadsheet. parameters file in its entirety from the spreadsheet.The slope profile file was created from parameters stored in the The next iteration may include a macro to generate the model

Slope profile file for all ore types into a single block.The sheet combined the Line Type 25 and Line Type 26 lines

derived from the metallurgical model.FIG 8 - Slope segment modelling for ramps. The processing recovery for each element for each ore type,

Line Type 26

metallurgical model.The Processing Cost for each ore type, derived from the

Line Type 25

from the metallurgical model.The Processing Throughput Factor for each ore type, derived

Line Type 21

Selling costs and prices for each element

Line Type 20

Results formatsLine Type 18

optimisation parameters workbook include:Lines in the model parameters file generated from the

parameters file.

FIG 7 - Detail from slope sectors.




R

e

l

a

t

i

v

e

V

a

l

u

e

T

o

n

s

(

0

0

0

)


not significantly changed.The results are shown in Figure 10.material being lost in the south. The overall tonnage of ore wasBest case, worst case and scheduled case analyses were run.additional material being mineable in the north and east, but

run on the 100 100 50 model. There is quite a significant change in the pit outline, with pit limits rather than a project value, the final optimisation was

from the previous version of the model, the A04 pit.As the objective of the exercise was to determine the ultimate

Figure 11 superimposed on the ultimate pit outline generatedslope modelling also had to be repeated at this scale. change in the waste/ore ratio. The two outlines are shown inundertaken in order to check the parameters and process. The scheduled case NPV is maximised, and Pit 15 where there is a

An initial run on a model reblocked to 200 200 100 was Two pit shells were considered of interest, Pit 18 where the

FIG 10 - Optimisation results.

Pit Num ber

11 12 13 14 15 16 17 18 19 20

0% 600,000685,136

10%

829,191 800,00020%

875,647 989,09430% Waste

944,406 1,000,000Ore 1,091,603

40%Best Case 1,024,515

1,218,41250% Scheduled Case 1,200,000

1,149,838Worst Case60%

1,400,00070%1,339,414

80%

1,600,00090%

99.11% 99.78% 100.00%

100%1,800,000

99.80% 99.81%

110%

120% 2,000,000

FIG 9 - Adjustment of slope angles at ramp elevation.

for Ramp

Slope Angle Adjusted

Overall Slope Angle

structure arc generation.

minimum number of benches in

benches to coincide with Four-XThe resulting slope is calculated over 3

Include ramp in the slope.

Slope adjusted at ramp elevation to

R WOOLLER




KUCC s sector analysis method. (Whittle Programming Pty Ltd: Melbourne).delivering similar results Four-X confirmed the validity of Wooller, R, 1995. Where Four-D Ends.. ., Optimising With Whittle

optimisation, served to give KUCC confidence in Four-X. ByREFERENCE

Running a parallel exercise, sector analysis and Four-X

CONCLUSION Corporation for permission to publish this paper.Technical Services Limited and Kennecott Utah Copper The author wishes to thank the management of Rio Tinto

final pit slope angles.On the south side the current pushback has been redesigned to

ACKNOWLEDGEMENTS been enlarged to include more payable material.Boy Corner on the north west of the pit the final pushback has

Canyon. potential north pushback is being investigated. In the Highlandincreasingly in determining the ultimate pit limits at BinghamFigure 12). A pushback has been added to the east side and a

As refinements are made to the model, Four-X will be usedDesign changes have been made to the ultimate pit shell (see

optimisation.FIG 11 - Four-X comparison at 5890 elevation. FIG 12 - Changes to the ultimate pit design following Four-X

slope angles. A04 Pit redesigned at ultimate piteliminated.

Current pushbackpushback

South side

Current Pit

NPit 15

enlarged.Pit 18N added.Corner pushback inv estigated. pushbackHighland Boy pushback being East sidePotential North


Whittle_TakingFourXToLimits.pdf

Documents