Power Geotechnical Pty Ltd

ABN 88 463 460 421

PGCA USER MANUAL

V2.1.0.3

June 2016

Tel: +61 2 4861 7825

Mob: +61 4 1243 0547
[email protected]
PGCA User Manual V2.1.0.3

Contents
Introduction....................................................................................................................................... 4
Hardware Requirements .................................................................................................................... 4
Model Interface ................................................................................................................................. 4
DMM models ............................................................................................................................ 8
Shutoff Settings ....................................................................................................................... 11
Menu Items ..................................................................................................................................... 12
The File Menu ............................................................................................................................. 14
New Project ............................................................................................................................. 14
Open Project ............................................................................................................................ 16
Open........................................................................................................................................ 17
Save ........................................................................................................................................ 23
Import ..................................................................................................................................... 24
Exit.......................................................................................................................................... 24
The View Menu........................................................................................................................... 25
Orthogonal Views .................................................................................................................... 25
Extracted Blocks...................................................................................................................... 25
The Calculate Menu..................................................................................................................... 26
Generate Recovery Data .......................................................................................................... 26
Generate Graphs ...................................................................................................................... 28
Generate recovery data for drawpoints ..................................................................................... 28
Generate recovery data for individual drawpoints..................................................................... 28
Height of draw ......................................................................................................................... 29
Summary by Level and Xcut .................................................................................................... 30
SLC Recovery Classes ............................................................................................................. 31
SLC Ring Grade Profiles ......................................................................................................... 31
Recovery Data with Level Information .................................................................................... 33
Generate All ............................................................................................................................ 33
Generate All – Single Report ................................................................................................... 34
The Optimisation Menu ............................................................................................................... 35
Calculate NPV from Recovery Data......................................................................................... 35
Optimise Schedule using NPV ................................................................................................. 36
Generate Multiple Footprints ................................................................................................... 40
The Scheduling Menu .................................................................................................................. 46

June 2016 2
PGCA User Manual V2.1.0.3

Footprint Generator ................................................................................................................. 46

Combine Lift Schedules ........................................................................................................... 57
Create SLC Schedule ............................................................................................................... 57
Visualise SLC Schedule........................................................................................................... 59
Zoned SLC Schedule ............................................................................................................... 60
The Tools Menu .......................................................................................................................... 62
Advanced Flow Settings .......................................................................................................... 62
Batch Mode ............................................................................................................................. 67
Design ..................................................................................................................................... 70
DXF ........................................................................................................................................ 74
Inspect Drawpoints .................................................................................................................. 79
SLC Blast to Ore ..................................................................................................................... 79
Appendix A - Formatting of input files ............................................................................................ 81
Schedule format for SLC ......................................................................................................... 81
Schedule format for block caving ............................................................................................. 81
Block model format ................................................................................................................. 82
Drawpoints file for block caving .............................................................................................. 82
Development file for SLC ........................................................................................................ 83
Ring type file for SLC.............................................................................................................. 83
Undercut file for Block Caving ................................................................................................ 84
Marker data input file .............................................................................................................. 84
Appendix B - Model Calibration...................................................................................................... 85
Appendix C - Formatting of DXF Files............................................................................................ 87
Appendix D - Effect of Model Resolution (cell size) on width of draw in PGCA.............................. 88

June 2016 3
PGCA User Manual V2.1.0.3

Introduction
This manual describes the method of use and functionality of the PGCA software package. PGCA
(Power Geotechnical Cellular Automata) was originally a stand-alone cellular automata modelling tool.
It has now been expanded to include other modelling tools and functions, but retains the name PGCA.
It can be tailored to simulate flow for all forms of caving mining. For the purposes of this document,
the two main caving mining areas will be addressed (sublevel and block caving). Other variations such
as panel caving or incline caving can be grouped into one of these areas depending on whether they
have one production level (block caving) or multiple advancing production levels (sublevel caving).

Hardware Requirements
PGCA runs on Windows platforms. The speed of model simulation will depend on processor speed,
memory availability and size of the model being run. Most modern computer processors will be
adequate. Whether a model can be loaded into memory will depend on random access memory (RAM).
Cellular automata (CA) models are currently standardized to require 8MB RAM. Therefore the
computer being used should have 8GB or more RAM available. Licenses are limited to individual
computers and cannot be shifted or copied to alternative computers without assistance from Power
Geotechnical (PG). The License file supplied with the executable must be located in the same directory
as the executable.

Model Interface
Figure 1 shows the model interface and describes some of the controls which can be used for an SLC
PGCA model.

1
11
2
3

5 8

6
7
9
10

June 2016 4
PGCA User Manual V2.1.0.3

Figure 1: Model interface options for a sublevel cave example

Features of the modelling interface shown in Figure 1 are:

1. Function Drop Down Menu: Commands described in the section following;

2. Screen View: View is selected from the View Dropdown menu. Desired coordinate or
bearing for view can be input into the Model Coordinates text boxes on the bottom left of the
screen. A slider for bearing is provided, and bearing can also be adjusted using keypad
arrows. The scroll bar on the right of the screen can be used to scroll in the direction of view.
Zooming in and out of the view can be achieved using the mouse wheel;

3. Model Dimensions: These are automatically generated by the program based on the drawpoint
layout and block model dimensions;

4. Grade Legend Control: This denotes the maximum grade by which the grade legend is
automatically calibrated. The value selected is divided into three grade increments and used to
generate the legend on the top left of the screen;

5. Shut-off inputs: Constraints by which shut-off modelling is controlled, if it is activated in the

modelling schedule (discussed in detail in the Shutoff Settings section);

6. Draw Input Width: Control which allows width of draw in the model to be adjusted;

7. Colour by Material Type: Toggle to show the Material Field from the block model instead of
grade field. Material type is read directly from the mandatory field of this name in the block
model. Material type values up to 21 can be used to indicate parameters like rock type, height
above production level, resource code etc. Material type 19 is reserved to indicate fines,
material type 20 solid unmovable material, and material 21 alternate flow bias properties;
Fines identified as Material type 19 in the block model will show up as yellow on the screen
when the colour by material box is checked (e.g. Figure 4). Material identified in the block
model as Material type 20 will be made solid and immovable. Solids show as black when the
colour by material box is checked.

8. Fines Migration Tab) – helps user to produce different flow properties for different rock types
within the model. This contains three text boxes for input parameters:
o Fines Probability, which sets the width of draw of the fines using a decimal value.
The large the decimal value the narrower the width of draw for fines particles
(generally set to a relatively narrow value due to the small size of the fines);
o Fines Mobility, which sets the rate of travel of the fines as a multiple of the rate of
travel for normal particles;
o IMZ/IEZ ratio, which sets the rate at which the movement zone grows relative to the
extraction zone (generally leave this value at the default value).

The “Show Subsidence” toggle allows the a function which automatically replenishes
waste at the top of a model to be switched;

June 2016 5
PGCA User Manual V2.1.0.3

9. Grade View Field: Allows the grade of interest to be viewed, where different grades are being
tracked. Once a model has started running this cannot be changed, as it requires the block
model to re-load. The field loaded automatically when the program opens will be the first in
the list of grade fields in the block model, and this field will also be the one referenced by the
shutoff algorithm;

10. Use Cave Shapes. When this is toggled on at the start of a simulation, the program will
request a folder in which the cave shapes are stored. These shapes must be named according
to the number of the period in which they will be opened. For SLC this will be the difference
between the start date of the model and the desired opening date in days e.g. 4076.DXF. For
block caves it will be the production period in which the shape is to be opened. Once selected,
the shapes will automatically open when the appropriate modelling period is reached, and
constrain flow.

11. When hovering the mouse over the model, the porosity reading for that specific area is shown
(see circled number in Figure 5). The location of the voids producing swell in the model can
be indicated on the screen as black particles by clicking on the porosity reading, and toggled
off by clicking the porosity reading again.

Figure 2 shows a CA block caving example. Most of the features for block caving are similar to those
for SLC, but some differences are numbered and described below:

Figure 2: Model interface options for a PGCA block cave example

1. The Shutoff Criteria tab has a different format (discussed in the shutoff section);

June 2016 6
PGCA User Manual V2.1.0.3

2. "Period" is used to measure modelling progress rather than "Simulation Date" as in the SLC
version. This is because the schedule modelling inputs (daily, monthly or yearly) are not
specified on the schedule input file format for block caving;

Figure 3: The Fines Migration Tab

Figure 4: Block caving model showing fines as material type 19

June 2016 7
PGCA User Manual V2.1.0.3

Figure 5: Block caving model showing porosity reading with voids toggled on (black particles)

DMM models
DMM (Differential Mixing Model) is a modelling option which uses analysis of the algorithms used in
the CA models to create a more rapid modelling option. DMM models generally complete life of mine
simulations in less than a minute. As they are intended as rapid high level approximations of CA, many
of the more specialized functionality available in the CA models are not available in the DMM models.

Figure 6 shows a DMM block caving example. The block model is shown on the screen when the
model is loaded, and the view can be manipulated using the View menu item as for CA models. The
projects are created and saved in the same way as CA models (described further below). Once a model
has been run, zones of recovery will be shown on the screen (e.g. Figure 7) and recovery files output
directly to a folder selected by the user. Due to the simplicity of the DMM model format, only the
width of draw is changeable in the user interface for DMM models.

June 2016 8
PGCA User Manual V2.1.0.3

Figure 6: Model interface options for a DMM block cave example

DMM models generally calibrate as well as CA models for free flow situations. Figure 8 shows a
comparison of reported grades for a DMM model from an operational mine when compared to the
forecast grades for the same project in the CA model. Over the life of the mine the difference between
the two is less than 1%. DMM models may not be suitable for situations where cave propagation, fines
migration or pit rilling are significant factors. Visualisation, modelling input options and reporting
functions are also currently more limited in DMM than in CA. However it is very useful for rapid high
level simulations, and where many simulations are required such as in optimization cases (discussed in
subsequent sections)

June 2016 9
PGCA User Manual V2.1.0.3

Figure 7: Extracted zones shown in DMM block caving example

Figure 8: Example of DMM grades plotted as a % of PGCA grades

DMM for SLC is currently under development.

June 2016 10
PGCA User Manual V2.1.0.3

Shutoff Settings
Shutoff modelling is only available in CA models. Figure 9 shows default shutoff settings for an SLC
(left) and a block caving model (right).

Figure 9: Shutoff settings for SLC projects (left) and for block caving projects (right)

The input parameters shown for SLC shutoff modelling are defined as follows:

 Instantaneous or Sample Average shutoff type determining the shutoff algorithm the program
will use (described in more detail below)
 Sampling rate – Increment at which measurements are taken at the drawpoint to determine
whether a drawpoint is to be shut off;
 Initial Shutoff (Tonnes) - Tonnes to be drawn before any shutoff is applied;
 Initial Shutoff Grade - Shutoff grade to be applied after minimum draw is exceeded;
 Final Shutoff (Tonnes) - Tonnes to be drawn before the “Final Shutoff Grade” is to be applied;
 Final Shutoff Grade – Shutoff grade to be applied after the “Draw Factor” is reached;
 Allowable empty buckets – The maximum number of times that a drawpoint can report that no
ore is available to draw in a given production period before draw is curtailed at that drawpoint
due to it being open. When generating results, an automatic output file called
“SLCRingCompletion.csv” reports which rings have shutoff due to this figure being exceeded.

The input parameters shown for block caving shutoff modelling are less complex than the SLC options.
The block caving settings are defined as follows:

 Instantaneous or Sample Average shutoff type determining the shutoff algorithm the program
will use (described in more detail below)
 Sampling rate – Increment at which measurements are taken to determine whether a drawpoint
is to be shut off;
 Initial Shutoff (Tonnes) – percent of the scheduled tonnage which must be drawn before shutoff
can be applied;
 Final Shutoff Grade – Shutoff grade to be applied after initial shutoff tonnage is reached;
 Use shutoff check box – determines whether shutoff will be used in a model run;

Two separate shutoff algorithms are supported. Instantaneous shutoff can be used to simulate geology
controlled draw control practice for the forecasting of metal recovery. It samples the drawpoint at the

June 2016 11
PGCA User Manual V2.1.0.3

intervals specified by the Sample Rate setting, and shuts the drawpoint down if the grade is below the
specified grade at that point in time.

Sample average shutoff is useful in optimisation of draw, and planning of draw control strategies. This
method computes a rolling average at the drawpoint over the tonnage specified by the Sample Rate, and
shuts the drawpoint down if the rolling average of the grade falls below the specified grade for the
cumulative tonnage drawn at that drawpoint.

Figure 10 shows the effect of the different shutoff algorithms on outcomes. The upper graph shows a
ring set to shutoff at 0.9% using the instantaneous algorithm. The instantaneous grade falls below 0.9%
on a few occasions after the Draw Factor of (specified in this case at 4000t), and finally drops below
the 0.9% shutoff criterion coincidentally with the 200t sample window after a total of about 23,000t
drawn. However the rolling average at this point is still above the shutoff grade, and were sample
average shutoff used as the criterion for shutoff, the ring could continue in production. The lower graph,
shows the sample average shutoff applied in the same case, allowing shutoff to push out to
approximately 10 times the tonnage allowed by the instantaneous shutoff criterion.

Figure 10: Example of the effect of shutoff method on results

In SLC models, individual ring types can also have individual shutoff settings. This allows the user to
change shutoff behavior in different parts of the mine. Use of different shutoff settings for different
rings is controlled using the Ring Type input file, explained further later in this document.

Menu Items
Before loading a file, only functions in the File, Optimisation, Scheduling and Tools menus can be
activated. The other menus are greyed out, as their functions are not yet available. This greying out of
menu items is generally used to show which functions are available to at various modelling stages,
according to the model type being run, and licensing availability.

June 2016 12
PGCA User Manual V2.1.0.3

Figure 11: Example of greying out of non-usable menu items

The following sections will explain each of the menu items available.

June 2016 13
PGCA User Manual V2.1.0.3

The File Menu

Figure 12 shows items available in the File menu.

Figure 12: Items available in the File Menu

New Project
To load a model, the user must either open an existing project, or create a new project. To create a new
project, select the “New Project” command. This will open the form shown in Figure 13. The tab
chosen will determine the model type to be used (DMM or PGCA). Within each tab, a choice of Sub
Level Cave and Block Cave exists. To create a sublevel cave, check the appropriate box, and likewise
for a block cave.

For sublevel caves, each project requires schedule, block model, development and ring type files, as
shown in Figure 13. Click on the open file icons to navigate to the files to include in the project. These
files must be formatted correctly. Discussions of the correct formats for these files are contained in
Appendix A. Model resolution for sublevel caves is currently fixed to 1.25m.

Figure 13: New Project form for a sublevel cave

June 2016 14
PGCA User Manual V2.1.0.3

For a block cave, the New Project form is slightly different (Figure 14). Drawpoint, Schedule and Block
model files are required. The use of an Undercut file is optional. Examples of the correct formats for
these files are also shown in Appendix A.

Figure 14: New Project form for a block cave

As each file is selected, relevant information will be shown on the screen (Figure 15). It is useful to
check this information against expected values to help avoid using incorrectly formatted files in creation
of a project.

June 2016 15
PGCA User Manual V2.1.0.3

Figure 15: Example of information provided when creating a new project

Once all of the input files are selected, a block resolution should be selected from the dropdown menu
(Figure 16). The model defaults to the smallest equi-dimensional block size it can create based on
available memory and cell size in the input block model. However commonly there are other larger
block sizes available. Larger block sizes will reduce memory used (allowing multiple models to be run
in much quicker simulation times), and can generally be calibrated as well as smaller block sizes.

Figure 16: Selecting a block resolution before creating a model

Once a block resolution has been selected, click on the “Create” button to build the project. This may
take a few minutes to complete and will result in the model loading. Once the model is loaded, select
“File>Save Project” from the drop down menu, and give the project a name (Figure 17). The user can
then click on the Start button to begin the simulation. Before running, the program will ask what the
results file is to be named. Although all input and output files for PGCA and DMM are text files, project
files should be named with the extension .prj and extracted block files with the extension .ebf.

Figure 17: Saving a project

Open Project
To open a previously created project, select the “Open Project” command (Figure 18).

June 2016 16
PGCA User Manual V2.1.0.3

Figure 18: Opening a project

The user will be prompted to select a project name, and then the model will load, and can be started by
clicking on the start button. Before clicking on the start button, ensure that the settings required for the
model run are correct. Some points to remember include:

 Select the appropriate draw control inputs;

 Select the appropriate width of draw;
 Select the grade field you wish to view from the "Field" dropdown. This grade field cannot be
changed once the model is started. Click out of this box once the grade field is selected as it
is possible to scroll through the grades, and doing so inadvertently will cause the model to
continually reload new grade fields;
 If the "Use Cave Shapes" box is toggled on, the program will request a folder from which to
extract the cave shapes. The files in this folder must be named by the period in which they are
to be opened (e.g. 2047.DXF). As the program runs, the shapes will automatically open and
constrain flow.

Open
The Open function gives the user the options shown in Figure 19.

Figure 19: Options available for the "Open" function

June 2016 17
PGCA User Manual V2.1.0.3

Open Block Model

The "Open>Block Model" command will open a block model for viewing on the screen at the resolution
of the cells in the block model.

Open Model State

The "Open>Model State" will open a model state (a full flow model resolution model, from which a
previous run can be continued, or a residual model shown at full resolution can be viewed). To open a
model state, the original project upon which that model state is based must first be loaded (the command
is greyed out until a project has been loaded). The model state will only contain the field that is loaded
on the screen during the model run. Other grades not loaded on the screen when the model was being
run will not have been preserved.

When opening a Block Model or Model State, make sure another model isn't already loaded into
memory. Opening multiple models at the same time may cause computer memory to reaching its
available limit. A user can open multiple models in different directories, but if these models exceed the
available memory, it is likely that the computer will need to be restarted (with loss of unsaved data).

Open Cave Shapes

The "Open>Cave Shapes" command is greyed out until a project has been opened. Once a project is
open, five cave shape options are available under this submenu item (Figure 20). A cave shape is a
DXF file used to constrain the model so that only areas known to have caved can flow to drawpoints.
When loading a cave shape DXF, it is important that the schedule to be run does not contain production
drawpoints outside the cave shape, or flow may not be constrained as expected. DXF files can be
formatted in a number of ways; however the program only recognises some of these, which are
discussed in more detail in Appendix C.

Figure 20: The open cave shapes submenu

From DXF
The "Open>Cave Shapes>From DXF" command uses standard DXF formats discussed in Appendix C
to open a cave shape within a project. This can be done to provide a limiting boundary at the start of a
modelling run, or simply to view a cave shape. Figure 21 shows an example of a cave shape being used

June 2016 18
PGCA User Manual V2.1.0.3

in a modelling run. (some flow has occurred outside the cave shape on the right hand side due to a
drawpoint being placed outside the cave shape).

Figure 21: Example of a cave shape opened from a DXF

From Points
The "Open>Cave Shapes>From Points" command uses a simple points file in the format shown in
Figure 22 in place of a DXF to produce a cave shape or limiting boundary. The purpose of this function
is to allow the user to construct their own limiting boundaries simply using a spreadsheet. The file
should have an extension ".pnt", and should contain space delimited X, Y, and Z coordinates, with each
new point on a separate line. To produce a boundary impervious to flow using a points file, the points
must be spaced within 5m of each other.

Figure 22: Format for a points file

An example of a cave shape created using a points file is shown in Figure 23

June 2016 19
PGCA User Manual V2.1.0.3

Figure 23: Example of a points file used to create a cave shape

From RAW
The "Open>Cave Shapes>From RAW" command produces a cave shape from a .raw file, which has
the format shown below. This consists of X, Y and Z coordinates for the three points of a series of raw
triangles.

Figure 24: Format of a RAW file

June 2016 20
PGCA User Manual V2.1.0.3

Final Failure DXF

The "Open>Cave Shapes>Final Failure DXF" command allows the user to specify a DXF which will
create a solid boundary outside which flow cannot pass, creating a final failure envelope. An example
of the use of such a failure DXF is shown in Figure 25.

Figure 25: Example of use of a final failure DXF

Hangingwall DXF
The "Open>Cave Shapes>Hangingwall DXF" command is designed to allow a limiting boundary to
open on the hangingwall of a steeply dipping orebody to help simulate a situation where hanging wall
failure is expected to be slow (and to preclude waste entering the production drawpoints too early).
This function is under development and is greyed out accordingly.

Convert Pit to Sky

The "Open>Convert Pit To Sky" command allows use of a DXF to extract a pit shell from a block
model. This is done by opening a project, and then selecting the Convert to Pit Sky command. This
will bring up the form shown in Figure 26. A flat layer of waste can be added to the top of the pit to a
selected thickness, or if a thickness of 0 is selected no waste will be added.

June 2016 21
PGCA User Manual V2.1.0.3

Figure 26: The Load Pit DXF form

The waste added to the base of the pit will be material type 19 (fines) and will travel according to the
properties used in the Fines Migration tab. The grades in the block model will show in a grid allowing
the user to assign a grade for each to the pit material.

Once the pit shell is removed from the block model, the project can be run. If a project is run and
closed, and opened later for calculation of grades from the .ebf file, the pit shell must be extracted using
the same procedure as when the model was originally run before calculation of grades can be completed
accurately (if it is not extracted, the model will assume that the fines in the pit have the grades of the
blocks which were there originally, rather than as waste which should be the case).

Figure 27: Example of fines added at base of a pit

June 2016 22
PGCA User Manual V2.1.0.3

Save
Options in the File>Save menu are shown in Figure 28

Figure 28: Options in the "Save" function

Block Model
The File>Save>Block Model function will save the residual model from a run that has been completed
back to the original block model format. This command must be initiated while the Project is loaded.
The saved block model file will incorporate the voids contained in the PGCA model, which will affect
the density of the cells in the Block Model file, however the contained metal will be an accurate
reflection of metal remaining in the model. This block model can then be opened later and viewed
using the File>Open>Block Model command.

If the user wishes to run a project with a pit many times (e.g in batch mode) or a pit in a DMM model,
it is possible to open the project and load the pit dxf (as discussed above), and then save the block model
out. This will save the approximate shape of the pit and any waste it contains (at the resolution of the
original block model). The saved blocks in the pit region will contain no density or grade and will be
shown as material type 5. Because the pit shell will be less precise when fitted to the block model, there
will be a slight difference in grade for the block model cells on the margin of the pit (estimations of
metal difference in testing of a typical model showed a difference of approximately 0.002% diffrence
in contained model metal).

Model State
The File>Save>Model State command will save a residual model in its full resolution cell size format,
so that it can be opened again later and viewed, or a simulation can be continued. The model state will
only contain the field that is loaded on the screen during the original model run. Other grades not loaded
on the screen when the model was being run will not be preserved. A previously saved model state can
be opened by using the File>Open>Model State command.

Options
File>Save>Options will generate the options shown in Figure 29.

June 2016 23
PGCA User Manual V2.1.0.3

Figure 29: Options available for the File>Save>Options command

Save Once, Period number: will generate a model state at the selected period. This allows the user to
pre-select the save point without having to manually stop the model.

Save Periodically, Interval: will generate a model state at intervals selected by the user. So if there are
100 periods in the schedule, and the user inputs the value 20, then the model will save 5 times (each
model state as large as 5GB). This could take a lot of storage space, therefore one needs to be careful
with the number of model states saved. Save by level, Level RL: is a one-time option which allows you
to generate a save state when production on a nominated level in an SLC model is reached.

Import
The only option currently available under the Import Menu is the Import Block Model function. This
function allows the user to import a block model and output for use in a PGCA model. The model can
be cut to specified dimensions. Options also exist to sort the model, create a hangingwall contact for
modelling of inclined orebodies, and for the insertion of production drives on the contact.

Exit
The command File>Exit will close the program.

June 2016 24
PGCA User Manual V2.1.0.3

The View Menu

Figure 30 shows options available under the View Menu

Figure 30: Options available in the View menu

Orthogonal Views
The Looking North, Looking South, Looking East, Looking West and Plan View commands enable
the user to change to an orthogonal view in the interface once a model has been loaded. Views can
also be changed using these commands while a model is running. A section on any bearing can be
viewed by inputting the desired bearing in the model coordinates box on the lower left corner of the
interface.

Extracted Blocks
The Extracted Blocks command enables the user to show extracted blocks from a model run mapped
against the original block model in its discretized form before the run commenced. This allows a
view of areas in the model from which model blocks have been recovered, and also the areas which
were not recovered. Before running a model, the user is requested to select an extraction file (.ebf
file) in which to save the modelling results, and it is this extraction file which must be selected when
the Extracted Blocks command is selected.

If the user wishes to view a model's residual state (i.e. the actual location of blocks at the end of a
model run, rather than the original position of blocks eventually extracted, then a model state needs to
have been saved previously, and is loaded using the File>Open>Model State command.

June 2016 25
PGCA User Manual V2.1.0.3

The Calculate Menu

Figure 31 shows options available under the calculate menu for PGCA block caving models. Figure 32
shows options available under the Calculate menu for PGCA SLC models.

Figure 31; Options available under the Calculate Menu for PGCA Block caving models

Figure 32: Options available under the Calculate Menu for PGCA SLC models

Generate Recovery Data

The Generate Recovery Data command generates three files: “Recovery.csv”, “RecoveryByRing.csv”
and “DevelopmentRecovery.csv” for SLC projects, and “Recovery.csv”, RecoveryByDrawpoint.csv”
and “UndercutRecovery.csv” for block caving projects. These files will be output into the same
directory as the project file from which they are being generated.

For SLC models, the “Recovery.csv” file delivers an output like that shown in Figure 33. This shows
production for each day between the first ring and the last ring in the production schedule, regardless
of whether production occurred on every production day.

June 2016 26
PGCA User Manual V2.1.0.3

Figure 33: Typical output from a Recovery.csv file for an SLC model

The “RecoveryByRing.csv” file delivers an output like that shown in Figure 34.

The last column in this file contains a field called EstimatedFiredTonnes. This is an estimate of tonnage
in the ring based on the shape of the ring, and the density of a single block located 5m above the top of
the drawpoint. As such it can be considered an estimate suitable for broad comparisons. It is considered
generally reasonable as all of the inividual cells in the model take their density values from considerably
larger block model cells.

Figure 34: Typical output from a Recovery by Ring.csv file for an SLC model

The “DevelopmentRecovery.csv” is shown in Figure 35. This file indicates the amount of metal pre-
depleted by the development before production began. It is not scheduled out, and is based on 5m x
5m drives.

Figure 35: Typical DevelopmentRecovery.csv output file

For block caving models, the “Recovery.csv” file delivers an output like that shown in Figure 33, and
the RecoveryByDrawpoint.csv file an output like that shown in Figure 36

June 2016 27
PGCA User Manual V2.1.0.3

Figure 36: Typical output from a RecoveryByDrawpoint.csv file for a block caving model

Generate Graphs
The Generate Graphs command outputs the results from the Generate Recovery Data command into an
Excel file which is named specifically by the user when executing the command. Automatic scatter
plots are generated for Metal Recovery, Grade and Production tonnes in separate worksheets in this
Excel file. An SLCRingCompletion tab is also generated for SLC projects.

Generate recovery data for drawpoints

The Generate Recovery Data for Drawpoints command is only relevant to block caving models. It
generates a "DrawpointTonnes.csv" file, which shows tonnes extracted at each drawpoint (e.g. Figure
37), and accompanying files showing metal and grade extracted at each drawpoint for each grade field
in the block model using a similar format. Also generated is a “HeightOfDraw.csv” file. These files
will be output into the same directory as the extracted blocks file from which they are being generated.
A further discussion of the HeightOfDraw.csv file is shown further below, and the file’s format can be
seen in Figure 40

Figure 37: Example format for a "DrawpointTonnes.csv" file

Generate recovery data for individual drawpoints

The Generate Recovery Data for Drawpoints command is only relevant to block caving models and
produces a form like that shown in Figure 38.

June 2016 28
PGCA User Manual V2.1.0.3

Figure 38: Drawpoint history form

This allows the user to select an extracted blocks file file and a drawpoint named in this file and
produce a production history for that drawpoint such as is shown Figure 39. This is useful for
analysis of individual drawpoints at a more detailed level. The file will be output into the same
directory as the extracted blocks file from which it is generated.

Figure 39: Example of output from the history of an individual drawpoint command

Height of draw
The Height of Draw Command generates a report like that shown in Figure 40 for an SLC project. This
report shows cumulative height of draw for each SLC drawpoint in 500t increments. The same
command for block caving produces cumulative height of draw by drawpoint and period (Figure 41).
The file will be output into the same directory as the extracted blocks file from which it is generated.

June 2016 29
PGCA User Manual V2.1.0.3

Figure 40: Example Height of Draw output for an SLC project

Figure 41: Example Height of Draw File for a block caving project

Summary by Level and Xcut

The Summary By Level and Xcut command is only relevant to SLC models and generates summary
recovery information by Level and Crosscut in a file called "SummaryByLevelAndXcut.csv", which
has a format like that shown in Figure 42, and will be output in the same directory as the project file.

Figure 42: Typical output from a Recovery by Level Command

June 2016 30
PGCA User Manual V2.1.0.3

SLC Recovery Classes

The SLC Recovery Classes command is only relevant to SLC models and generates recovery
information with associated information on the recovery class the ore was coming from (i.e. Primary,
secondary, teriary etc.). The output file is named “SLCRecoveryClasses.csv” and will be output in the
same directory as the project file.

Figure 43: SLC Recovery Classes form

SLC Ring Grade Profiles

There are two submenu items available under the SLC Ring Grade Profiles command. These are
intended to allow calculation of grade histories for an individual ring, and for multiple rings in an
SLC schedule.

Individual Ring Grade Profile

The Individual Ring Grade Profile command is only relevant to SLC models and generates incremental
production information for a specific ring nominated by the user. On selection of this submeny item,
the form shown in Figure 44 will appear. Select an extracted blocks file and a ring name for which a
grade profile is desired. The Generate Profile button will activate the command.

Figure 44: Form for the Individual Ring Grade Profile Command

The output file generated by this command will be named the same as the ring selected and will be
generated in the same directory as the project file. It will contain information similar to that shown in
Figure 45:

June 2016 31
PGCA User Manual V2.1.0.3

Figure 45: Typical output from a specific ring recovery output

Bulk Ring Grade Profiles

Before running this function, make sure that any pit shapes used in the original modelling run to be
assessed are open in the project.

Selection of the menu item will produce the form shown on the left in Figure 46. Select an extracted
blocks file first. If all of the ring histories in the schedule are desired, check the “Generate All Grade
Files” checkbox. It can take up to half an hour to process a project with a large production schedule.
If only a selected number of grade profiles are desired, don’t check the “Generate All Grade Profiles”
box, and instead load a list of individual ring names from a csv form as shown in the example on the
right of Figure 46.

Figure 46: Form for the Bulk Ring Grade Profiles function (left) and format for list of specific rings (right)

Two files are output when this function is run and will be located in the same directory as the project
file.

The “Bulk Ring Grades Instantaneous.csv” file shows the grades for each ring for 100t production
increments. The “Bulk Ring Grades Cumulative.csv” file shows the cumulative production grades for
each ring in 100t increments. An example of a comparison of data from an individual ring is shown in
Figure 47. This shows that the individual ring history (which outputs approximately bucket sized
increments) is more variable than the instantaneous grade profile (output in 100t increments), and how
both of these are more variable than the cumulative grade profile for the ring.

June 2016 32
PGCA User Manual V2.1.0.3

Figure 47: Example of Instantaneous and Cumulative Grade Profiles and an Individual Ring History for an
individual ring.

Recovery Data with Level Information

The Recovery Data with Level Information command is only relevant to SLC models can take some
time to calculate. When complete it produces a report (“SLCRecoveryClasses.csv”) similar to that
shown in Figure 48 in the same directory as the project currently open. This includes all production
information for each grade field for each day in the production schedule as well as class information
such as Primary, Secondary etc.

Figure 48: Typical output from a Recovery Data with Level Information command output

Generate All
The Generate All function is only relevant to block caving models and produces all of the block caving
reports (output to the same directory as the project file), and leaves the extracted blocks file open on the
screen on completion. This is useful for generation of documents for external reporting on modelling
work.

June 2016 33
PGCA User Manual V2.1.0.3

Generate All – Single Report

The Generate All - Single Report function is only relevant to block caving projects. It generates a
HeightOfDraw.csv file as well as a file called “Full Report.csv” which contains grades and height of
draw for each drawpoint in all production periods (e.g. Figure 49). The reports are output into the same
directory as the project file from which they are calculated. This command also leaves the extracted
blocks view on the screen.

Figure 49: Example of output from the Generate All - Single Report command

June 2016 34
PGCA User Manual V2.1.0.3

The Optimisation Menu

The Optimisation Menu currently allows the options shown in Figure 50.

Figure 50: Options under the Optimization menu

Calculate NPV from Recovery Data

This option allows the user to generate an estimate of NPV from any modelling run which has been
proviously completed. When selected, it will open a form as shown in Figure 51.

Figure 51: Form used to calculate NPV from a previously completed modelling run

The user is asked to select a directory in which the previous modelling run sits. When this is selected,
the model will read the recovery files in this directory and populate the form with the available block
model fields (example in Figure 52). Block model fields that contribute to NPV can be given a dollar
value. When setting the dollar value, also set the units field to the appropriate unit. At present, gold
and silver should be be in ounces per tonne and all other fields in dollars per tonne. If a field is not
valid for this calculation, it can be left at 0 and will not be used.

Operating Cost should should include mining cost, processing cost and general administration costs

Once this form has been completed, the calculation can be completed clicking on the “Generate NPV
of Schedule” button.

June 2016 35
PGCA User Manual V2.1.0.3

Figure 52: Example of populated NPV calculator form

Output for this process comprises a file called “NPVOutput.csv” file which will be placed in the
directory containing the recovery files being assessed. An example of the format of this file is shown
below (Figure 53)

Figure 53: Example of NPVOutput.csv file

Optimise Schedule using NPV

This option allows the user to create hill of value information for a situation where a basic layout and
scheduling information are assumed or known. If required some of these inputs can be generated using
the footprint generator. The form used for this function is shown in Figure 54. Single footprint or
multiple footprints options can be selected.

In the input tab of this form, some fixed parameters are selected comprising:

 Drawpoints file;
 Sequence file;
 Rampup file;
 Block model;
 Base Parameters file (as described in the Tools/Batch mode section);
 Location of directory where the output files are to be placed.

June 2016 36
PGCA User Manual V2.1.0.3

Figure 54: The Input files tab for the Schedule optimization form

Once these fields have been filled, the user should move to the NPV settings tab, which is shown in
Figure 55. The fields from the block model will be shown, and values can be input for calculation of
NPV as described in the previous section. Once the NPV settings tab has been populated, the user
should move to the Scheduling Parameters tab (Figure 56). This will allow the user to set up a batch
mode run which will test different widths of draw, production rate and shutoff values. In the example
shown, 5 realizations have been selected for each of the three variable scheduling parameters. This
would result in 5x5x5 DMM modelling runs being completed (125 runs), which would take about 4
hours in total.

Each of the modelling runs will automatically create its own subdirectory in the selected output
directory. There are two phases to the modelling process. In the scheduling phase, schedules and
tonnage files will be created in each of the created subdirectories. In the modelling phase, DMM models
will be run and results output into each of these directories. At the end of the modelling process, a
summary file called NPVSummary.csv will be output into the selected modelling directory. This will
list each modelling option run and the NPV generated by each. This information can then simply be
converted to hill of value graphs to give estimates of sensitivity of the case to each of the variable
modelling parameters. An example of such a hill of value graph is shown in Figure 59. Note that since
only two variables can be graphed in these graphs, separate hill of value graphs must be created for each
width of draw realization.

June 2016 37
PGCA User Manual V2.1.0.3

Figure 55: The NPV settings tab

Figure 56: The Scheduling Parameters tab

June 2016 38
PGCA User Manual V2.1.0.3

If the multiple footprints option is selected, an automatic sequence will be generated for each footprint
(i.e. the Select Sequence File option will be greyed out in the Simulation Files tab). Currently these
automated sequences are started at the drawpoint with the highest grade. Plans exist to change this
starting point to the centre of gravity for metal content of the footprint. Once the most suitable footprint
has been selected, a more desirable sequence can be generated by the user in the scheduler and used
again in the single footprint option. The other tabs in the multiple footprint option are identical to those
for the single footprint option.

Figure 57: Progress bar showing multiple schedules being created for each of the footprints being assessed

Figure 58: Example of NPV Summary.csv file

June 2016 39
PGCA User Manual V2.1.0.3

Figure 59: Example of Hill of Value graph created from output of the schedule optimization process

At present fixed costs and metal values are used for these calculations; however variable metal price
and cost assumptions (as a function of production rate and /or time) are likely to improve the
functionality of the schedule optimizer.

One advantage of Optimise Schedule Using NPV function is that footprints saved out from the Generate
Multiple Footprints function (described subsequently), which use NPV as a basis for footprint
generation can be assessed for NPV based on use of different shutoff grades in the actual mining
process.

Generate Multiple Footprints

This option allows the user to rapidly create footprints and schedules for comparative purposes using
only an appropriately formatted block model as a starting point. Once this option is selected, import
the block model to be used as shown in Figure 60

Figure 60: Create Multiple Footprints selection

June 2016 40
PGCA User Manual V2.1.0.3

After a few seconds, the Layout screen shown in Figure 61. Select the appropriate layout options, and
press “Generate Dummy Drawpoint Layout”. A dummy layout will appear as shown.

Figure 61: The Layout tab

Once the dummy layout appears, move to the Level Placement tab, which will bring up a screen as
shown in Figure 62. The upper slider is used to locate the lowest footprint level. The lower slider is
used to place additional levels at intervals determined by the Level Increment input box. The minimum
and maximum draw height inputs allow the user to filter out unrealistic drawpoint heights. Finally, a
generalized ramp up file must be selected, which will be applicable to all drawpoints in every footprint.
This file should contain an arbitrary drawpoint name, and the tonnes for each period in the ramp up on
a single line, all space delimited. Once this tab is complete, move to the NPV Settings tab (shown in
Figure 63)

The top half of the NPV settings tab is similar to the NPV settings used in previous sections of the
Optimisation menu. The lower half of the tab allows the user to select variations in draw width and
minimum NPV for inclusion in the footprint. Once these settings have been filled in, the user can select
the “Generate Footprints by NPV” button. This will activate the batch mode and start running all of the
options selected. A progress bar will show work proceeding, and at the end of the process a “Footprint
Creation Complete” tab will appear (Figure 64). At this point, the user can move to the View Footprints
tab.

June 2016 41
PGCA User Manual V2.1.0.3

Figure 62: The Level Placement tab

Figure 63: The NPV settings tab

June 2016 42
PGCA User Manual V2.1.0.3

Figure 64: Completion of footprint creation

In the View Footprints tab, the user can view, modify and save out footprints for the scheduling process.
As shown in Figure 65, each of the footprints generated is listed with associated relevant information.
By clicking on a line in the data shown, the user can view the particular footprint which will be shown
in the screen on the right. By right clicking in this screen, the user can edit drawpoints for inclusion (as
shown previously in Figure 73). Once happy with a footprint, the user can check the “Use” box to save
relevant footprints out to the scheduling stage. The format of a saved footprint file is shown in Figure
66, and contains drawpoint name, X, Y, Z, width of draw, tonnes, and NPV. At this point the user
should move to the Scheduling Parameters tab (Figure 67).

The Scheduling Parameters tab allows the user to select variations on production rate to be applied to
each of the footprints. An automatic sequence will be generated for each footprint starting at the
drawpoint containing the highest metal value. Batch mode will place all of the modelling runs in the
selected output file, and also generate an NPVSummary.csv file listing the results. This information
can be used to manually fine tune the footprints selected and run different sequencing and scheduling
options from the Scheduling menu.

June 2016 43
PGCA User Manual V2.1.0.3

Figure 65: The View Footprints tab

Figure 66: Example of a saved footprint file.

June 2016 44
PGCA User Manual V2.1.0.3

Figure 67: The Scheduling Parameters tab

June 2016 45
PGCA User Manual V2.1.0.3

The Scheduling Menu

The Scheduling Menu can be used before a project has been opened, to allow the user to create a
project to be run later. The options under this menu are shown in Figure 68.

Figure 68: Options available under the Scheduling Menu

Footprint Generator
The “Footprint Generator” menu item allows the user to create a block caving drawpoints file from
scratch. When selected it produces a form as shown in Figure 69.

Figure 69: The “Generate a Block Caving Footprint” form

Select the “File” menu in the top hand corner, and choose “Create” to begin (Figure 70).

June 2016 46
PGCA User Manual V2.1.0.3

Figure 70: Options under the “Create” submenu

At this stage, there will be a choice between generating a footprint directly from the block model, and
importing a footprint from the third party “Cave Planner” software.

Footprint From Block Model

If “Footprint From Block Model” is selected, the user will be prompted to open an appropriately
formatted block model file. Once a block model is selected, a form like that shown below will appear
(Figure 71). The user can select a layout type, and specify parameters relative to that layout. Once this
has been done, select “Generate Dummy Drawpoint Layout”. This will produce a preliminary overlay
of drawpoints. If the general layout is considered acceptable, the user can move to the “Level
Placement” tab.

Figure 71: Setting up a layout in the Footprint Generator

Selection of the level placement tab will bring up the form shown below (Figure 72). The user can
select the level RL by moving the “Lift Settings” slider, or entering the value in the box shown.

June 2016 47
PGCA User Manual V2.1.0.3

Figure 72: Locating footprint RL in the Footprint Generator

Once a level for the footprint has been selected the user can move to the “Footprint Generator settings”
tab, which will produce a form with the layout overlay in plan view. The user can then input the
following parameters:

 Minimum Grade: The minimum grade that a drawpoint must produce;

 Minimum Assigned Tonnes: The minimum tonnes that each drawpoint must produce at the
total minimum grade.

Once these parameters have been chosen, the user can select “Run Footprint Generator”. This may take
a short time to process, after which a preliminary footprint will appear on the screen. The user can then
use the mouse (hold down left click) to create a window for selecting drawpoints to modify, and right
click for a menu of modification actions (Figure 73). When manual editing is complete, the user can
select “Save Drawpoint Locations” from the file menu. This will save a space delimited drawpoints
file which can be imported directly into the scheduling process.

June 2016 48
PGCA User Manual V2.1.0.3

Figure 73: Manually modifying a footprint

Footprint From Cave Planner

If “Footprint From Cave Planner” is selected from the File menu file (Figure 74), the user will be
prompted to open an appropriately formatted cave planner file. This must be a comma delimited .csv
file.

Figure 74: Importing a footprint from third party cave planner software

When the file has been selected, the user can select the desired layout parameters (Layout Type and
Geometry Parameters). Once these have been specified, the user can select “Extract Schedule From
Cave Planner File”. This will produce a form as shown below (Figure 75).

June 2016 49
PGCA User Manual V2.1.0.3

Figure 75: Footprint generated from Cave Planner file

This indicates the tonnes assigned to each drawpoint, and the layout of these drawpoints as defined in
the Layout tab on the left of the screen. These user can then modify these parameters and re-select the
“Extract Schedule From Cave Planner File” option to show the layout on the screen. Once the layout
is as desired, the user can select “Save Drawpoint Locations” from the file menu. This will save a space
delimited drawpoints file which can be imported directly into the scheduling process.

Create Block Caving Schedule

Once a footprint has been generated, a schedule can be created. To do this, select "Create BC Schedule"
from the scheduling menu.

When the “Create BC Schedule” option is selected, the form showed in Figure 76 will appear. The
message box appears to prompt the user to open the File menu (circled) and import a drawpoint file
(Figure 77). This is required to begin the scheduling process. The drawpoints file must be correctly
formatted as previous versions of the software.

June 2016 50
PGCA User Manual V2.1.0.3

Figure 76: Opening screen for block caving schedule creation

Figure 77: Loading a Drawpoint File

Sequencing
Once a drawpoints file has been loaded, a form like that shown in Figure 78 will appear, showing the
layout of the drawpoints loaded from the drawpoint file. In this form the user can either ask the program
to create a sequence, or load a sequence which has been manually created. If the program is used to
create a sequence, three inputs are required.

1. The dropdown menu is used to select a starting drawpoint from the drawpoints contained in the
drawpoint file.
2. The front angle is used to indicate the direction of advance (for instance 90 degrees has been
used for the sequence illustrated in Figure 78).
3. The X to Y growth ratio determines the relative growth rate of the fronts (illustrated in Figure
79).

The refresh button is used to generate the sequence, which is contained in a file called “Generated
Sequence.csv” located in the directory from which the drawpoints were originally selectedd. This is in
the default format required for sequences used in later scheduling steps.

More complex automated sequencing capacities may be addressed in later versions, however if the user
wishes to use a self generated sequence, this is catered for in the “I have my own sequence option”.
Once a sequence has been loaded or created, the “View Sequence” slider can be used to visually check

June 2016 51
PGCA User Manual V2.1.0.3

that the sequence is appropriate before proceeding to the next stage, generation of tonnages to be
extracted from each drawpoint.

Figure 78: Sequencing a schedule

June 2016 52
PGCA User Manual V2.1.0.3

Figure 79: Use of the X to Y Growth Ratio in the Block Caving Scheduler

Tonnage Generation
The Tonnages form is illustrated in Figure 80. In this form the program can be asked to generate
tonnages for extraction, or a manually generated tonnage file can be used. If the program is asked to
generate tonnages, two options are available. The first is a rapid first estimate scheduler which
completes tonnage etimation in a matter of seconds (DMM), the second is a tonnage generator based
on PGCA flow modelling. Before selecting a tonnage generation engine, a block model must first be
loaded (marked 1 in Figure 80). On loading a block model a popup window will indicate some data
mainly relevant to the DMM engine.

The user must next select a tonnage generation engine. If DMM is selected (marked 2), the program
will use the following settings to generate tonnages. The first three of these are common settings which
also apply to the PGCA engine:

 Shutoff Grade: the program will read the first grade in the block model as the basis of this
shutoff;
 Min Draw Height: all drawpoints will be required to draw to at least this height;
 Max Waste Drawthrough: The greatest amount of waste that a drawpoint can draw through to
carry in ore above. This setting is intended to allow the user to partially manage risk related
ore ore high above the drawpoints which has a smaller chance of reaching the drawpoints than
ore immediately above the drawpoints;
 Draw width

Figure 80: Generating drawpoint tonnages

When the DMM engine has completed a tonnage estimate, a window such as that shown in Figure 81
will appear.

June 2016 53
PGCA User Manual V2.1.0.3

Figure 81: DMM tonnage estimate notification

Once this window is dismissed, the screen will update showing tonnages estimated for each drawpoint
to give an indication of where levels of draw on the footprint are highest and lowest. This screen can
also be refreshed manually

Three files will have been output to the directory from which the block model was selected:

1. “DPGrades.csv” showing forecast grades (for the shutoff grade field only) by drawpoint by
period. These are not scheduled yet, and are only relative to the starting point for each
individual drawpoint;
2. “tonnages.csv” showing tonnes and grade for the drawpoints. This is automatically selected as
the default input file for the scheduling phase;
3. “HeightOfDraw.csv” showing height of draw in stages for each drawpoint relative to its starting
point.

If the PGCA engine is used to generate tonnages (marked 3 in Figure 80, the three common settings
indicated above will be used, and in addition, two additional settings are required.

1. Block resolution: selected from a dropdown menu as in regular PGCA models (see Figure 16);
2. Probability setting: related to width of draw as for regular PGCA models (see discussion in
Appendix D)

When the PGCA tonnage generation engine is selected, a window such as that shown in Figure 82 will
appear showing the estimated time required to complete the estimation. This is calculated based on
model size and schedule length and will be longer than that required for the DMM engine. The
advantage of using the PGCA tonnage estimation is that it is currently more efficient than the DMM
tonnage estimate (i.e. sometimes finds more of the available tonnes), and will eventually allow the user
to estimate tonnages using most of the modelling factors which PGCA allows, and which DMM (a more
high level estimator) is not designed to handle.

Figure 82: Estimate of time required for PGCA tonnage engine

June 2016 54
PGCA User Manual V2.1.0.3

If “Yes” is selected in the form shown, a PGCA window will open behind the tonnage generation
window, and run in the background, with only the Simulation Progress bar showing (Figure 83).

Figure 83: Example of PGCA tonnage generator running behind the tonnage form

When the estimation is complete, a window showing the estimated tonnage total will appear (as in
Figure 81) and the user will be prompted to refresh the screen to show the tonnage estimate on the
screen. The “DPgrades .csv” and “tonnage.csv” files will be available in the root directory. The
scheduling phase is the next step in the work flow. If the PGCA tonnage generator has been used in
the scheduling process, it will be necessary close the program and reopen it before running the
schedule in PGCA.

Scheduling
When the scheduling tab is selected, the form shown in Figure 84 will appear. The visualisation on
the right side of the form will appear as a blank in this stage of the process. To generate a schedule,
the following parameters must be identified:

 Start date: Selected from a dropdown menu;

 A Drawpoint Ramp Up File (e.g. Figure 85), indicating the rampup profile to be used for all
drawpoints. This file must have the drawpoint names in the first column (in the same order as
the drawpoint file), and the desired extraction rate by period in the subsequent columns);
 Option to Limit Production ramp up (check box): - if checked, this will allow the user to input
a file which constrains the rate at which the mine can ramp up. This is sometimes needed when
the production rate allowed by the ramp up file exceeds the rate at which the mine is expected
to be able to ramp up as a whole. It is formatted as a list of values representing production rate
by period (Figure 86);
 Maximum Monthly Mine Production Tonnes: The constraint for individual drawpoint
production rate;

June 2016 55
PGCA User Manual V2.1.0.3

 Drawbell Development Rate – Per Month: The rate at which drawbells can be opened for
production (Figure 86);
 Monthly Development Rate From File: Allow the user to select a file containing a variable
production rampup rate

Figure 84: The scheduling form

Figure 85: Example of a Drawpoint Ramp Up file

Once these parameters have been input, selection of “Generate Schedule” will complete the process. A
schedule formatted for use in PGCA or DMM models will be generated. The schedule can be viewed
in the right hand window of the scheduling screen. By scrolling through the schedule using the slider
bar, the user is able to see the drawpoints which are open in the schedule at any particular period. Active
drawpoints are highlighted.

June 2016 56
PGCA User Manual V2.1.0.3

Figure 86: Example of files to limit production rampup (left) and for variable drawbell opening rate (right).

At present this scheduler does not account for undercutting. To allow for this, the user must take the
final schedule and remove the tonnes and metal attributable to the undercut, and redistribute them in
the output schedule according the undercut schedule. Allowance for undercutting in the model is
planned for development in later versions of the software.

Combine Lift Schedules

The "Combine Lift Schedules" command is under development and is greyed out.

Create SLC Schedule

The first item available in the scheduling menu is “Create SLC Schedule”. This allows the user to
generate an SLC schedule. The form for this function is shown in Figure 87. This is designed to allow
scheduling out of tonnages generated from running life of mine PGCA forecasting models in shutoff
mode.

Two types of input files are required for the scheduler. These are a drawpoints file and a tonnage file
(neither of which contain header rows). The drawpoints file contains columns for Ring Name, X,
coordinate, Y coordinate, Z coordinate, Ring type, and Dependency. The dependency column shows a
ring name for a precursor ring which must be fired before this ring can be scheduled (if such a precursor
ring exists). This is often used to ensure that slot and other irregularly oriented rings are fired in the
correct sequence.

Other input production constraints are:

 Start Date for the schedule;

 Starting ring for the schedule;
 Front Orientation;
 Level Lag;
 Maximum Ring Production Tonnes (the maximum tonnage produced from a ring in a 24 hour
period);
 Maximum Mine Daily Production Tonnes;
 Minimum Ring Turnaround (Days);
 Minimum Slot Turnaround (Days);
 Max Daily Rings Fired.

June 2016 57
PGCA User Manual V2.1.0.3

Once created, the schedule can be viewed using the Schedule Visualiser (described further below) and
directly imported to be run in PGCA models.

Figure 87: The SLC scheduling form

June 2016 58
PGCA User Manual V2.1.0.3

Figure 88: Example of a drawpoints file for the Create SLC Schedule function

The tonnage file (Figure 89) contains a list of ring names and the tonnes to be drawn from them.

Figure 89: Example of an SLC scheduler tonnage file

Visualise SLC Schedule

The "Visualise SLC Schedule" command is designed to allow the user to visualize a PGCA formatted
SLC production schedule. The File menu in the visualiser allows the user to open any correctly
formatted schedule, and the view menu allows views looking north, east or in plan to be selected. The
slider bar at the bottom of the screen allows the user to move to a desired date in the schedule. Rings
that have been produced up to that date are shown in red, and rings which are yet to be produced are
shown in blue.

June 2016 59
PGCA User Manual V2.1.0.3

Figure 90: Example of schedule viewed in the SLC Visualiser looking East (left) and in plan view (right)

Zoned SLC Schedule

This function is designed to modify the results of a shutoff schedule from an SLC modelling run to
produce an evenly distributed tonnage based schedule. When the function is activated, the form shown
in Figure 91 will appear. A schedule file, matching “RecoveryByRing.csv” and
“SLCRingCompletion.csv” files should be selected. The search radius value selected will determine
how far each ring looks when calculating its new production tonnage. A large search radius value will
generate a smoother production tonnage gradient across levels, but this will also produce higher tonnage
variation in individual rings from their value in the original shutoff schedule.

Figure 91: Form for the Zoned SLC Schedule function

When the Create button is selected, Production tonnages will be smoother across individual levels, but
not vertically between levels. Rings in a schedule with fixed tonnages (i.e. not shutoff rings) will
retain their original production tonnages and rings which have shut down due to voids (i.e. opened to
surface) will not have their production tonnages zoned. Two output files will be generated in directory

June 2016 60
PGCA User Manual V2.1.0.3

from which the un-zoned schedule was selected. These will be “Unzoned Combined Schedule.csv”, a
file containing all of the rings with no zoning of production, but containing constrained tonnages
where applicable, and “Zoned Tonnage File.csv”, containing all rings with constrained and zoned
tonnages where applicable.

June 2016 61
PGCA User Manual V2.1.0.3

The Tools Menu

The Tools Menu allows the options shown in Figure 92.

Figure 92: Options available in the Tools Menu

Advanced Flow Settings

The Advanced Probability Settings sub-menu contains three sub-commands which allow the user to
tailor flow behaviours in more detail (Figure 93)

Figure 93: The Advanced Flow Settings command

Angled Flow
The Angled Flow command produces the form shown in Figure 94. This allows the user to bias flow
by changing the individual probability values for movement for any of the blocks around each block.
For example, to bias flow to the north, probabilities for the blocks in the top line of the Top Layer would
be increased relative to the other probabilities in this layer. The Material 21 tab is greyed out as it not
ready for use. It is being developed to allow different biases in the model controlled by denoting them
as material type 21.

June 2016 62
PGCA User Manual V2.1.0.3

Figure 94: Advanced Probability Form

Fragmentation Settings
The Fragmentation Settings command allows the user the option to set a link between a Fragmentation
field in the block model, and draw widths. When this option is selected the form shown in Figure 95
will appear. The form will be populated based on the Fragmentation field in the block model (as long
as the user has defined one). It will set a series of fragmentation ranges for which the user can input
specific draw widths and mobility factors. The “Min Size” field will automatically update based on the
values the user places in the “Max Size” field.

June 2016 63
PGCA User Manual V2.1.0.3

Figure 95: The Fragmentation Draw Width Form

If a fragmentation field is to be used in the block model, the mean fragment size value should be used,
as this is the size which has the greatest influence on width of draw. Mean fragmentation at the
drawpoints will be reported in the “Recovery.csv” file and in the “PGCA - GradeByDrawpoint –
Fragmentation.csv” file which can be generated using the “Generate Recovery Data for Drawpoints”
function under the Calculate toolbar. The value reported will not account for secondary fragmentation
however, and will only represent the mean primary particle size based on the inputs provided in the
block model. A capability to estimate secondary fragmentation in the modelling cycle and generate
drawpoint by drawpoint assessments in the form of full fragmentation distributions for each drawpoint
is not yet a possibility.

Mobility Settings
The Mobility Settings command will bring up the form shown in Figure 96. This allows the user to set
width of draw and mobility factors for each individual rock type shown in the “RockType” field, if this
field is included in the block model. This will enable rock types with different material properties to
behave in different ways in the modelling process.

June 2016 64
PGCA User Manual V2.1.0.3

Figure 96: Example of a filled Rock Type Draw Width form

The check box at the bottom of the form will allow the user to define whether the model assesses an
average of all of the surrounding blocks to define the draw width and mobility, or just the block
immediately above the void being filled by movement.

The test model shown in Figure 97 contains a fault (diagonal) with a mobility factor set at 3. The only
difference between the two models shown is that the check box has not been ticked in the model on the
left. This results in the effect of the fault being more pronounced (as it takes only one fault block
immediately above the void to set the mobility at 3). This widens the draw on the intersection with the
fault. On the right, the effect of the relatively small amount of fault blocks is overwhelmed by the
influence of the surrounding blocks, resulting in the draw zone flowing through the fault, and retaining
the general draw width defined by the predominating rock type surrounding the fault.

June 2016 65
PGCA User Manual V2.1.0.3

Figure 97: Example of the effect of checking the box for use of an average draw width based on surrounding blocks

Marker Trial Data

The Marker Trial Data command will activate the form shown below

Figure 98: The Marker Trial Data Form

This command is designed to allow a mine to calibrate a PGCA model against marker trial results.
Three input files are required. The project and extracted block files are standard PGCA files. The
Marker data file must be specifically formatted as shown in Figure 127.

Two output files will be generated in the location of the project file being assessed: “summary
markerdata.csv” and "detailed markerdata.csv". The summary file contains general information such
as percentages recovered in each recovery class in model and mine (e.g. primary, secondary and tertiary
recovery classes), as well as average draw width and height. The detailed file contains information on
recovery status for every marker both in the mine and in the model.

June 2016 66
PGCA User Manual V2.1.0.3

Batch Mode
The Batch Mode command will bring up the form shown in Figure 99. Batch mode is designed to allow
the user to set up multiple PGCA or DMM runs at one time. These runs will then be completed
automatically, with results files deposited in automatically generated output folders.

Figure 99: The Batch Mode form

The Project Properties tab requires the following inputs:

 A DMM or PGCA project file which has been previously created;

 A specifically formatted base parameters file (format shown in Figure 100);
 A designated output directory in which to deposit batch modelling results;
 Check the box marked “Save model state at the end of each run” if you want to see the model
state at the end of the run. Model states can be multiple gigabytes in size, so use of this check
box is not generally recommended.

The base parameters file is a space delimited text formatted file which reads in the basic parameters for
a PGCA or DMM run. A DMM SLC or PGCA project will require a draw width in the first line, while
a PCCA block caving model will require a probability factor. The value for ShutoffType (second last
line) determines the shutoff mode (0 for Instantaneous and 1 for Sample Average). Specific project
types will extract the appropriate information from this file and ignore irrelevant information.

June 2016 67
PGCA User Manual V2.1.0.3

Figure 100: Example of a correctly formatted PGCA Base Parameters file

Once the Project Properties inputs have been completed, the user can then choose the tab corresponding
to the type of batch run that they would like to complete. There are three types if batch modelling run
supported. These are:

 A parameter sweep, for generation of sensitivity models on a range of different parameters;

 A multiple schedule run, for modelling of different schedules for the same set of parameters;
 A multiple parameter file run, for modelling of individually created base parameter files.

The Parameter Sweep tab is shown in Figure 101.

Figure 101: The Parameter Sweep tab

June 2016 68
PGCA User Manual V2.1.0.3

First select the type of variable that is required from the drop down menu (Flow Width is shown here).
Click on the Add button and a row will be added to the list. In this row, input the starting value, the
increment and the number of realizations required. In this example widths of draw at 12m, 14m, 16m
and 18m will be run. Multiple variables can be run in the same batch run. Each item in the batch run
list will reference the Base Parameters File for all input parameters other than the variable shown in this
form. Once all items to be modelling have been added, click on the Run All Parameters button to start
the model.

Results for each modelling run will be output in separate sub directories under the Base Output
Directory selected at the beginning of the process. For PGCA models, these will include the following:

 Recovery.csv and RecoveryByDrawpoint.csv or RecoveryByRing.csv files;

 An ebf file;
 A copy of the relevant Base Parameters File for reference;
 If selected, a model state which can be opened in the PGCA project to view the location of cells
at the end of the modelling run.

For DMM models outputs will include:

 Recovery.csv file;
 GradeByDrawpoint.csv and MetalByDrawpoint.csv files for each grade element;
 A copy of the relevant Base Parameters File for reference.

The Multiple Schedules tab is shown in Figure 102. To add a schedule to the list, click on the Add
button and navigate to the schedule to be run. Once all schedules have been added, click on the Run
All Schedules button to start the batch modelling process. The Multiple Parameters tab works in the
same way, using different Base Parameters files as selected by the user.

June 2016 69
PGCA User Manual V2.1.0.3

Figure 102: The Multiple Schedules tab

Design
The Design sub-menu is intended to contain commands which will help the user in creation of PGCA
projects. At present it contains two items (Figure 103).

Figure 103: Contents of the Design sub-menu

Convert Strings to Rings

This function is designed to automatically generate SLC rings from a string file. This means that once
strings for production drives have been generated, the rings for these drives can be set up quickly and

June 2016 70
PGCA User Manual V2.1.0.3

easily. The form for this function is shown in Figure 104.

Figure 104: Form for the Convert Strings to Rings function

The string file to be imported should be a comma delimited csv file in the general form shown in Figure
105. The specific names in the header can be changed, however the values for X, Y, and Z must be in
columns 3, 4 and 5, and the level must be in the 10th column.

Figure 105: General format of the string file from which rings are to be generated

An example of the output from this function can be seen in Figure 106. This can be used directly as a
development file in an SLC schedule, and also as the basis of a schedule (once dates, tonnages and ring
types have been added to the ring locations). It should be noted that the ring Z coordinates will match
the string Z coordinates, so if the strings are placed on the floor of the drives, it will be necessary to
move them up to the backs of the drives before initiating the function.

June 2016 71
PGCA User Manual V2.1.0.3

Figure 106: Example of output of the Strings to rings function

String Mode
The String Mode sub-menu is designed to allow the user to directly generate strings by drawing them
on the screen. This is so that drives can be created following the profile of the orebody and then
modified later in excel if required. At present this sub-menu is under development and is greyed out,
however a brief description of the functionality being developed follows.

Start String
Start string allows the user to click on the screen to start drawing a string. An example of where this
option might be required is shown in Figure 107 where the outline of the orebody is not uniform and a
rough guide is required by drawing strings by eye.

June 2016 72
PGCA User Manual V2.1.0.3

Figure 107: Example of use of the start string function

Write Strings
Write strings allows the user to output the strings drawn on the screen to a file.

Figure 108: Example output of the write string function

June 2016 73
PGCA User Manual V2.1.0.3

DXF
The dxf menu contains a number of functions for manipulation and generation of dxf files (Figure
109)

Figure 109: Functions available under the DXF submenu

Translate DXF
The Translate DXF command is designed to translate the coordinates of a DXF file to synchonise it
with the coordinates being used in the model. By entering the desired shift in coordinates, checking the
box and hitting the save button (Figure 110), the translation parameters will be saved by the program.
Any DXF which is then loaded into the model during that modeling run then will be translated by these
coordinates automatically.

Figure 110: Translate DXF form

Generate DXF from SLC development

The Generate DXF from SLC development command will generate a dxf file showing the location of
development rings for an SLC project. This can be opened for viewing with various software packages
for viewing in context with other pre-existing objects such as stopes or pit shells for design purposes.
Before executing the command, the SLC project on which it is to act must be opened. When the
command is executed a dialogue will be opened asking the user to name the output file. After naming
the output file, select Save and the file will be output in the selected directory.

June 2016 74
PGCA User Manual V2.1.0.3

Figure 111: Example of dxf file of development output from PGCA

Generate DXF with Rings from SLC Schedule

This command enables to user to show the SLC rings as well as the development in a dxf file. The
source SLC project should be opened first. When the command is executed a dialogue will be opened
asking the user to name the output file. After naming the output file, select Save and the file will be
output in the selected directory. Note that only the lower half of the rings are shown. This allows
undercut rings to be shown as half rings (as they are usually fired), but still gives the impression of full
SLC rings within the centre of the production levels.

June 2016 75
PGCA User Manual V2.1.0.3

Figure 112: Example of dxf output showing rings and development

Generate Caveback Shapes from Schedule

This command allows the user to generate a rough approximation of cave back propagation based on a
production schedule. These cave shapes are only indicative as they do not make reference to rock mass
properties. They are calculated using estimates of swell and the draw schedule only. Therefore they
should be used only for high level estimations.

The source project should be opened before executing the command. The program will determine if an
SLC or block caving project is open and generate a form appropriate to the mining method. An example

June 2016 76
PGCA User Manual V2.1.0.3

of this form is shown in Figure 113. For SLC models, a period is one day in the schedule, while for
block caving models, time allocated for a production period in the schedule is undefined (can be a
month, quarter or year).

The user should select an output folder which should be reserved for use of only the dxfs being
generated. If creating a folder through PGCA, make sure to click outside the text box naming the folder
before continuing. A swell factor should be input, along with the period interval between the shapes
being generated. By hovering the cursor over the Period Interval box, the user can find out how many
periods the schedule is due to run for, and determine an appropriate interval accordingly. Once the
input information has been entered, select “Generate” and a message indicating that the cave shapes
have been generated will appear shortly afterwards.

Figure 113: The generate Caveback Shapes from Block Caving Schedule Input form

The cave shape dxf files can be viewed in visualization software for comparison with surrounding
structures and shapes (e.g. Figure 114). When the cave shapes have been generated, the “Use Cave
Shapes” box at the bottom right of the screen will automatically be checked. When run, the program
will automatically use the directory in which you have placed the cave shapes when running the model
(Figure 115). If use of cave shapes is not required in the modelling run, the cave shape box can be un-
checked before running.

Note that for a schedule in which cave shapes are generated at 5 period intervals, the first cave shape
will be named 0.dxf (indicating to PGCA that it should be opened at period 0), but it will show the
estimated propagation of the cave for rock scheduled to be drawn from periods 1 through to 4. The
next shape 5.dxf will open in period 5, and represent all rock scheduled for draw up to the end of period
9. The example cave shape shown in Figure 114 is shown in Figure 115 being used in the PGCA model
from which it was generated.

June 2016 77
PGCA User Manual V2.1.0.3

Figure 114: Example of cave shape generated from block caving schedule

Figure 115: Example of cave shape being used after generation from schedule

June 2016 78
PGCA User Manual V2.1.0.3

Inspect Drawpoints
The Inspect drawpoints tool allows the user to identify the name of a drawpoint in a footprint by
hovering the mouse over it in plan view. When Inspect Drawpoints is selected, a cursor inspector
window will open at the bottom of the screen showing the name of the closest drawpoint or SLC ring
to the cursor, and the distance this drawpoint is from the cursor.

Figure 116: Identifying drawpoints on the screen using the cursor inspector

SLC Blast to Ore

This function is designed to allow hanging walls in sub-vertical SLC mines to fail after the production
level has passed below the current level. To activate this function, click the function in the menu and a
tick will appear next to it indicating that it is active. This must be done before starting the project run.

Figure 117 SLC Blast to Ore activated

In order for this option to work, cells in the block model which are to be treated as a solid hangingwall
must be designated material type 3 in the block model. When each ring is fired, any blocks on the edge
of the ring which are material type 3 will remain solid and unmovable until production on that level is
complete

Figure 118: Block model showing material type 3.

In operation, this will produce an outcome such as that shown in Figure 119, with hangingwall failure
delayed. This is in keeping with experience at some SLC mines with sub-vertical hanging walls, and
in some cases produces better calibration results against mine recovery data.

June 2016 79
PGCA User Manual V2.1.0.3

Figure 119: Example of model with SLC blast to ore operating.

June 2016 80
PGCA User Manual V2.1.0.3

Appendix A - Formatting of input files

Several types of input files are required for PGCA and DMM. Generally these files must follow a
defined format (usually space delimited text). These are as follows:

Schedule format for SLC

The Schedule format for SLC is as shown in Figure 120. This comprises a date for schedule start in the
top row, followed by columns containing space delimited text containing the fields Ring ID, Ring X,
RingY, Ring Z, Date Fired, Tonnes Drawn, Ring Dump Angle (0=vertical)., Ring Dump Direction, and
Ring Type. A Ring Dump Direction value of 10 degrees will move the top of the ring forward by 10
degrees in the direction specified by the Ring Dump Direction field. The date can be in the following
formats for the example date 1st April 2014:

 1/4/2014;
 01/04/2014;
 4/1/2014;
 04/01/2014

Input date formats must be internally consistent for individual schedules. All PGCA outputs will be in
the format 01/04/2014.

A header row is not used. All rings must be located within the bounds of the block model or an error
will result. For reporting purposes, ring names should separate level, cross cut and ring using # marks.

An additional row for ring production duration can also be added after the ring name field. If added,
this will distribute the production in that ring evenly over the number of days specified. If this field is
omitted, the program will assume a default value of 1, and draw all of the tonnage shown on the day
the ring is blasted, as in earlier versions of the program i.e. schedules from previous versions can still
be run. If used, ring duration values must be greater than 0.

Figure 120: Schedule file format for SLC models

Schedule format for block caving

The schedule format for block caving shows drawpoint names in the first column, and production for
each modelling period in succeeding columns (Figure 121). No header lines are required. The order of
drawpoints in this file should be the same as in the drawpoint file.

June 2016 81
PGCA User Manual V2.1.0.3

Figure 121: Schedule file format for block caving models

Block model format

The block model file format for both SLC and block caving models is shown in Figure 122. This
comprises a header line which contains the header titles “X”, “Y”, “Z”, “Density”, “Material”, and any
grade/rock type fields. The “Material” field should contain a number between 1 and 21. Each of these
numbers shows a different colour when the “Colour by Material” box is checked in the simulation
interface. This box can be toggled on and off while a simulation is in progress. The remaining fields
contain grades or rock types, and should be numerical, with the exception of an optional “RockType”
field which can be alpha-numeric. Non-grade fields included in the block model will show results that
are not meaningful in the reports.

Material types 19-21 are special fields which affect the flow of rock in the model. Material 19 denotes
fines, material 20 solids, and material 21 alternate flow bias properties, so these field types should not
be used unless the specific property of the field is required.

Figure 122: Block Model file format

Drawpoints file for block caving

A "Drawpoints" file is only required for block caving models. The format for this file is shown in
Figure 123. This consists of a column for drawpoint name (in the same order as the schedule file), the
the Y,Y, and Z coordinates of the centre of the drawpoint brow, and a width of draw column (which is
only referenced in DMM models if different widths of draw are required for different drawpoints. If
these values are missing, the DMM will use the width value shown in the Simulation Properties

June 2016 82
PGCA User Manual V2.1.0.3

interface). No header row is required. All drawpoint locations in drawpoint files must be located within
the bounds of the block model, or an error will result.

Figure 123: Drawpoints file format for block caving

Development file for SLC

A Development file is only required for SLC models. Its format is shown in format in Figure 124. This
comprises columns containing a Development task ID, X, Y, Z coordinates for this task, and the
orientation of the development. The Z coordinates for the development file should correspond to the
backs of the drive. The program will use this information to create standard 5x5 drives in the model.
When reporting, the model will produce a “DevelopmentRecovery.csv” report which gives the tonnes
and grade extracted by each development file record. However as development scheduling is not
currently supported, the dates when this development is extracted will be unspecified.

Figure 124: Development file format

Ring type file for SLC

A Ring Type file is only required for SLC. Its format is shown in Figure 125. This file requires a
header row containing the titles shown in line1 of Figure 125. On subsequent lines parameters for
different types of ring are recorded. Parameters shown in columns B to D are related to ring geometry.
Parameters shown in columns F to J are related to individual shutoff settings for each ring type. If the
parameters for the individual shutoff settings are absent, the model will default to the settings in the
user interface for that ring type. A limit of 100 ring types is applicable for this file.

Figure 125: Ring file format

June 2016 83
PGCA User Manual V2.1.0.3

Undercut file for Block Caving

An undercut file is only required for Block Caving models. Its format is shown in Figure 126. This
file requires an ID field, X, Y, Z coordinates for the collars of the undercut rings, and undercut height,
and the proportion of each undercut ring to be drawn.

Figure 126: Undercut file format

Marker data input file

A marker data input file is required for the Marker Trial Data command. It contains information on
Marker installation locations in mine coordinates as well as Recovery information. The file must be
space delimited and have the following information in rows: Marker ID, Ring Name, Marker Plane,
Marker Hole Number, X, Y, Z, Date Recovered and recovery class (as shown below). Recovery
categories need to be as follows: “Primary”, “Secondary”, “Tertiary”, “Quaternary”, “Above4”, or
“Unrecovered”.

Figure 127: Marker data input file format

June 2016 84
PGCA User Manual V2.1.0.3

Appendix B - Model Calibration

Model calibration is an important process if PGCA is to be used to forecast metal recovery for mines
which already have a production history, and is solely the responsibility of the licensed model user. It
is the user's responsibility to ensure that their models calibrate effectively before using the model
outputs, or failing availability of calibration data, benchmarking against existing operations where
similar mining practices are in use.

A useful practice to follow to calibrate the model follows.

1. Gather historical ring production data, and mill reconciliation data. Compare tonnages
extracted in both datasets, and where possible, reconcile the two data sets;

2. Run the PGCA model and compare the results against the mill reconciliation data, varying flow
width until model calibrates. To achieve a wider flow width in SLC models, the actual draw width can
be input. Projects should be recalibrated whenever a new block size is used (because of mesh effects
discussed in Appendix D), or when historical reconciliation or input data changes. It is also best to re-
run a calibration case when a model update is received.

For indicative purposes, Figure 128 shows the relationship between width of draw and draw width
factor in a single drawpoint block caving model, for a 100m high extraction zone with a cell size of
1.25m. Early results from a single ring SLC model indicate that a similar width of draw relationship
exists for SLC. In any case, selection of width of draw factor for the calibration process can be guided
more by the value required to make the model calibrate, than a fixed requirement to obtain a given width
of draw in the model. If expected width of draw does not produce the best calibration results, further
basis for estimation of width of draw (e.g. marker trials) are useful.

Probability Factor vs Width of Draw for

40 100m height of draw in block caving
model
30
Width of draw (m)

20

10

0
0 0.2 0.4 0.6 0.8 1
Probability Factor y = 10.497x-0.327
Width R² = 0.9978

Figure 128: Relationship between width of draw and probability factor in PGCA for an 100m high extraction zone

By running the model at a range of draw width settings, it should be possible to bracket the historical
production performance of the mine, and zero in on the most appropriate draw width settings to use in
the model. The modeller should not expect to achieve parity against reconciliation on a ring by ring, or
day by day basis. This is because too many sources of variation exist in production performance to be

June 2016 85
PGCA User Manual V2.1.0.3

accounted for at such a small scale in the model. However, over longer periods, it should be possible
to achieve an acceptable calibration result.

Although there is short term variation in the calibration results, total metal extracted over the calibration
period should be equivalent to that recognised by the mill (as the standard agreed as the basis for
calibration). This is important because from a forecasting perspective, the modeller wants to be as
certain as possible that the right amount of metal is available in the mine for extraction for future
production.

When forecasting, it is important to recognise that PGCA can only estimate based on the production
schedule provided. It cannot estimate which rings will not be fired due to poor ground conditions,
equipment breakdowns etc. Therefore factors for these types of delays may need to be accounted for
after the schedule is produced.

June 2016 86
PGCA User Manual V2.1.0.3

Appendix C - Formatting of DXF Files

DXF files come in many different sub-formats, and PGCA only recognises a small subset of these at
present. DMM does not support use of DXF files. To check whether a DXF if likely to open in PGCA,
open the file in notepad and search for the terms "VERTEX" or "3DFACE". These are terms that PGCA
looks for in extracting data from these files. Figure 129 gives an example of DXF files with these terms
shown which can be opened by PGCA. Work on opening more forms of DXF file is ongoing.

Figure 129: DXF formats that PGCA reads

June 2016 87
PGCA User Manual V2.1.0.3

Appendix D - Effect of Model Resolution (cell size) on width of draw in

PGCA
In block caving models, a width of draw value at 60,000t drawn (about 100m high using with an average
density of 2.7) is used. As PGCA generates flow shapes which grow in width with growth in height, a
defined height of draw or tonnage is required before a width of draw can be specified. The width of
draw setting has been calibrated from test modelling carried out at different cell sizes, relating
probability of block movement in the model to width of draw for these different cell sizes. Figure 130
below shows an example of test model results plotted in cross section. The overlain ellipsoid shape
indicates the 99th percentile of blocks recovered for this test case. The 99th percentile case is used so
that outlier blocks will not artificially inflate the width of draw taken from test. The X and Y scales
relate to block location rather than a distance. As this case shows a 2m cell size, the values on the axes
would need to be multiplied by 2 to give a value in metres.

Figure 130: Example of model run to relate width of draw and probability factor for PGCA block caving drawpoint

June 2016 88
PGCA User Manual V2.1.0.3

A series of models were run for 1.25m, 2m, 2.5m and 5m cell sizes to generate relationships for each.
Figure 131 shows an example of the relationship generated for models run at the 2m block size. These
relationships have been coded into the software so that the program automatically detects the cell size
at which a model has been created, and then generates the appropriate probabilities of block movement
to match the width of draw setting selected by the user in the input screen.

Figure 131: Example of test modelling results and the fitted equations generated from these results

Because PGCA is cell based, changes in cell size can produce changes in results. For this reason a
model run at a certain draw width at one cell size will be slightly different from the same model run
with the same draw width at another cell size. This is because there are rounding effects associated
with the use of larger block sizes, and also because results change slightly in every model run due to
the probabilistic nature of the algorithms used in PGCA. For this reason it is best to use the same cell
size for all model runs when generating results for a specific mine. Once a model has been calibrated
at this cell size, it should be understood that recalibration will be required if the user changes to an
alternative cell size.

Generally it is not recommended to run a model at a cell size which is larger than a quarter the width of
draw at which the model is being run. This allows the model to create more realistic draw shapes.

June 2016 89