Академический Документы
Профессиональный Документы
Культура Документы
Tel +1 604.684.6550
Fax +1 604.684.3541
www.gemcomsupport.com
Copyright 2012 Gemcom Software Australia Pty Ltd. All rights reserved.
This software and documentation is proprietary to Gemcom Software Australia Pty Ltd.
Gemcom Software Australia Pty Ltd publishes this documentation for the sole use of Surpac and Whittle
licenses. Without written permission you may not sell, reproduce, store in a retrieval system, or transmit
any part of the documentation. For such permission, or to obtain extra copies please contact your local
Gemcom Software Office.
While every precaution has been taken in the preparation of this manual, we assume no responsibility for
errors or omissions. Neither is any liability assumed for damage resulting from the use of the information
contained herein.
All brand and product names are trademarks or registered trademarks of their respective companies.
This manual has been designed to provide a practical guide to the many uses of the software. The
applications contained within this manual are by no means exhaustive as the possible uses of the
software are only limited by the users imagination. However, it will give new users a starting point and
existing users a good overview by demonstrating how to use many of the functions in Whittle and Surpac.
If you have any difficulties or questions while working through this manual feel free to contact your local
Gemcom Software Office.
Contributors
Darrienne Thobaven
James Willoughby
Product
Surpac v6.3
Whittle v4.5
Table of Contents
Introduction .............................................................................................................................. 5
Requirements ........................................................................................................................................... 5
Objectives ................................................................................................................................................. 5
Workflow ................................................................................................................................................... 5
Table of Contents
Page 4 of 308
Introduction
Pit optimisation allows you to produce a model for pit design based on a block model with real world
constraints. This document introduces the theory behind the strategic mine planning process and
provides detailed examples using the block modelling and Whittle interface functions in Surpac, and the
Foundation, Multi-Analysis, Advanced Analysis and Milawa modules of Whittle.
Requirements
Before proceeding with this training guide, you need:
Objectives
The objective of this tutorial is to give you the skills to complete an optimised mining study using an
existing block model and other known inputs.
Workflow
The process described in this tutorial is outlined below:
Page 5 of 308
A theoretical discussion
Overview
A theoretical discussion
Overview
Before you start to use the software, the theory behind mine planning and pit optimisation is discussed
below.
Course Overview
support-aa@gemcomsoftware.com
Exits
Toilets
Drinks
Breaks/Lunch
Questions
Page 6 of 308
A theoretical discussion
Course Overview
Course Goals
Attendance Requirements
Page 7 of 308
A theoretical discussion
Course Overview
Course Overview
The Company
Page 8 of 308
A theoretical discussion
Course Overview
Page 9 of 308
A theoretical discussion
Course Overview
Page 10 of 308
A theoretical discussion
Strategic Planning
Strategic Planning
Overview
Strategic Planning
Situation Analysis
Economic Evaluation
Decision-Making Behavior
Page 11 of 308
A theoretical discussion
Strategic Planning
Page 12 of 308
A theoretical discussion
Situation Analysis
Situation Analysis
Situation Analysis
Commodities Market
Share Market
(Income)
(Value)
Page 13 of 308
A theoretical discussion
Situation Analysis
Economic evaluation
NPV Analysis
Page 14 of 308
A theoretical discussion
Situation Analysis
NPV Analysis
Page 15 of 308
A theoretical discussion
Decision-making Behaviour
Decision-making Behaviour
Decision-making Behaviour
Corporate Objectives
1.
2.
3.
4.
5.
Page 16 of 308
A theoretical discussion
Decision-making Behaviour
Decision-making behaviour
Neutral to risk
Averse to risk
Sating the markets thirst (maximum reserves)
Cost positioning
Other
Neutral to risk
Maximize NPV
Risk-adjusted discount rate takes into account
opportunity cost and additional risk
Add positive NPV projects to your portfolio
When choosing mutually exclusive projects, choose
the one with the highest NPV
Page 17 of 308
A theoretical discussion
Decision-making Behaviour
Decision-making behaviour
Averse to risk
Choose shorter mine life, smaller pits, flatter slopes
Conservatism can lead to better chance of project
recovery.
Results in lower NPV
Decision-making behaviour
Page 18 of 308
A theoretical discussion
Decision-making Behaviour
Decision-making behaviour
Maximisation of reserves
Increase the metric by which the share market will
assess your value.
Possibly sacrifice NPV.
Increases propensity for share price to increase,
leading to increase in shareholder wealth.
Increase access to hedging and loans.
Page 19 of 308
A theoretical discussion
Decision-making Behaviour
Neutral
Averse
Sating the market
Cost positioning
Other
Decision-making behaviour
Undesirable Behavior
Mining to a prescribed cut-off.
Mining to a prescribed stripping ratio.
Page 20 of 308
A theoretical discussion
Decision-making Behaviour
NPV Curve
Design Criteria
Page 21 of 308
A theoretical discussion
Decision-making Behaviour
Maximum NPV
Of the mine
NPV
$m
Cost of
Production
NPV for the
$/oz
Selected pit
3500
300
3000
250
Decrease NPV
2500
200
2000
Increase
reserves
1500
150
1000
500
100
50
0
1
10
13
16
19
22
Pit number
Page 22 of 308
25
28
31
34
37
A theoretical discussion
Exploration
Principles of
Pit
Optimisation
Corporate
Objectives
Project
Scoping
Strategic
Mining
Software
Sensitivity
Analysis
Scheduling
Page 23 of 308
Risk
Analysis
A theoretical discussion
Mine Planning
Overview
Mine Planning
Optimisation Algorithms
Slopes
Commodity Price
Whittle Concepts
Costs
Mine Planning
Mine Planning
What does mine planning mean to you?
Page 24 of 308
A theoretical discussion
Mine Planning
Page 25 of 308
A theoretical discussion
Mine Planning
Reserve definition
Scoping study
Prefeasibility study
Feasibility study
Bankable feasibility study (BFS)
One-off question
Resource definition
Page 26 of 308
A theoretical discussion
Resources
Cut-off Schedule
Reserves
Long-term Schedule
Short-term Schedule
Page 27 of 308
A theoretical discussion
Page 28 of 308
A theoretical discussion
Inputs
Inputs
Mining Factors
Human Nature
Interest Rates
Economic Uncertainty
Power
Costs
Capital
Costs
Dynamic and
constantly changing
Inflation
Commodity
Prices
Labour
Costs
Contractor
Costs
Mother Nature
Geological Uncertainty
Jointing
& Bedding
Lithology
Constantly changing
as more information
is collected
Faults
Mineralization
Limits & Types
Complex
Page 29 of 308
A theoretical discussion
Inputs
Geological uncertainty
Mineralisation limits
Mineralisation grades
Faults
Jointing and bedding
Lithology
Hydrology
Constantly
changing as
more
information
is collected
Topography
Economic uncertainty
Interest rates
Inflation
Dynamic
and
constantly
changing
Power costs
Capital costs
Commodity prices
Contractor costs
Labour costs
Page 30 of 308
A theoretical discussion
Inputs
Technical parameters
Slope angles
Tonnage & grade model
Dilution
Mining recovery
Milling recovery
Rock properties, continuity, & bulk density
Mining methods
Production Rates
Economic parameters
Discount rate
Mining costs
Processing & overhead costs
Commodity price
Page 31 of 308
A theoretical discussion
Pit Optimisation
Quick Review
Pit Optimisation
Page 32 of 308
A theoretical discussion
Pit Optimisation
AIR
WASTE
MINERAL
What is Optimum?
NPV
Potential value
improvement
Extra value
foregone
Strategy 1
Strategy 2
Selected
Pit size
Pit size
Page 33 of 308
A theoretical discussion
Pit Optimisation
Project Sensitivity
$3,000
Pit Value
$2,500
B
$2,000
A
$1,500
$1,000
$500
$0
0
2,000
4,000
6,000
8,000
10,000 12,000
Pit Tonnes
Page 34 of 308
A theoretical discussion
Pit Optimisation
Price
Costs
Slopes
Small Pit
Pit Size
Larger Pit
Quick Review
Page 35 of 308
A theoretical discussion
Optimisation Algorithms
Optimisation Algorithms
Optimisation Algorithms
Floating Cone
and
Lerchs-Grossman
Page 36 of 308
A theoretical discussion
Optimisation Algorithms
Quick Review
Page 37 of 308
A theoretical discussion
Slopes
Slopes
Slopes
What is an ARC?
C
An arc is a relationship between
two blocks. An arc from block A to
block B indicates that, if A is to be
mined, then B must be mined to
expose A. The reverse is not true.
If B is to be mined, block A may or
may not be mined.
Page 38 of 308
B
If A is mined
so is C
A
A theoretical discussion
Slopes
Arc Relationships
Arc from
A to B
If A is to be mined, B
must be mined to
expose A
The reverse is not true
If B is to be mined, A
may or may not be
mined
Benches
Desired Slope
Page 39 of 308
A theoretical discussion
Slopes
Slope Accuracy
Summary
Page 40 of 308
A theoretical discussion
Page 41 of 308
A theoretical discussion
In other words
Ore tonnages
Grades
Recoveries
Product price
Mining cost
Milling cost
Overheads
What is Value?
Page 42 of 308
A theoretical discussion
Revenue
Costs
$327.75
Revenue
Costs
-$242.25
Page 43 of 308
A theoretical discussion
But wait!
= [(0.25*150
Costs
[(230.625)+ (1177.875) - (1200) ]- (180)
1408.50- 1380
$28.50
Page 44 of 308
A theoretical discussion
Value =
The Section in square
Brackets must => 0
In other words
Page 45 of 308
A theoretical discussion
Marginal Cut-off
Marginal Grade
CostP
Recovery* Price
Summary
Page 46 of 308
A theoretical discussion
Commodity Price
Commodity Price
Commodity Price
Page 47 of 308
A theoretical discussion
Commodity Price
Commodity Price
Commodity Price
Price most powerful influence on profit.
Future prices influenced by:
Historic/Statistical (trend, cyclical, random).
Forces driving change (in market analysis context).
Strategic responses:
Forecast future prices.
Plan to accommodate future uncertainty.
Decision points, mine life considerations.
Price control (hedging, become a price-setter).
Page 48 of 308
A theoretical discussion
Whittle Concepts
Whittle Concepts
Whittle Concepts
Block line
I J
19 36
19 36
Parcel line
I J
20 36
20 36
K No parcels MCAF
PCAF
tonnes
1 1
1 SULF
1.000
2160.
1.000
159.
2160.
769.
K ROCK
tonnes
1 1
1 SULF
1.000
2160.
Page 49 of 308
metal1
1.000
97.
metal2
2160.
305.
A theoretical discussion
Whittle Concepts
Gemcom
Medsystem
Datamine
Vulcan
Whittle (Block 1,1,1)
Whittle vs Surpac
Whittle works in IJK which is number of blocks in x,y,z
Surpac works in N,E,RL which is coordinates in y, x, z
Whittle does not store gradeonly contained metal
Surpac has user blocks and (smaller) sub-blocks
Whittle has blocks and parcels
MCAF and PCAF and slope attached to blocks, ROCK and
grades attached to parcels
Page 50 of 308
A theoretical discussion
Whittle Concepts
Whats a Parcel?
Part of a block for which rocktype, tonnage, and metal
content are known
The total tonnage of the parcels may be the same or less
than the block tonnage
If the parcel tonnage is less than the block tonnage, the
remaining tonnage is undefined waste (unknown rocktype)
If a block has no parcels, the total tonnage of the block is
undefined waste
Neither the position of a parcel within a block, nor its shape
is defined UNLIKE sub-blocks in Surpac
Parcels
C1
Ore A
Tonnes A Metal A
Ore B
Tonnes B Metal B
Ore C1
Tonnes C1 Metal C1
Ore C2
Tonnes C2 Metal C2
Ore C3
A
C2
B
Tonnes C3 Metal C3
C3
Undefined Waste
Tonnes A
Tonnes A
Metal B
Tonnes B
Tonnes B
Metal C1
Tonnes C1
Tonnes C1
Metal C2
Tonnes C2
Tonnes C2
Metal C3
Tonnes C3
Tonnes C3
+
Undefined
Waste
Page 51 of 308
A theoretical discussion
Whittle Concepts
Parcels
In the previous diagram, each parcel is evaluated separately
to determine destination
Each parcel within a block can go to a DIFFERENT
destination i.e. ore and waste in the same block
Parcel size is important for selective mining issues
Summary
Whittle works in I,J,K not N,E,RL
Whittle stores contained metal, not grades
Whittle has blocks and parcels similar to Surpac blocks and
sub -blocks
Unlike Surpac, the location and shape of each parcel is unknown
And the sum of parcel tonnage may be less than block tonnage
(the rest is undefined waste )
Each parcel can go to a different destination
Page 52 of 308
A theoretical discussion
Page 53 of 308
A theoretical discussion
Equipment size
minimum working space
multiple SMUs in equipment size block
use template in Minimum Mining Width
Page 54 of 308
A theoretical discussion
X offset
Z offset
Expanded Model
Page 55 of 308
A theoretical discussion
Adjusting Framework
Page 56 of 308
A theoretical discussion
Slopes
Page 57 of 308
A theoretical discussion
Bearing
Slope Angle
20
42
77
38
135
46
180
42
260
38
335
44
60
60
55
45
45
55
Constant
Constant
Slope
Slope
60
55
45
60
Page 58 of 308
A theoretical discussion
Slope Bearings
Slope Considerations
Rule of thumb bench search for arc generation in each
sub-region
(max(x,y) x 8) / z
Page 59 of 308
A theoretical discussion
Summary
Four possible slope methods
Remember Whittle uses bearings not wall positions.
Remember slopes are smoothed between bearings
Use overall slopes and make allowance for ramps
Whittle employs rule of thumb to calculate number of
benches for arc generation
Page 60 of 308
A theoretical discussion
Mining Dilution
Whittle defines dilution as extra waste taken to the mill
Ore tonnes are increased while metal is the same so
grade decreases.
For a dilution of 5% on 100t @ 2 g/t, the resulting ore is
Page 61 of 308
A theoretical discussion
Mining Recovery
Whittle defines mining recovery as ore taken to the waste
dump
Ore tonnes and metal are decreased so that grade is the
same
For a recovery of 5% on 100t @ 2 g/t, the resulting ore is
ORE
Mining Recovery
(extraction loss)
MINING LIMIT
Profit =
[ (Ore*Grade*Recovery* (Price-CostS)) - (Ore*CostP) ] - Rock*CostM)
x Mining Recovery
x Mining Recovery
x Dilution
Page 62 of 308
A theoretical discussion
Costs
Costs
Costs
Calculating Costs
Types of Costs
Units
Mining cost
$/tonne mined
Processing cost
$/tonne processed
Selling costs
$/unit of product
Page 63 of 308
A theoretical discussion
Costs
Page 64 of 308
A theoretical discussion
Costs
Page 65 of 308
A theoretical discussion
Costs
Page 66 of 308
A theoretical discussion
Costs
Reference Block
Reference Block :
Used to calculate CAFs
Represents the base case
scenario for mining
Typically positioned at the
top of the pit
Does not have to
physically exist
Page 67 of 308
A theoretical discussion
Costs
Profit =
[ (Ore*Grade*Recovery* (Price-CostS))
- (Ore*CostP) ] - Rock*CostM)
x PCAF
x MCAF
Summary
Cost must be in units of relevant activity
Any cost that continues when the production stops should NOT
be included in the optimisation
Do NOT add a stripping allowance or be conservative with the
parameters
CAFs are calculated from a reference block
The reference block does not have to exist but should give
CAFs >=1.
Page 68 of 308
A theoretical discussion
Pit Shells
Pit Shells
Revenue Factors
A simple way of scaling values to provide a range of pit shells
which are optimal for different prices
If the base price (RF=1) of gold is $800/oz
RF = 0.5 has a gold price of $400/0z
RF = 2.0 has a gold price of $1600/oz
Page 69 of 308
A theoretical discussion
Pit Shells
Plan
Page 70 of 308
A theoretical discussion
Pit Shells
Summary
Nested pits are generated by the revenue factors
Page 71 of 308
A theoretical discussion
Page 72 of 308
A theoretical discussion
Selection is done by comparing the grades of the material with precalculated processing cut-offs
If it does not satisfy the cut offs, it is treated as waste
Page 73 of 308
A theoretical discussion
Profit =
[ (Ore*Grade*Recovery* (Price(Price-CostS)
CostS) - (Ore*CostP
(Ore*CostP)) ]
- Rock*CostM
Rock*CostM))
Summary
Page 74 of 308
A theoretical discussion
Input
Finance
Mining
Commodity Price(s)
Price(s)
Costs
Discount Rate
Ore Selection
Corporate Objectives
Geotech
Slope Stability
Metallurgy
Recovery
Geology
Costs
Tonnes
Processing Rates
Grade
CutCut-off Grades
Rocktypes
Page 75 of 308
A theoretical discussion
Course Workflow
Course Workflow
Course Workflow
Page 76 of 308
A theoretical discussion
Course Workflow
Process Overview
In Surpac
In Whittle
Validate geological
parameters
Add cost & slope info
Generate reports for
validation
Export block model
In Surpac
Import res or msq or dxf
Design pits
Export pit design
In Whittle
import pit design
Scheduling & analyses
Page 77 of 308
A theoretical discussion
Course Workflow
Page 78 of 308
A theoretical discussion
Course Workflow
Page 79 of 308
A theoretical discussion
Course Workflow
Advanced Analysis
Schedule
Best Case (inner shell out)
Worst Case (top down)
Specified (using pushbacks)
Detailed Cash flow
DCF (Discounted Cash Flow Analysis)
NPV (Net Present Value = Discounted Cash flows)
Adjustment for Time Costs
Variables changing with Time
IRR (Internal Rate of Return)
Sensitivity
Spider Diagram
Advanced sensitivities
(Custom Graphs & Advanced Analysis)
Advanced Optimizations
Cut-off Optimization
Stockpiles
Page 80 of 308
A theoretical discussion
Course Workflow
Other Tools
Reblocking Tools
Parcels
Combining & Splitting Blocks
Calculating MCAFs & PCAFS
(Cost Adjustment Factors for each block)
Block Value Calculation
For Checking
Variables
Expression
Changing by time
Page 81 of 308
Overview
Requirements
The data files required for this exercise are: bg_id2_070313.mdl, topo2007.dtm, topo2007.str
Getting started
Please note: This manual uses version 6.3 of Surpac.
Because this type of work is repetitive and can generate large numbers of files, it is vital to have a
structured file naming convention. For this course the data structure consists of several folders, named
according to data function, beneath a root folder of c:\training\data as shown in the following image:
1. In Windows Explorer, create a separate directory for the training data and beneath this directory
create a new directory called 1_originaldata and several others as shown above. Copy the data
files into the 1_originaldata directory.
VERY IMPORTANT Do NOT put spaces in the directory names.
Page 82 of 308
Getting started
2. Before you start Surpac, right-click on the Surpac icon on the desktop. Change the Start in
directory to the one that you have just created.
3. Start Surpac.
4. Make it so that only the Main Menu and Block Model Menu are displayed by placing the mouse
pointer over a space on a menu or toolbar and right-clicking on the names of the menus you want
to show. Also display the Block Model Toolbar in the same way.
Note: Any instructions in this tutorial about accessing block model functions in this tutorial, refer
to either the Block Model menu, or the Block Model toolbar.
5. In the Navigator, right-click the training folder, and choose Set as work directory.
6. Under 1_originaldata, drag the block model file bg_id2_070313.mdl from the Navigator into
Graphics.
The block model is displayed in the Status bar at the bottom of the Surpac window as shown
below:
Page 83 of 308
Getting started
One of the important features of the block model to note is the number of user blocks in each direction
160 in Northing, 100 in Easting, and 75 in RL giving a total number of user blocks of 1.2 million. Whittle
has a limitation of 99999 blocks in any direction, but this block model does not have to be modified for
this limitation.
In addition, you can see from the list of attributes that three attributes exist which define gold grade, and
an additional three attributes could define the rock type litho, rock and weathering. An attribute called
sg exists for specific gravity, that is density. sg is a mandatory attribute. You can also see that some
attributes have a background value of -99.
8. Click Apply in the form to close it.
Page 84 of 308
Getting started
Type
Decimals
Background
Au_cut20
Real
-99
Au_cut15
Real
-99
au_uncut
Real
-99
ave_dis
Real
-99
count
Integer
Litho
Character
Sed
min_dist_samp
Mined
Real
-99
Character
Num_samp
Integer
-99
Pass
Integer
Rescat
Character
UNCLASS
rock
Character
sg
Weathering
Real
Character
1.8
fresh
Only one attribute for grade is exported to Whittle so it is necessary to ask the person supplying the
geological data which gold grade is relevant. For this exercise, the au_cut15 attribute is the appropriate
attribute.
Often you have a situation where the original data is the resource block model which should not be
modified. Therefore the next step is to make a copy of the block model to preserve the original data.
Page 85 of 308
Getting started
9. In the Navigator, highlight the 2_optimisation_bm directory, right-click, and choose Set as work
directory.
10. Choose Block model > Save As, delete the Model name, and type bg_opt_070322 to indicate
that this model is created for the purposes of optimisation.
Page 86 of 308
Getting started
2.05
0.50
0.45
0.41
Total
16.79
15.39
16.66
16.90
0.00
0.00
0.00
0.00
Total
14.74
14.89
16.21
16.49
Cost
Value
Unit
Comment
Administration
1.5 million
$ pa
Head office
Personnel
0.5 million
$ pa
Consumables
150,000
$ pa
Ore Haulage
3.00
$/t hauled
Contractor costing
400 RL 320 RL
< 320 RL
Ultramafics
38 (1)
42 (4)
Sediments
40 (2)
48 (5)
Porphyry
44 (3)
55 (6)
Metallurgical Recovery
Weathering attribute
Recovery
Page 87 of 308
Getting started
Hardcap
94%
Oxide
94%
Transition
94%
Fresh
90%
In this exercise, mining and processing costs have been defined by RL, weathering and lithology. You will
use attributes to hold the mining and processing costs. The metallurgical recoveries are listed by
weathering type, and schedules are required by weathering as well. The geotechnical zones are defined
by RL and lithological code. Next you will investigate the block model for relevant data.
10. Choose Display > Display block model to display the entire block model.
The following form appears:
11. Fill in the form as displayed in the previous image, and click Apply to draw the model with
block faces.
Page 88 of 308
Getting started
To get a better indication of where the ore deposit is located in the block model, you can add a graphical
constraint to display only the blocks where the gold value is greater than zero.
Page 89 of 308
Getting started
Page 90 of 308
Getting started
Note: For all functions that you can run either from the menu or from a toolbar, the first instance in this
document will refer to the menu command, and the subsequent instances will refer to the toolbar button.
For example, you can access the graphical constraint function from the toolbar button as shown below:
14. Rotate the constrained block model to gain familiarity with the deposit.
Next you will run a block model report to investigate the values in the weathering attribute.
15. Choose Block model > Report.
16. Enter the information on each form as shown, and click Apply on each form.
Page 91 of 308
Getting started
For the following form, some of the options are explained for clarity.
Format Headers?
This option allows you to flag whether you wish to reformat the header fields in the table. When this
option is selected, all underscores are replaced with a space and each word begins with a capital letter.
This formatting generally produces neater looking results. When this option is cleared, the headers retain
their original formatting. This is useful for maintaining the integrity of data when importing and exporting
between applications.
Use Volume Adjustment?
Make a volume adjustment if the physical volume of a block is different to that implied by the dimensions
of the blocks in the model. For example, an expansion factor for excavated material, or if differential
dilution and reduction has been applied. Select `Yes' if you wish to use pre-determined volume
adjustments for blocks.
Geometric Grouping?
This allows you to report by groups of solids, strings, or DTMs.
Pivot Compatible?
If you output a report to CSV, and select Pivot Compatible, you can then insert a pivot table in Excel and
easily create a SUM or an AVERAGE for the data, or do other analysis using a pivot-table.
Page 92 of 308
Getting started
By selecting to report all values for the character attribute of weathering, any blocks with a value of null
will also be reported.
Note:
The Keep blocks partially in the constraint check box changes the way the block is selected for inclusion in
the constraint. The default method in Surpac uses the centroids to define whether the block is inside or outside
the constraint. When Keep blocks partially in the constraint is selected, will include a block in the report if
ANY part of the block is inside the constraint (instead of the centroid being inside) Be careful using this check
box, because if you use it in reports such as the mineral inventory report, and do not use it in conjunction with a
volume adjustment factor, you will overestimate your resource.
Page 93 of 308
Getting started
Leaving the constraints form blank as shown in the previous image will report on all blocks contained in
the block model.
The report appears similar to the following one. The weathering attribute contains the values of fresh,
transition, oxide and hardcap. Note that there are no blocks with a null weathering value. Some blocks
do contain the default au_cut15 value of -99 because the average values of Au Cut15 is close to -99. No
air blocks are present.
17. Create similar reports for the attributes of litho, rescat and rock.
Do any of these attributes contain null values? How many tonnes have been classified? Which attribute
appears to be the basis for the sg values?
Page 94 of 308
Getting started
These reports show that it is reasonable to use weathering as the basis for the rock code required by
Whittle, litho for the geotechnical definition and to keep the following attributes for further processing:
Au_cut15
Sg
Litho
Weathering
Page 95 of 308
Getting started
Next, for comparison purposes, you will note the size of the block model file.
18. In the Navigator, right click on the block model and choose Properties.
The file size is displayed in the message window 65,953,976 bytes. There are 1024 bytes in a KB, so
the file size in MB is approximately 62.9 MB.
19. Check that Block Model Audit is on by looking at the
button on the right side of the Block
Model Toolbar. If it is not recording, turn it on by clicking the button.
You can record a block model audit with a macro, which is very useful for both an audit trail and for use in
the next optimisation study. You will look at this function later in the course.
To speed up processing time and reduce hard disk space, you will delete the attributes that are not
required.
Note:
If attribute 2 is a calculated attribute that references attribute 1, you must delete attribute 2 before attribute 1.
Tip: Press TAB to navigate to the next line in the previous form, or press ALT+TAB to move to the
previous line in the form.
22. Save the block model and display the file size again in the Message window using the
Navigator.
The new block model is 30.1 MB in size which is less than half the size of the original. This is the
expected result because you have deleted more than half of the attributes.
Page 96 of 308
Getting started
ktonnes
Au g/t
Au koz
Fresh
1,949.0
1.81
113
Transition
1,917.5
1.65
102
Oxide
887.5
1.76
50
Hardcap
101.5
1.42
Total
4,855
1.73
270
Page 97 of 308
Rock Code
Rock Code
Whittle requires a code, which defines rock type, with character values that are no longer than 4
characters in length. For this rock code, Whittle has a limit of 250 values in total, that is 250 different rock
codes. Although it is not necessary, you will ensure that every block in the model has a value for the rock
code and that characters are uppercase to make it easier to read the report.
The weathering attribute is the basis of the rock code, but you need to define the air blocks as well. You
will modify the weathering attribute to include air blocks, and then add a calculated attribute called rock
which will satisfy the requirements of Whittle. Air blocks are defined by the topo2007 dtm. (It is vitally
important to check that this dtm covers the entire surface area of the block model.) You can assume
that the rest of the values for the weathering attribute are defined correctly.
1. Drag topo2007.dtm into Graphics to confirm that it covers the block model.
4. Click Yes when you are prompted to overwrite the block model.
Page 98 of 308
Rock Code
It is vitally important that you perform a validation after every modification to the block model.
You can perform validations using
a graphical constraint
a block model report
colouring of the block model
block ID of individual blocks
Next you will visually check the weathering values by adding a graphical constraint for each value.
5. Add a graphical constraint to display the air blocks as shown in the following image:
Do the air blocks cover the entire area of the block model? Was the top of the model set too low?
Remove the graphical constraint and constrain the block model so that you can see each of the other
weathering types in turn. How thick are each of the weathering types?
6. Check that there are no null values in the weathering attribute.
You can find null values in two ways:
generate a block model report that is unconstrained and the tonnage of null values will be
reported
add a graphical constraint with a space for the value e.g. weathering=
(The form will not
accept a blank value for weathering but if you insert one space this is read as a null value.)
Next you will add the new rock code as a calculated attribute.
7. Choose Attributes > New.
8. Enter the information on the form as shown, and click Apply on the form.
Page 99 of 308
Rock Code
The complete formula is as follows. You can copy and paste it into the Add attributes form.
iif(weathering="air","AIR",iif(weathering="hardcap","HCAP",iif(weathering="oxide","OXID",
iif(weathering="transition","TRAN","FRSH"))))
Character values are case sensitive so AIR and air are different values. Be very careful with capitals
and spelling. Note also that Surpac automatically converts the attribute name to lowercase regardless of
the case you enter in the form.
Tip: For more information about the syntax of expressions, open the Surpac help and search for the
topics: Expressions, and Block Maths.
Support desk tip
Problem hunting syntax errors
An easy way to find syntax errors in an expression is to copy the expression to Word or some text editor
and do the following:
Ensure an if statement has a , after the test, and value1 parts for example
iif(test,value1,value2)
9. Check the file size of the block model after the addition of this calculated attribute.
10. Validate the new attribute and check that all blocks have a value for the rock attribute by creating
a new block model report.
Rock Code
Whats the difference between the block maths and the assign value functions?
The main difference between the two functions relates to saving the block model. The Assign Value
function is found under the Estimation menu and any function under that menu prompts for a Block Model
Save after the function is complete. Functions under the Attribute menu (including Block Maths) do not
prompt for a Block Model Save.
For the Assign Value function, it is not necessary to perform a Block Model Save. Use of either function is
personal preference. If Block Maths is used, and something went wrong with the model immediately after
the function (and before any manual save) then the changes will be lost from the model. If the Block
Maths function is used then it is important to regularly save the model manually using Block Model Save.
This training guide uses the Assign Value function to avoid manual saving of the model.
Gold Grade
Gold Grade
The Whittle export function requires at least one grade attribute. There are several possible conditions
that are invalid. You should test for these and rectify them if they exist:
negative grades
non-zero grades in blocks defined as air
Gold Grade
The majority of the blocks in the model have the background value of -99 for gold grade.
Next, you will remove this constraint and display the air blocks with grades above 0.
4. Choose Constraints > New graphical constraint.
5. Enter the information on the form as shown, and Apply the form.
Note that each line typed into a constraint form can be used in a variety of combinations. Options for the
combinations include AND, OR and brackets. AND is equivalent to the intersection so that in the previous
example, for a block to be included in the constraint, it must fulfill BOTH lines a and b. OR is the
equivalent of a union. For a block to be included it must fulfill one or both of lines a and b.
The block model is displayed as shown in the following image:
Gold Grade
Blocks exist for both conditions, and you will modify these blocks next.
6. Choose Constraints > Remove all graphical constraints.
7. Choose Estimation > Assign value.
8. Enter the information on the forms as shown, and Apply each form.
Gold Grade
Next you will verify that the attribute assignment has given the expected result by viewing the attributes
for a random block.
9. Choose Attributes > View attributes for one block and select one of the blocks at the top of
the model.
The following form appears.
11. When you are finished with the form, click Apply.
12. Press ESC to stop examining blocks.
13. Visually verify that no blocks exist with negative gold grades by using the graphical constraint
function.
14. Using the Assign value function, set all grades to 0 for air blocks and then check that the change
was successful.
Page 105 of 308
Gold Grade
Density
Density
The Whittle export function requires an attribute to define density. It is illogical for an air block to have a
density other than zero. If this occurs you should change the sg value.
1. Visually display air blocks with a non-zero density.
The block model appears as shown:
2. Set the density of the air blocks to 0 and visually check that the changes were successful.
3. Check for other illegal values such as a negative sg or a zero sg in blocks that are NOT air.
4. Using the Navigator, display the size of the block model, which is approximately 32.7 MB.
5. Generate a new block model report using the rock.bmr format file for all blocks with au_cut15 >
0.5.
Density
You can compare the previous report with the same report for the id2 block model to quantify the
changes in tonnes, grade and contained metal.
Mineral inventory au_cut15>0.5g/t
Rock
Hardcap
ktonnes
Au_cut15 g/t
ktonnes
Au_cut15 g/t
Au koz
98.3
1.45
-0.1%
0.0%
0.0%
887.5
1.76
55
0.0%
0.0%
0.0%
Tran
1,917.5
1.65
112
0.0%
0.0%
0.0%
Fresh
1,949.0
1.81
124
0.0%
0.0%
0.0%
Total
4,852
1.73
296
-0.1%
0.0%
0.0%
Oxide
Note that the changes have only occurred in the top-most rock code where the air definition has
influenced sg and grade.
6. Check that there are no blocks with sg=0 beneath the topo and that no negative values for sg
exist.
It is a good idea to make this check unless that situation is intended.
Assessment
Assessment
Theory Assessments
Draw a NPV/Tonnes curve
Which pit has the highest NPV?
Which pit is the least sensitive to change?
What controls slope accuracy in Whittle?
How does Whittle store the location of the block?
What is the difference between parcels and subblocks?
Exercise 1 Assessments
Slopes in Whittle
Slopes in Whittle
Slopes in Whittle
Slopes in Whittle
Slopes
Slopes
Slopes in Whittle
Slopes
Benches
Desired Slope
400 RL 320 RL
< 320 RL
Ultramafics
38 (zone 1)
42 (zone 4)
Seds
40 (zone 2)
48 (zone 5)
Porphyry
44 (zone 3)
55 (zone 6)
You could add the attribute as an integer, and then fill it using the Assign Value function or you could add
it as a calculated attribute. You will try both methods to illustrate the difference in file size and time
required.
First you will add the attribute as an integer.
1. Choose Attribute > New, and enter a name of zone, type of integer and background value of 1.
2. Using the Assign value function, fill the attribute with the zone numbers (1 to 6) shown in the
table above.
3. When you are finished, display the properties of the block model file.
The integer attribute adds 4 MB to the file size of the block model.
4. Delete the zone attribute using Attributes > Delete.
5. Display the block model file size, and check it has returned to the previous size.
6. Choose Attributes > New.
7. Enter the information shown below, and Apply the form.
The calculated attribute does not increase the block model file size. The calculate attribute is faster to
add, quicker to process, and you can edit the formula without having to delete the attribute and add it
again.
Costs in Whittle
Costs in Whittle
Cost Handling
Costs in Whittle
Mining Costs
Single value (Reference cost in Whittle)
Value or expression for rock codes (Mining CAF in Whittle)
Single expression (Whittle)
MCAF (Mod file attribute in Surpac)
Processing Costs
Costs in Whittle
Processing Costs
The extra
Costs in Whittle
Costs in Whittle
Reference Block
Reference Block :
Used to calculate CAFs
Represents the base case
scenario for mining
Typically positioned at the
top of the pit
Does not have to
physically exist
Calculating Costs
Types of Costs
Units
Mining cost
$/tonne mined
Processing cost
$/tonne processed
Selling costs
$/unit of product
Costs in Whittle
What to Include?
Costs to Include
Incremental (or variable) Costs
Directly proportional to tonnage or unit of product
Wages, fuel, explosives etc
Costs must be included in the associated activity (e.g. drilling in mining)
Include
Costs in Whittle
Dayworks
Mobilisation/Demobilisation
Site Clearing & Topsoil
Geological
Grade control
In-pit supervision
Surveying
Support (Geotechnical)
Dewatering
Costs in Whittle
Camp costs
Administration
Capital replacement
Communication
Legal Services
Safety & Training
Any Other Fixed Costs
Costs in Whittle
Costs in Whittle
Example:
Gold CIP/CIL
Costs in Whittle
OR
OR
Costs in Whittle
Mining
Processing
Comments
Blasting
Load & Haul
Miner's Wages
Reagents
Stockpile management
Mill Wages
Stop ?
YES
Costs in Whittle
Mining
Processing
Comments
Stop ?
Exploration
Costs
Cost of Mill
Must be depreciated
over the life of a
project (or pit)
NO
FIXED COSTS
Mining
Processing
Comments
These stop when
mining stops but
should they be mining
or processing costs?
Stop ?
YES
Costs in Whittle
Mining
Processing
Rehabilitation
Maintenance
Comments
Stop ?
Mining
Processing
(complex)
Comments
Stop ?
Repayments,
fees and interest
Costs in Whittle
Compensation
Mining
Processing
Comments
Stop ?
Road ?
Maintenance
$980,000
1,000,000t
$0.98/t
4,000,000t
$0.24/t
Costs in Whittle
Mining Waste
Mining Ore
Extra for
mining Ore
Milling Ore
$1.05
$1.87
$0.82
$8.25
$1.17
$2.15
$0.98
$8.25
$1.29
$2.47
$1.18
$8.25
$1.41
$2.84
$1.43
$8.25
$1.53
$3.26
$1.73
$8.25
Bench
Mining
Waste
Milling Ore
Extra for
Mining Ore
Time Cost
Milling Cost
COSTP
$1.05
$8.25
$0.82
$0.98
$10.05
$1.17
$8.25
$0.98
$0.98
$10.21
$1.29
$8.25
$1.18
$0.98
$10.41
$1.41
$8.25
$1.43
$0.98
$10.66
$1.53
$8.25
$1.73
$0.98
$10.96
Costs in Whittle
Time
Cost
Mining
COSTM
Milling
Ore
Extra for
Mining Ore
Milling Cost
COSTP
$1.05
$ 0.24
$1.29
$8.25
$0.82
$9.07
$1.17
$ 0.24
$1.41
$8.25
$0.98
$9.23
$1.29
$ 0.24
$1.53
$8.25
$1.18
$9.43
$1.41
$ 0.24
$1.65
$8.25
$1.43
$9.68
$1.53
$ 0.24
$1.77
$8.25
$1.73
$9.98
Bench
Example 1
Cost of Processing Mill
Costs in Whittle
Costs in Whittle
Example 2 - Trucks
If the expected life of the mine is shorter than the life of the
trucks treat same as mill
If life of mine is long progressive truck purchases would stop
if mining stopped. Cost of replacement must be averaged over
the life of the mine and treated as a time cost
Costs in Whittle
Example 4
Loan for Initial Costs
Example 5
Loan for Recoverable Costs
Costs in Whittle
Costs in Whittle
Costs in Whittle
Value
Unit
Comment
Administration
1.5 million
$ pa
Head office
Personnel
0.5 million
$ pa
Consumables
150,000
$ pa
Ore Haulage
3.00
$/t hauled
Contractor costing
The metallurgical recoveries and costs have been provided on the basis of weathering while grade
control will be conducted in the hardcap and oxide as sampling of ditchwitch lines in porphyry material
only. The grade control for transition and fresh material will be blast hole sampling of porphyry material
only.
Possible processing costs are shown in the following table:
Processing Cost ($/t ore)
Porphyry Only
Admin
Haulage Personnel Consumables
Process L&H add GC
Hcap
1.88
3.00
0.63
0.19
9.00
0.05
Oxide
1.88
3.00
0.63
0.19
9.00
0.20
Tran
1.88
3.00
0.63
0.19
10.25
0.27
Fresh
1.88
3.00
0.63
0.19
10.50
0.30
Ultramafics and Sediments
Admin
Haulage Personnel Consumables
Process L&H add GC
Hcap
1.88
3.00
0.63
0.19
9.00
0.05
Oxide
1.88
3.00
0.63
0.19
9.00
0.20
Tran
1.88
3.00
0.63
0.19
10.25
0.27
Fresh
1.88
3.00
0.63
0.19
10.50
0.30
2.05
0.50
0.45
0.41
Total
16.79
15.39
16.66
16.90
0.00
0.00
0.00
0.00
Total
14.74
14.89
16.21
16.49
Note that the load and haul component listed above is only the extra cost of mining ore, over and above
the cost of mining waste. For example, in the hardcap, waste costs $1.31/BCM to mine while ore costs
1.31 + (0.05*sg) = $1.40/BCM to mine.
For this exercise the mining costs and processing costs are contained in the mod file. You should add
those costs as attributes in the Surpac block model.
MCAFs
MCAFs
Instead of adding the mcaf as a real attribute and using the assign value function based on RL, you can
define the mcaf as a formula (line of best fit) and add it as a calculated attribute. The table of mining costs
on page 6 gives the total unit mining cost in $/BCM. A reference mining cost must be chosen so that the
MCAFs can be calculated. Choosing the highest defined cost of $1.69/BCM as the reference cost and
dividing by the density of 1.8 for hardcap, the reference cost (or COSTM) is $0.94/t.
You will add the mining cost per BCM to the model as the formula based on RL, and use calculated
attributes for mining cost per tonne and MCAF. You can use a spreadsheet to graph the mining cost
($/BCM) versus RL, and a line of best fit found as shown in the following graph:
Mining cost $/bcm vs RL
y = -0.0437x + 18.453
R2 = 0.991
9.00
8.00
7.00
$/bcm
6.00
5.00
4.00
3.00
2.00
1.00
0.00
200
250
300
350
400
RL
When you generate the formula from a graph using Excel, make sure you use a graph type of XY Scatter
graph rather than a Line graph. Also always check the formula before using it, by manually calculating the
cost for a given RL.
1. Add the following calculated attributes:
Attribute
decimals
expression
mincost_pbcm
-0.0437*_zcen + 18.453
mincost_pt
iif(sg=0,1,mincost_pbcm/sg)
mcaf
mincost_pt/0.94
Note that it is important to test for impossible conditions and deal with them in any expression. When a
block has sg = 0, then mincost_pt has a divide by zero error if no test is included. In this case, a warning
message appears in the Message window and the attribute is given a value of -1. There is no fatal error
and it is possible to export a model with these negative values.
Be careful, when you add attributes, to ensure that the type is calculated and not real. If you choose
real, the formula will be seen as a description and all of the blocks will have a value of 0. An easy way to
MCAFs
tell that the attribute is calculated while the form is displayed is that the background value is greyed out. If
you choose real, the background value is white.
Note for using calculated attributes: The benefit of using calculated attributes is that you can change the
formula without having to delete the 3 formulae. However if you modify mincost_pbcm, only the updated
values for mincost_pbcm are immediately shown. To see the correct values for the mincost_pt and mcaf
attributes, you must close and open the block model.
2. Visually check some blocks for the expected values of the new attributes.
PCAFs
PCAFs
For this exercise, processing cost is dependent on both lithology and weathering. If the processing cost is
only dependent on weathering, you could enter it in Whittle because you have chosen weathering as the
basis for the rock code.
In Whittle, each rock code has its own reference processing cost rather than one processing cost for all
codes as you used for mining cost.
You will use the hardcap processing cost as the reference cost (or COSTP) for the two groups of costs
porphyry and others (ultramafics and sediments).
Porphyry
Total
Pcaf
Hcap
16.79
1.0000
Oxide
15.39
0.9166
Tran
16.66
0.9923
Fresh
16.90
1.0066
Ultramafics and Sediments
Total
Pcaf
Hcap
14.74
0.8779
Oxide
14.89
0.8868
Tran
16.21
0.9655
Fresh
16.49
0.9821
COSTP for all rock codes is $16.79.
1. Add a pcaf attribute to the block model in Surpac, and assign it the pcafs shown in the previous
table using a calculated attribute.
The relevant formula is:
iif(rock="AIR",1,iif(litho="Porphyry",iif(rock="HCAP",1,iif(rock="OXID",0.9166,iif(rock="TRAN",0.9
923,1.0066))),iif(rock="HCAP",0.8779,iif(rock="OXID",0.8868,iif(rock="TRAN",0.9655,0.9821)))))
2. Save the block model.
3. Validate that the pcaf attribute has been correctly set, either by viewing attributes for several
blocks, or by running a block model report.
PCAFs
All of the block model functions performed on the model so far, have been stored in the block
model audit trail.
4. To display the audit trail, choose Block Model > Display Model Audit Trail.
A form similar to the following one appears.
You can change or delete the comments shown next to each function. You can also save the
audit trail as a macro and use it in the future.
5. Click Save as Macro in the Audit Trail form, type the macro name of bmaudit, and Apply the
form.
6. Right-click on bmaudit.tcl, and select Edit.
Note:
Be careful not to double-click the tcl file in the Surpac navigator because this will run the macro.
You could use parts of this macro on another project with only small modifications.
Support desk tip
A valid Whittle model file does not contain any mcafs or pcafs that are zero or negative. In this case you
have generated the mcafs and pcafs so that they cannot be 0 or negative. When mcafs or pcafs are
defined in a different way, always perform a check for 0 or negative values:
The Surpac export function requires an attribute or value for mcaf and pcaf however it is possible to leave
the input box blank. If no attribute or value is entered, Surpac will automatically use the first real attribute
in the attribute list sorted alphabetically, in the place of the particular caf attribute. For this reason,
ALWAYS have an attribute or value for mcaf and pcaf even if it only contains the value of 1 for all blocks.
The rock code for air is required and also the rock code for waste. If it is known that some blocks
definitely have no grade and there is no need to differentiate between the rock codes for waste, then you
can enter a single rock code for waste. Usually the different types of waste will be required for scheduling
or cost purposes and it is best to allow the optimisation process in Whittle determine what is waste and
what is ore. In this case, it is acceptable to enter a code for waste that does not exist in the block model.
The form requires that you enter a value.
The export function also has an option to validate the Surpac block model for Whittle requirements. Many
of the checks that have been performed manually are checked BEFORE a mod and par file are created.
After the validation is complete, a report is written. If any blocks contain values that are illegal, the value
of WTLERROR is written to an attribute created for error identification purposes. You do not have to
perform validation before export, however it is recommended that you validate each time an export is
performed.
The form contains additional fields for Block value (a mod file format not covered in this course), mine
number (if multiple projects are to be seen as one project in Whittle) and an entire section on Pit List file
export.
The volume adjustment attribute is used to modify the ore tonnes in a block but NOT the total block
tonnes. This attribute is a factor less than one i.e. 0.8 not 80%. If the block has 100t @ 2 g/t, without the
volume adjustment the ore parcel contains 200g. If the volume adjustment was 0.5 then the total block
tonnes remain 100t and the parcel contains 50t @ 2 g/t for 100g.
1. Choose Block Model > Export > to Whittle.
2. Enter the information in the following form, and Apply the form.
Attribute Type
Rock code
Character or calculated
Mcaf
Pcaf
Zone
Integer or calculated
Sg
Mine number
Integer or calculated
Block value
Grades
The error report is automatically opened in your default text editor. If the validation failed, no mod and par
file will be created. An example of a validation report from a block model that passed the tests is shown
below.
The export report is also opened in the default text editor. This will be used for validating the import into
Whittle later on.
After the validation and export reports are displayed, a par file, and a mod file of 75.8 MB size, are
created.
3. Open the mod and par files in a text editor.
4. Compare the mod file to the file format shown on page 50.
What data do you think the columns in the following par file represent?
In this situation, the block model has an extra attribute added called error, and the blocks with negative
grades have the value of WTLERROR written to that block. The remaining blocks have a NULL value. To
display the problem blocks, use a graphical constraint such as that shown below of error != (space)
If the block model validation fails for any of the other reasons listed, the same results as above will occur.
Assessment
Assessment
Exercise 1 Assessments
What checks need to be performed on the grade
attribute/s?
What function would you use to change the sg of Air
blocks to zero?
How can you problem hunt a syntax error in an
expression?
What are the 4 ways to define slopes in Whittle?
Exercise 1 Assessments
How are the CAFs calculated?
What should always be generated before exporting
the mod and par file?
What information is contained in the par file?
What are the 2 types of information that could be used
for mcaf or pcaf in the Surpac export?
After a short wait and a loading, please wait message, the main work area is displayed as shown below:
menu bar
Tool Bar
Data Pane
Navigation Pane
This is the main Whittle window. It has several sub-sections. At the top of the window you will see the
familiar looking title, menu and tool bars. Below, on the left, is the Navigation pane. You build up an
analysis model consisting of various data, functions, and results in the Navigation pane.
On the right side is the Data pane, which displays a set of tabs containing data, settings and results. At
the very bottom is a status bar. On the right side of the status bar is your licence details.
Name of node
Data status
Expand/contract
The items that you can see in the Navigation pane are called "nodes". Each node represents a set of
inputs, processes and results. The nature of the data, process and results depends on the type of node.
Lines link each node to its parent (above and to the left) and its child or children (below and to the
right). Parents provide data and results upon which the node depends. Children depend on the data and
results in a node.
Icon: Illustrates the sort of data and results the node contains.
Description: User-defined text.
Status line: Shows the status of the data and Results.
Display toggle: Click on it to show or hide child nodes.
STRATEGIC MINE PLANNING SOFTWARE
Working in the Navigation pane
Click on a node (left mouse button). The data and results associated with that node are displayed in the
Data pane.
Right click on a node (right mouse button). A menu appears which shows all the operations that you
can perform on that node (including export).
Click on a display toggle. All nodes in the tree with children have a display toggle. Click on the toggle to
display or hide the child nodes.
The blue
highlight
indicates
this line is
selected
Tabs
Click Accept to
commit your
changes
The Toolbar is probably the easiest way to learn how to set up and control the Whittle programs. The
SAVE (floppy disc symbol) OPEN (opening file symbol) ADD (plus sign) and the three running men icons
will help you perform most tasks.
At certain nodes you can view the optimal shells, model and schedule. When this is the case, the
small inverted cone icon on the right side of the toolbar is available. When there is nothing to view the
icon is unavailable.
When you become proficient with Whittle you will find that working in the data and navigation panes is
easy and you will probably use the right-mouse click to activate a context sensitive menu. This menu
displays all items that can be used at that node.
1. Select the Project node (the top node in the Navigation tree), and click on the Description tab.
2. In the data pane, type some text at the end of the Description box, and Accept the changes.
3. Select the block model node, and click on the Dimensions tab. Look at the block model
dimensions.
4. Click the Help button.
A help topic specifically related to the Dimensions tab is displayed (that is, a context-sensitive
help topic).
5. Look at the differences between the pit shells node and the operational scenario node.
6. Click the Messages tab of the Operational Scenario node, and note that no messages are
shown because the node has not yet been run.
7. Right-click on the block model node, select Add and note that only Slope Set, Import
Surfaces, Reblock Block Model, and Grade Tonnage Graph are available.
8. Click on the Basic Slope Set node, and select the Description tab. Change the description of
the slope set node, and click Accept.
9. Copy and paste a branch.
Next you will run all of the nodes.
10. Select the project node, and click the Run All Nodes button.
When the nodes are running, a running man icon appears in the bottom left corner, next to the progress
bar, and the interrupt button
Licensing
When each node is successfully run, a green check mark appears next to the icon and the
Output, Report and Messages panes display results and messages.
11. Select the pit shells node, select the Mining tab, change the reference mining cost, and
Accept the changes.
12. Highlight the pit shells node, and select the Run From button.
Licensing
Whittle software has similar security to Surpac. You can install Whittle on any PC, but you need a
hardware sentinel (also known as a dongle), and a licence file to run most Whittle commands. The Whittle
licence file is a .udt file.
To import the licence file into Whittle, plug in your sentinel, click File > License Validation, select Import
license, click Browse, and browse to the .udt file.
If you do not have a .udt file yet, but you have purchased a licence, you can download the licence from
Gemcoms licence servers. To do so you will need to use your login credentials for the Gemcom Support
website www.gemcomsupport.com.
To download a .udt file, in the License Selection form, first select Request a license from a Gemcom
Support portal. Then, with your sentinel number listed, click Online, log on with your Gemcom Support
credentials and download your licence file.
Note: There is more information about licensing available in licensingTransitionGuide.pdf in the Whittle
installation directory.
After you load the licence file, two expiry dates, MaintenanceExpiry and LicenseExpiry, are written to your
C:\Users\<user>\AppData\Roaming\Gemcom\Whittle\<version>\fx.ini file. If the computers system
date is after LicenseExpiry, then no version of Whittle will work.
For licensing queries contact the support desk support-aa@gemcomsoftware.com.
Important directories
Whittle is installed by default to C:\Program Files\Gemcom\Whittle\<version>\.
The training data is installed by default to C:\Users\Public\Gemcom\Whittle\<version>\projects.
The ini file is stored in C:\Users\<user>\AppData\Roaming\Gemcom\Whittle\<version> by default.
Important directories
The programs directory contains the group of exe files which make up the Whittle software executables.
You could run these executables from a command prompt.
The projects directory contains the Marvin and other training datasets. It is recommended that you place
any project data in a separate area. When a new project is created, called for example gold, Whittle
creates a directory called gold at the location that you specify. Two project files called gold with the
Page 159 of 308
Important directories
extensions fxp and fxb are also created along with a subdirectory called working_gold. The working
directory contains all of the files generated by Whittle when work is completed.
When a node is run, it creates a series of files called FXPE_1.(ext) for the first node run, FXPE_2.(ext) for
the second node run, and so on. The files with a pr? extension are print files for the node. For example,
an optimisation print file is called pro and a reblock print file is called prr. The print file is the file displayed
in the report pane of each node.
It is not recommended to edit files in the working directory because Whittle generates and handles them
automatically. Place any files relevant to the Whittle project in the directory above (in this example) gold
not gold_working.
The fx.ini file shows what modules you are licensed to use. When you select File > Preferences, the
Master INI File field shows you the location of the ini file.
Important directories
Important directories
The label has come off the dongle. How do I know what the dongle number is?
Normally the dongle number is on a label on the dongle. If the label has come off, there is a small DOS
program available from supportdesk that can communicate directly with the sentinel and show the
number. The program is run from a DOS prompt. Contact the Gemcom support desk if you dont know
the hardlock number.
Preferences
You can store some files in fixed or comma delimited format. You can store some files as binary rather
than text to speed up processing time and reduce file size. The Spreadsheet headings setting
determines the format of the headings for the files that are output from schedules and analyses. Note that
error messages are given in terse technical codes.
System limits
To see the limits that are built-in to Whittle, choose Tools > Show System Limits.
Important directories
Note:
Important directories
The maximum number of characters in a filename including the path is 256 characters.
Important directories
6. Click Next.
9. Click Finish.
A status message is displayed showing the percentage of the model that is loaded.
After the mod file has been loaded, the define element type codes form is displayed. It is possible to
change the name of any elements but we will leave it as the default of the first 4 characters of the grade
attribute name.
The new project is loaded and some nodes are automatically created as part of the import. The project
node has been successfully run and has a green tick beside it. Some nodes such as the pit shells node
require more input before running and have a red cross.
For this course, the nodes from the slope set onwards are not required and you will delete them.
1. Highlight the New Slope Set node, and click on the scissors icon to cut that branch.
The fxp file in the title bar has an asterisk next to it, indicating that the project has been modified
but not saved. The project will be saved every time a node is run but no nodes have been run
yet.
2. Click on the floppy disk icon to save the project. The asterisk in the title bar disappears.
Model origin
Model dimensions
Block size
Total tonnes for each rock type
Total contained metal
Average mcaf, pcaf and total contained metal for a few benches
You will run the block model node which includes a grade and tonnage report.
1. Select the Block model node and change the node name in the Description tab to 4 x 5 x 4
(many parcels)
2. Check the information in the data pane of each tab.
Are the model dimensions, origin and block size correct?
3. In the Formats tab, change the Element units to gram.
4. Click Accept.
5. Select the Tonnage Regions tab.
The Def Mass (or default tonnage) applies to any blocks that are not explicitly defined in the mod
file. Def Mass is used by the Reblock and Pit Shells nodes. It is possible to only export part of
the model by constraining the export in Surpac and you would need to use default mass for this
also. If the blocks excluded are air blocks, then Def Mass should be 0. If the blocks excluded are
waste blocks then Def Mass should be the single value for tonnage to be applied to all excluded
blocks in the tab above. Remember that it is incorrect to exclude a combination of air and waste
blocks from the model due to the application of Def Mass.
Using a default density of 2.75, a block of dimensions 4 x 5 x 4 has a mass of 220 tonnes.
If some default waste, or waste tonnes not included in parcels, exist, that waste is not reported
on the Summary tab.
Compare the tonnes and contained metal with those from the Surpac report in the first exercise.
The difference in tonnage between the Surpac and Whittle model is 1 tonne from a total of 244
million tonnes, and the variance in contained metal is 0.001%.
Note that the minimum grade in the model is 0g/t and the maximum is 15 g/t as expected.
Page 173 of 308
13. Compare the bench listing for tonnes, grade and contained metal with the report produced in the
first exercise.
Has the export been successful?
Volume adjustment
If a volume adjustment attribute containing the value of 0.5 for all blocks was included in the export of this
data, the summary and report tabs are shown below.
The difference between total tonnage and parcel tonnage is not included in the Summary tab but shows
in the report tab as WASTE. Note that the item labeled TOT is only the total of the tonnes within parcels,
not the total within the model. It does not include the undefined waste shown as WASTE.
The import will abort, the node will display a red cross in the status icon and if the check data box is
chosen the following message is reported:
The specified model file (import_0.mod) could not be found
This is a fatal error that you must rectify before further processing can take place.
*** WARNING: THE MODEL FILE CONTAINED ONE OR MORE ZERO VALUE POSITIONAL CAF
FACTORS.
The optimisation will complete and produce an optimal shell that mines all blocks with positive grade (due
to zero costs). The process time is much longer than a project with no zero value cafs. For this training
data set, the optimisation process time will be 10 times longer if zero cafs are included. For this reason, it
is vitally important to read the message and report tabs taking note and appropriate action for warning
and error messages.
In Whittle v4.1, there is a fatal error on import (displayed below) and no further processing can take place.
For this reason always use the validate function in the Whittle Export function in Surpac.
Negative Sg
If a mod file containing negative values for tonnage is imported into Whittle, no warning or error
messages will occur. For the example below, the sg for all OXID blocks was set to -1.8. Whittle has
simply removed all of the parcel lines with a rock code of OXID. Processing is able to continue and a
result is produced which is incorrect. You must rectify this situation before proceeding.
Grade/Tonnage Graph
Grade/Tonnage Graph
For future comparison purposes and to further investigate the project, you will generate a grade tonnage
graph.
1. Right-click the Block Model node and add a Grade Tonnage Graph.
There are two main types of format for the graph: histogram of tonnages within specified grade
ranges or as a traditional line graph of tonnes and average grade vs cut off grade.
2. On the Definition tab, under Graph type, choose Cut-off.
3. Go to the Definition tab and next to the User defined cut-offs box, click Add Range.
You will use a lower cut-off grade of 0.5 g/t to avoid the large tonnage of waste material skewing
the graph.
4. Fill in the Range Cut-off form as shown, and click OK.
8. Select the 2
Grade/Tonnage Graph
Block Size
9. Do not click on OK because this would exit the larger graph view. Instead click on the Graph tab
to show the graph as below:
The Grade Tonnage Graph is a useful tool to view the grade distribution. You will use the graph later in
the course.
Block Size
There are several block sizes used in optimisation
It is important to complete a sensitivity analysis to understand which are the driving factors for the project,
and to reduce risk. In order to complete the sensitivity quickly, a larger block size is chosen. The smaller
number of blocks will usually give just the same shape of graph with a very small shift of absolute value.
The block size for the block value calculation is chosen relative to the smallest mining unit (SMU) to
simulate what can practicably be mined. If parcels exist at a size smaller than the SMU, make sure the
issue of ability to extract the ore at that size has been addressed or combine the parcels within the block.
The issue of SMU could affect the block size, mining cost, dilution or mining recovery.
Reblock
Reblock
In the Report tab of the Block Model node, find the number of blocks, and number of parcels, that might
be processed. This model has 1.2 million blocks and 1.6 million parcels that could be processed. The
block size is 4 x 5 x 4 and the parcel size is 2 x 2.5 x 2. The mining costs are based on an excavator and
truck fleet that can selectively mine 8 x 10 size blocks on a 4 m bench. The resulting dilution is assumed
to be 5% with a mining recovery of 100%.
A block size of 8 x 10 x 8 is chosen for the block value calculation and requires that the model be
reblocked to this size with a maximum of 1 parcel per rock type per block. The reblock will result in a total
of 300,000 blocks. The sensitivity will be performed on 12 x 10 x 8 blocks giving a total of 100,000 blocks.
In either case, the model must also be extended to the north to ensure the optimal pit does not hit the
edge of the model. First we will complete the reblock, slopes and optimisation for the sensitivity analysis
and then repeat the process for the block value calculation.
1. Highlight the block model node, right-click, and select Add > Reblock Block Model.
2. Change the description of the new reblocked block model node to 12 x 10 x 8 (1 parcel) Used
for Sensitivity.
On this node, you could add relevant notes, for example relating to the assumed fleet.
You can use the Reblock node to add expensive blocks (for example, some existing
infrastructure that would be expensive or impossible to move), extend or truncate the framework
or split/combine blocks.
If an optimal shell reaches the edge of a model, the slopes are simply extended vertically as if the
extension contained air blocks. The optimal shells should always be checked visually for
unexpected results. You can use the Adjust Framework tab to avoid the problem of a shell
hitting the edge of the model.
During the reblock program, any block on the side or bottom of the framework is replicated into
the nearest blocks in any adjacent extension. Any element content of these replicated blocks is
removed but the rock-type, pcaf, mcaf and the tonnage of the parcels is retained. One effect of
this replication is to extend any topographic surface horizontally outwards to the limits of the input
model framework.
Zone number is stored in the mod file for each block. If blocks are combined then the zone
number of the block with the lowest numerical co-ordinates is selected as the zone number for
the new block.
Reblock
Reblock
4. Click on the Adjust Blocks tab and enter the information shown below:
The blocks can be combined or split, and the resultant block is given the tonnage of the sum of the
component blocks.
When blocks are split, the tonnage and element content is shared equally between the output blocks. The
number of parcels in each output block is the same as in the original block.
Note that the maximum parcels per block is set to 1 so that grades are being averaged within each rock
type. Expressions for mcaf and pcaf can be added at this stage however you have no need to change the
values already stored in the block model.
If you are producing a model where the block size is similar to the selective mining size, you should limit
the number of parcels to one for each rock-type.
If you are producing a model for use in design or sensitivity work, you have to consider how many parcels
you need.
Assuming that the grades are different in the different parcels, it is important to retain a good description
of the grade distribution in the model as a whole, so that the tonnage processed will react realistically to
changes in cut-off. It does not have to be too detailed in a particular block . This is because the overall
mine behaviour is dependent on the grade distribution of groups of blocks rather than that of individual
blocks.
With a single element, five or ten parcels per rock-type in a block is probably enough. With two elements,
you might expect to have to square this, but five or ten is still probably enough. This is because the
tonnage discrimination is the same as for one element.
5. Click Accept.
6. Run the reblock node.
7. Click on the tabs that are available now that the node has been run.
8. Check the model dimensions, tonnages, contained metal and grades.
Remember that tonnes have been added so that the total tonnage and tonnage by rock code will be
greater than the original. What is the warning message? How many parcels does the new model contain?
What is the name of the new model?
Slopes
Slopes
Slopes are defined in Whittle by structure arcs. The relationship between blocks is contained in the
structure arcs file.
For example, in the diagram below if A is to be mined, B must be mined, however the reverse is not true.
If B is to be mined, A may or may not be mined. An arc is formed from A to B.
B
Arc
from A
to B
A
All slopes are translated into a large number of block relationships. It is wrong to assume we need an arc
from each block to every block which is above it. This is because arcs can chain as shown below.
B
If A is mined
so is C
A
Slopes
The slopes node creates a structure arcs file containing the relationships shown visually below.
Desired Slope
It is possible that the actual slope of the optimal shell is steeper than the desired slope that you enter. (it
will not be shallower). In the previous example, the structure arcs have been generated 8 benches to the
full height of the model. Because most models have many more benches, you can control the number
used to generate structure arcs. The greater the number of benches, the larger the number of structure
arcs and the slower the optimisation. The smaller the number of benches for arc generation, the greater
the slope error between desired and achieved slope.
The formula for the number of benches to search for in structure arc generation is
(max(x,y) x 8)
z
where x, y and z are the block dimensions. For this exercise, the max of x and y is 12 and the max of z is
8, so the number of benches should be 12. After running the slope node, always check the slope error
and increase the number of benches if the slope error is too large (for example > 2 degrees). A large
slope error in the optimal shell means that the actual pit design cannot achieve the tonnes, and grade
reported in the optimisation.
Whittle slopes are overall slopes. They should include an allowance for ramps and berms.
Slopes
Regardless of the definition method for slopes (rectangular regions, zone numbers or rock codes), slopes
are given by a bearing. It is very important to note that bearings are given instead of wall positions. Any
walls at right angles to the bearings, in a particular slope region, will have the given slope applied. In the
diagram below, the slope specified for a bearing of 45 degrees would be used by the program in the
positions indicated by the arrows.
Slopes
Be careful when specifying the bearings for slopes because slopes are interpolated between the bearings
as shown in the following diagram.
60
45
55
60
To ensure the walls have a constant slope, add bearings on either side with the desired slope.
60
55
45
Constant
Constant
55
45
60
Overall slopes for this project were given in the previous exercise and were the same for all bearings
within each zone.
1. Select the reblocked block model node, right click, and choose Add > Slope Set.
2. Change the description to Zone numbers, and in the Slope Type tab, select Zone numbers in
the Model file.
Slopes
3. Click on the Profiles tab and, at the top of the data pane, select Add Profile.
The new profile appears directly beneath the Add Profile button.
4. Change the slope to 38, and press ENTER.
5. Continue adding profiles until the slopes for all six zones have been defined.
Remember the formula for the number of benches to search for in structure arc generation.
(max(12,8) x 8)
8
In this example
(max(12,8) x 8)
= 12
8
6. When the profiles are complete, Accept the data.
Slopes
7. After you have accepted the data, match each slope profile to the correct zone number.
8. When all zones have the correct profile, Accept the data.
9. Click on the Check Data button to ensure the data is valid.
10. Run the slopes node.
The program generated 17.6 million arcs and took several minutes to run. A file with a .stu
extension (the structure arc file) has been created.
11. Click on the Report tab.
What are the average slope errors for the 6 zones?
12. Change the number of benches to 10 for profiles 1, 2, 4 and 5.
13. Rerun the node.
Note that the number of structure arcs has dropped to 13.1 million and the process time was reduced.
The smaller structure arc file reduces the process time of the pit shells node.
Slopes
At each Report tab, four icons are available in the top right corner of the data pane.
The first icon changes the font size of the report, the second performs a search. The third icon opens a
print dialog box, and the fourth allows you to save changes to the font size.
Be careful of the profile numbering in the Report tab. The first part of the report lists the slopes and
number of benches for arc generation by profile using 0 as the default profile number.
Directly beneath the section shown above, a plan view and the slope errors are listed for each profile
using 1 as the default profile number. In the error listing the profile numbers are increased by 1.
Slopes
Process Time
The factors that affect the process time for structure arc generation are:
Number of benches used for structure arc generation (more = slower, however this must be
considered against the slope error)
For any project, you will not know what the ideal number of arcs is (because of the factors listed above)
and the focus should be on the slope error rather than the total number of arcs.
Generally, aim for an average slope error of between 1 and 2 degrees. Anything less than 1 degree is
probably generating an unnecessarily large structure arcs file for no advantage and will slow down the
process time of both the structure arc and pit shell generation. A shallow slope angle (for example less
than 35 degrees can produce slow process times but it is not usually an option to change the desired
slope angle)
Slopes
What about blocks in Surpac that have different zone numbers in the sub-blocks?
In the mod file, the zone number is stored in the block line, NOT the parcel line. During the export in
Surpac, if a Whittle block contains blocks or sub-blocks with different zone numbers, the zone number
with the highest cumulative tonnage within the Whittle block is used.
Slopes
The previous error message is caused by the IJK definition of the regions having the same value. Whittle
interprets this situation as the regions overlapping.
You can resolve this problem by entering the value of 161 for the Min Y field of region 2.
Slopes
Assessment
Exercise 2 Assessments
What are the 2 elements to Whittle security?
Where is the ini file defined in Whittle?
What is the fxp file?
What type of tonnage will not be shown on the
Summary tab of the Import Block Model node?
What values are contained in blocks added in the
Reblock node?
Exercise 2 Assessments
What is the minimum number of parcels possible per
block?
What is the target number of blocks for sensitivity
purposes?
What is the target slope error in degrees?
Is the zone number stored in the parcel or block line of
the mod file?
Slopes
Pit Shells
Pit Shells
Revenue Factors
So far you have gathered data for costs, recovery, mining and processing limits and slopes that, coupled
with a gold price, would produce a single, optimal pit.
The previous block value formula determines the value of the block which utilises a single value for each
input and gives a single result, for example pit A shown in the following image.
Pit Shells
If the metal price increases, the value of every block within outline A increases in value or stays the
same. No block value decreases in value. Consequently, every block in outline A is still worth mining. In
addition, if we do another optimisation using the new values, the new outline, shown as B below, is
certain to include the whole of A. It may also include extra blocks that were not worth mining before, but
which now are worth mining.
Consequently, if you step the prices through a series of values, doing an optimisation for each, you obtain
a set of nested pit outlines, and this is, in effect, what the Pit Shells node in Whittle does. It multiplies all
of the prices by a series of "Revenue Factors" ranging, typically, from 0.3 to 2.0, and produces a pit
outline for each.
The reason for producing outlines for the smaller values of Revenue Factor is to produce inner pit shells
to highlight the best positions to start mining and to assist with the sequencing. The outer pits can help
determine a less risky location for infrastructure such as waste dumps or processing plant.
The outlines are usually very close together and form an almost continuous spectrum, where the change
in tonnage from one outline to the next is quite small. However, if the grade increases sharply with depth
or the ore body is discontinuous, large tonnage differences between adjacent pits can occur.
Because all the outlines conform to the pit slope requirements, it is simple to determine what sequences
are permissible when mining out a particular pit.
Pit Shells
You can enter revenue factors as a range with a start, end, and step or single values or a combination of
both. You can choose from two methods of revenue factor range generation: fixed intervals or geometric
intervals. The fixed intervals, as the name suggests, provide a constant step size in revenue factor. The
geometric intervals allow you to specify the required number of factors but will then produce more pits in
lower revenue factors and less in the higher factors (handy if looking for starter pits as well as a final pit).
This is illustrated in the following figure.
The generation of pit shells involves simple cashflow (it is not discounted) and does not take into account
the time value of money, or when a particular block is mined. The optimal shell assumes that all blocks
are mined today.
Ore Selection Method
Whittle has two methods of selecting parcels for processing and determining whether a parcel is ore or
waste. It can use cut-offs or cash flow.
When ore selection is by cut-off, ore is selected by comparing the grades of the material with precalculated processing cut-offs. If material does not satisfy the cut-offs, it is treated as waste. In the case
of multiple processes, the first eligible process is used. The grades are compared with the cut-offs of
each in turn, in the order in which they are specified in the Block Model node. For an explanation of how
multiple cut-offs are handled, see the Whittle help.
When ore selection is by cash flow, ore is selected by comparing the cash flow which would be produced
by processing it and the cash flow which would be produced by mining it as waste. If the cash flow from
processing it is higher, the material is treated as ore. If not, it is treated as waste. If more than one
processing method is applicable, the one which produces the highest cash flow is used.
In the simple cases (one process, no block co-ordinate based expressions affecting cut-off), ore selection
by cash flow produces the same result as that produced by the use of marginal cut-offs. However there
are some cases where this is not true.
Cut-off calculations are based on processing cost, element processing cost, price, selling cost, and
recovery and are calculated in a preprocessing stage before individual blocks are considered. If any of
the items are based on expressions that:
then you must use ore selection by cash flow, because the marginal cut-off, determined for the whole
mine, cannot be calculated.
In the simple case of a single process and no expressions, these two mechanisms should produce the
same result. In the case of multiple processes, where the cut-offs are not ranked, you must use ore
selection by cash flow. If you want to use ranked cut-offs, you must use ore selection by cut-off.
Pit Shells
For a simple case, such as this example where the ore selection method does not affect the destination
of a block, note that selection by cut-off includes the cut-off grade for each rock code in the report file.
This cut-off is based on the data for the reference block of that particular rock code. If pcafs are used,
which is true in our case, the cut-off grade shown in the report is NOT correct for any block with a pcaf
that is not 1.
Value mode is where the block value is either stored in the block line of the mod file or calculated from
the parcel values defined by the element indicated to store the parcel value in the user element tab. Both
ways of defining the block value require the VALU module and are not covered in this course.
When ore selection is by Profit mode, the parcel revenue element is used to determine the process path
which will return the best cashflow. In simple situations, the results are identical to ore selection by
cashflow but profit mode has the ability to be used in Stockpile and Cut-Off Optimisation (SPCO) for
multiple processes.
1. From the Slope Set node, right-click, and choose Add Pit Shells.
Note that the new node icon has a red cross and requires further input before processing.
2. Click on the Mining tab, and enter the following information:
Note that you enter a 5% dilution as 1.05. If the mining recovery were 5%, you would enter it as a factor
of 0.95. If you use both mining dilution and mining recovery, the order of application does not affect the
result because mining dilution affects the grade but mining recovery does not.
The rock type CAFs are not used because you have provided the mcaf in the model file. You have the
choice to give the costs as $/BCM or $/t. Currently it is set to $/t.
3. Click on the Processing tab, and, beside the Processing Paths section, click Add.
4. Enter the information for the HCAP rock code.
There is only one method of processing the ore, so you will call it MILL and use cut-off for the ore
selection. The reference processing cost for all rock codes is 16.79. Recovery is 0.94.
Pit Shells
5. Click OK.
6. Copy the first process definition to create entries for all the rock codes as shown below:
Note: The Recovery for FRSH is 0.9.
Pit Shells
Most of the settings in the Settings pane minimise the run time of the optimisation. Given the
speed of modern hardware, there is normally no need to do this. It is possible to optimise using
all blocks, only blocks in a sub-region, or only blocks defined in the model. In this case the first
and third options are the same, and the second is not relevant.
It is possible to include or exclude air blocks from the optimisation to speed up processing time. If
air blocks are not included, you cannot use the output in the Minimum Mining Width node.
There is also an option to choose which air blocks to include in the output file after processing.
Only the option to include all air blocks is usable with the Minimum Mining Width node.
An example of the requirement to include air blocks is as follows:
When you run a Pit Shells node, the background process periodically dumps all of the data from
memory to a temporary work file to enable a later restart. This can be extremely helpful if, for
example, a long optimization run is terminated because of power failure. The restart interval
determines how often the dump occurs. If the power source is unreliable or the run time is
expected to be very long, the restart functionality can be extremely useful.
The Optimization Control allows you to limit the number of iterations taken before results are
produced. If Maximum Iterations is selected, the result will NOT be an optimal shell.
The default setting for a new pit shells node is 86 revenue factors (0.3 to 2.0 using 0.02 steps).
This high number of revenue factors could potentially take a long time to process so use only one
revenue factor for a quick first pass.
10. Under the Produce nested pit shells heading, highlight the existing revenue factor range, and
click Delete on the right side.
11. Click on Add Single, and enter 1.0.
Pit Shells
Bench Phases are used when the mod file contains the value for the block (in this situation, the block
model is called a value model. You need an additional module of Whittle to use a value model) rather
than the metal and tonnes (as we have used). With the Value option, revenue factor ranges are not
available, so an alternative is used to produce the pits. Initially, only the top bench is considered for
optimisation. A pit optimisation is performed and if there are any blocks worth mining, this constitutes a pit
shell.
Next, a bench is added (now considering the top two benches). A pit optimisation is performed and the
result constitutes a pit shell.
Therefore, in the example below, the first pit produced is the starter pit on the top bench. The second pit
is the one which first uses blocks on the second bench. This process continues until the final pit which
uses the blocks on the last bench.
13. Because you do not have any constraints, nor any additional blocks, and you do not need to
change the CAFs or user defined elements, click Accept.
14. Change the description to make it easier to identify for example Sensitivity RF 1.
15. Run the node.
16. When the run is complete, click on the Output tab.
A brief summary appears which shows that the optimal pit at the nominated gold price has 3.4 million
tonnes of ore @ 1.63 g/t with a strip ratio of 4.31:1. Note that there was only 1 revenue factor, so the pit is
given a number of 1. If there were 10 revenue factors, they would be given numbers 1 10 in increasing
order of gold price.
Pit Shells
17. Click on the Report tab and note that tonnes and grade by bench for each pit is provided.
What are the cut off grades for each rock type? Note that these are only the cut off grades for blocks with
a PCAF of 1. What are the warning messages?
Pit Shells
Beneath the cut-off grade display, the total tonnes, tonnes processed, and grade processed are
listed by bench, starting with the highest bench (38 in this example).
18. Click on the Messages tab.
19. Find the name of the res file.
The output from the pit shells node is the res file which is the mod file with a new header section
and an extra column containing the number of the smallest pit that the revenue factor is part of.
Note that if the Use binary files option is selected in File > Preferences > File Outputs, the
extension of the results file is rzp not res. Here is an example of a res file in a text editor.
Pit Shells
20. Again click on the Output tab, and click the Open in spreadsheet button on the toolbar
The output has been saved to a csv file and opened in Excel:
22. At the data display box, select the option to view only the grades, and click OK.
Pit Shells
Tip: If you want to give the Viewer a white background, as shown in the previous image, click
Invert.
26. To check that the pit does not reach the edge of the model, select Show Topography.
Pit Shells
The pit has not reached the edge. On the western side, the extended topography is highlighted
by the flat area extending west to the edge of the model.
27. To change the colour range, click Scale.
After clicking OK, it is possible to use the Edit button to change each range if desired.
Pit Shells
Pit Shells
Next you will run the pit shells node again but using a revenue factor range of 0.2 to 2 in steps of
0.1.
32. Click on the Optimization tab and delete the single revenue factor.
33. Click Add Range, and complete the form as shown below:
Pit Shells
34. After filling in the Start Factor and End Factor boxes, click on auto next to the Number of
Factors, and then enter a Step Size of 0.1.
35. Click OK.
36. Change the node description to Sensitivity RF 0.2 2.0.
37. Run the node again.
38. Check any errors in the Messages tab.
39. Check the results in the Output tab.
The optimal shell now has a pit number of 9 (ninth in the revenue factor list) with the same values
for Rock and Ore as the previous run.
40. Click on the Report tab and note that tonnes and grade by bench for each pit is shown as below.
Pit Shells
42. Open the viewer and notice that you can display each pit, by selecting Show Pit and clicking the
up and down arrows beneath Show Pit.
MCAF definition method and complexity (more complex i.e. expressions = slower)
PCAF definition method and complexity (more complex i.e. expressions= slower)
For the first run on a project, use large blocks (~200,000 total) and a small number of revenue factors (for
example less than 20). You can use smaller blocks, and more revenue factors, in subsequent iterations.
The geometric interval definition is a good way of focusing on starter pits as well as the final pit.
Pit Shells
Negative Tonnes
If a mod file containing negative values for tonnage is imported into Whittle, no warning or error
messages are displayed and a result is generated. For the earlier example of setting all values of OXID
tonnage to negative, the revenue factor 1 pit is shown as follows:
The total and ore tonnes are reduced by the amount of OXID, and the shell is generated as if blocks
containing only OXID are actually AIR blocks. This result is incorrect.
A red exclamation mark on a yellow background next to a node means the node has aborted
unexpectedly. When this occurs, and you try to run the node again, a message like the following example
might appear in the Messages tab:
*** RESTART IS NOT POSSIBLE DUE TO THE STATE OF THE WORK FILE
^^^DELETE GENERATED FILES FOR THIS NODE AND RERUN
To enable the node to be rerun, you must first delete the work file. To do so, right-click the node, and select Delete
Generated Files.
Pit Shells
You can also use the Delete Generated Files menu command to delete all the outputs from a node and
reduce the size of the working directory. After the generated files are deleted, you cannot view the results
of the analysis until you have run the nodes again.
Operational Scenario
Operational Scenario
The Operational Scenario node defines most of the settings required for a schedule, pit by pit analysis,
or any other form of analysis. Most settings are similar to those that appear in the parent Pit Shells node
and the settings are the same as those in the Pit Shells node by default. It is the first place in the
optimisation process where time is applied by limits on mining, processing or elements and financial
parameters like discount rate, capital costs and explicit time costs.
The discount rate is used to calculate discounted cash flow analysis. This discount rate might include not
only the time value of money expected but also an adjustment for the risk or opportunity cost of the
project.
For the purpose of generating pit shells (and cut-off calculations), time costs must be implicitly included in
the mining costs, processing costs, or selling costs. However, for the purposes of cash flow analysis and
discounted cash flow analysis (which you do after entering settings on the Operational Scenario node),
it is more useful to show the time costs explicitly. In order to achieve this, you should:
Perform the optimisation (Pit Shells node) with time costs in the Mining, Processing or Selling
tabs included implicitly.
In the Operational Scenario enter the replacement capital cost
In the Operational Scenario enter the costs identified as explicit in the Explicit time costs
section
Note that the time costs that are factored into the Mining, Processing, or Selling tabs should be applied
against the item that is the limiting factor of the optimisation. Otherwise the intended time cost value will
not be correctly incorporated in the final result. Note also that different periods could have a different
limiting factor.
1. Right-click the Pit Shells node, and choose Add > Scenario.
Note that the Mining, Processing, Selling, Expressions and User Element tabs are the same
as those in the Pit Shells node with one important difference. The Operational Scenario tabs
now have a clock icon instead of the expressions icon to show that you can vary those inputs for
each period. The inputs can still be expressions or values but the input can be defined for each
period. The period of the project is defined in the original Block Model node in the Formats tab.
The period length is 1 year by default, although you can set the period length to 6 months, 3
months, or 1 month.
2. Change the Description to indicate the mining limit of 9 million tpa and processing limit of
800,000 tpa.
3. Enter the limiting values into the Limits tab.
Spider Graph
Note: The throughput factor is the relative speed of processing for a particular rock type.
For example, a rock type which is easy to crush might have a throughput factor of 1.2. You would
allow for this when applying milling throughput limits
The time cost in the optimisation was $1.5 million pa which was included implicitly as $1.88/t ore
processed. In this node, that value will be defined explicitly for the financial analysis that will
follow.
4. Enter the following information on the Time Costs tab:
Depending on the study requirements, other steps might include selecting the best cutback with
Pushback Chooser, scheduling with buffer stockpiles, modifications for minimum mining width or cut-off
grade optimisation. This course does not cover any of these options. Contact your local Gemcom office if
you are interested in an Advanced Whittle Techniques course.
Next, you will select the two pits for further work. You will then run the slopes, optimisation and graphs on
a block model for design purposes, and then export two shells for use in Surpac.
Spider Graph
Knowing which inputs a project is sensitive to, is at least as important as knowing what the NPV of the
project will be. Determining which inputs are the most sensitive allows you to review those inputs and
reduce the project risk. There is no point in spending days gathering mining cost data when the project is
insensitive to it. Sensitivities are generally tested by varying one input at a time and quantifying the effect
on a project measure. Often the project measure will be NPV or mine life or metal produced, depending
on the driver for the corporate objective. The definition of sensitive varies but one rule of thumb is a
project where the measure varies more than 25% when the input is varied 10%.
Spider Graph
Note that all the Whittle functionality used to date, is included in the foundation module. Spider graphs are
included in the Advanced Analysis Module (ADVA). To see what modules are available for the license
you are using, select Help > About.
The spider graph allows you to very quickly see the effect of multiple inputs against a user-defined
measure.
1. From the Operational Scenario node, click Add > Spider Graph.
2. Fill in the Schedule tab as shown.
Because the mine life of the final pit is relatively short, you will generate sensitivities on a final pit
of pit 9 with no cutbacks. The Schedule tab is common to all graphs and analysis nodes but not
all fields are relevant for each application. The specified case scheduling algorithm only affects
results if a result for the specified case is chosen in the Definition tab.
Next you will add the inputs you want to test.
3. On the Definition tab, click Add.
Page 217 of 308
Spider Graph
4. The Data Selector is displayed. You use this to find the variables you want to work with in this
case, to find out which variables the operation is most sensitive to.
5. Type mining capacity into the box in the lower left, and click Search.
Spider Graph
Note that items involving rock type can only be varied one at a time. For example, the processing
cost can only be varied for fresh or transition rather than all varied at once. Also note that slopes
must be run manually. For presentation in reports and flexibility, it might be easier to vary some
of the inputs manually and save the results to csv for graphing in a spreadsheet. The graphing
functionality of Whittle is not as comprehensive as a spreadsheet or dedicated graphing software.
It is suggested that you use the Spider Graph node to quickly determine the sensitive parameters
but not for presentation purposes.
This schedule assumes a top down sequence with no cutbacks. This sequence is called the
Worst case in Whittle (see discussion in next section)
8. In the Vary items by (%) box, type 10.
9. In the Set indicators at (%) box, type 15.
10. Set Value to display in output to Discounted Open Pit Value for Worst Case.
11. Run the node and then go to the Graph viewer.
Typical of a gold project the most sensitive input tested is gold price. The next sensitive inputs are
metallurgical recoveries. The project is less sensitive to mining capacity than mill capacity which is
expected in a project that is milling limited.
No inputs caused a variation in NPV of 25% or greater so the project is not particularly sensitive to any
one input.
Apart from the graph shown above, the spider graph node generates a csv file (with a single column of
data) and a print file as shown in the Report tab.
Spider Graph
It is also invalid to use a value that is based on complicated expressions or one that changes with time.
Assessment
Exercise 2 Assessments
What causes the nested pit shells in the Pit Shells
node?
Where is the reference mining cost entered in Whittle?
What number range should the mining dilution be?
What number range should the metallurgical recovery
be?
What is the cut off grade shown in the Report tab of
the Pit Shells node?
Exercise 2 Assessments
What affects the process time of the Pit Shells node?
What are the 2 ore selection methods?
What are implicit costs?
Where is the Discount Rate entered in Whittle?
What is the number for % change in value to measure
against for a sensitive parameter?
Assessment
After you have chosen the sequence, you can develop the schedule. The schedule allocates the parts of
the sequence to time periods. The schedule determines when the cashflow occurs. The optimal pit is
not known until a schedule is completed.
The 3 types of sequences available in Whittle are
1. best
2. worst
3. specified
Imagine a final pit to be mined with a series of possible smaller intermediate pits in cross section as
shown in the following diagram:
In the worst case scenario, no intermediate pits are mined. The final pit is mined top down, bench by
bench.
Top Down
Mining
This sequence is almost always feasible (and simple) to mine. Over the life of the mine, waste is mined
early while most of the ore is mined towards the end. The high cashflows occur at the end of the mine life
and hence give the worst NPV.
In the best case scenario, each pit is mined out before the next pit is started.
Depending on how many revenue factors were chosen, the sequence is not normally feasible (cutback
width less than minimum mining width, access is a problem, vertical rate of advance is not achievable).
The waste and corresponding ore are mined in a similar time frame so the cashflows are brought forward
and occur closer to the start of the mine life.
Inner
shell
out
The specified case is somewhere between the best and the worst case. You specify which cutbacks to
mine to gain the advantages of higher cashflow of the best case while still being practical and feasible.
The NPV lies somewhere between that of the best and worst case.
The Schedule tab is available in all of the graph and analysis nodes. The Final Pit pane is
unavailable because the Definition tab will override the input. It is possible to define a specified
schedule but because you are yet to choose the starter and final pit, there is not a specified case.
Because you are not using a specified case, the sections beneath Specified Case Scheduling
Algorithm and Specified Case Pushback Definition are irrelevant.
3. Leave the defaults for the Schedule tab, and select the Definition tab.
The default values are shown for the definition. You will graph all pits so leave the default of pits
1 to 0 in a step of 1.
4. Delete the values to display until only the following variables remain.
Note that there is no difference in the mine life of the best and worst case until the mine life is greater
than 4.3 years. The mine life of the pits around revenue factor of 1 is about 4 years and there are several
pits of mine life between 0.4 and 2 years which could be starter pits. The difference in cashflow between
best and worst case for pit 9 is around $3.5 million. The actual achieved schedule will be somewhere
between the best and worst case.
7. Go to the Graph Tab and click the magnifying glass icon.
8. To make the graph easier to read, select the Display Preferences tab and change the Scaling
Factor for the Y Axis to 1,000,000.
9. To return to the graph, click the Graph tab.
nd
10. In the 2
Notice how flat the value curves are around the revenue factor of 1. Even for the worst case, the
difference in NPV between pits 7 and 10 is less than $1 million in a total of $42 million. Conversely, the
Page 227 of 308
difference in ore tonnes is 500,000 t (0.6 years milling) of 3 million total tonnes. If the corporate objective
was maximum mine life, you might choose pit 10. If the corporate objective was lower risk through
decreased exposure to variation in gold price, or lower geotechnical risk through shorter mine life, you
might choose pits 7 or 8.
Also note an obvious jump in stripping ratio between pits 11 and 12. Looking at a starter pit, there are
several jumps in stripping ratio between pits 2 and 4.
The actual cashflow is somewhere between the best and worst case. The highest NPV for the worst case
scenario is pit 8. For this exercise, you will investigate pit 8 for the final pit, and pit 4 for the starter pit
because it has a low stripping ratio for 1.7 years of milling.
The output of a Pit by Pit Graph node is a csv file of the results shown in the Output tab, and a print file.
12. Click OK to exit the Graph Viewer.
13. Go to the Summary tab, and display the summary for pit 8.
The pit utilisation in the summary above is 72.4%. The formula for pit utilisation is shown below:
Rejected material is material for which there is a processing path (that is, a method/rock-type
combination) but which is not processed because its grade is not high enough. The pit utilisation is a
combination of metallurgical recovery and the contained metal in parcels beneath the cut-off grade.
Hence the cut off grade for that pit, in the pit by pit graph, is lower than that used in the pit shells node
and the ore tones are greater in the pit by pit graph.
In the case of a pit with revenue factor > 1, the reverse is true. That is, the ore tonnes are greater in the
pit shells node.
Note that the Waste (other) value is negative. Negative waste tonnages and strip ratios are possible due
to the way dilution is applied. When a mining dilution is specified, the ore tonnes within the block are
increased while the metal content remains the same. To ensure the total tonnes are correct, the waste
tonnes are decreased by the amount that the ore tonnes are increased. On benches where there is little
or no ore, the waste tonnes are negative.
There are several ways to deal with this issue.
The preferable way to handle dilution is in the block model in Surpac. You can apply dilution
skins or use changes in block size to simulate dilution.
You can set very small negative tonnages to zero.
You can set larger negative waste tonnes to zero, and decrease the ore tonnage by the negative
amount to ensure integrity of the total tonnes. In practise, benches of all ore are not likely to have
additional waste mined because this would mean mining outside the design.
A fourth option to avoid using dilution factors is to use reblocking in Whittle to simulate the effects of
dilution. If the reblocking function is used with a maximum number of parcels of 1, the amount of material
above a certain cut off grade can change. To quantify the changes due to reblocking you will use a grade
tonnage graph.
Tonnes
Grade
Grams
Ounces
4x5x4
3,519,177
2.15
7,575,319.5
267,211
12 x 10 x 8
3,730,928
1.74
6,508,896.2
229,594
211,751
-0.41
-1,066,423
-37,617
6%
-19%
-14%
-14%
Variance
%
The change in block size has resulted in a 6% increase in tonnes, and a 14% decrease in contained
metal. When applying the dilution factor in Whittle, the tonnes are increased, the contained metal stays
the same, and the average grade decreases.
Given that you have also included a dilution factor in the optimisation of 5%, and that the project is
sensitive to the dilution factor, it may be prudent to review the method for defining dilution and choose to
use either the dilution factor or the reblocking. The 12 x 10 x 8 block size was chosen to determine
sensitivities in this case, and it was not intended for final results. However when you do generate the final
pit optimisation, you will investigate the amount of dilution represented in the reblock, and then decide
whether or not to use a dilution factor.
Schedule Graph
Schedule Graph
The Schedule Graph node generates graphs and csv output for the defined inputs by period. The node
also produces an msq file which is similar to the mod file with 3 extra columns.
First, you will generate a simple worst case schedule of pit 8 to define the lower threshold.
1. At the Operational Scenario Node, add a Schedule Graph.
2. At the Schedule tab, choose a Worst case schedule for pit 8.
3. Accept the defaults in the Definition tab, and Run the node.
The Schedule Graph node has generated a csv file of results shown in the Output tab, a graph and a
summary for the total pit.
Schedule Graph
Note the large amount of waste mined in the first year in the graph below.
The node has also produced a mining sequence file (msq file). The file has the same format as the mod
file except:
Schedule Graph
In the block header line the zone field has been replaced by the period, fraction and pushback
fields
In the parcel line the destination of np (not processed) or MILL is added
Looking at the previous graph, the mill is filled each year but the mining capacity is not reached in
any year although the first year is close. This operation certainly has potential to defer waste
stripping and increase the NPV.
4. To investigate which benches are mined in each period, add a Bench Schedule beneath the
Schedule Graph.
5. Accept the default values and run the node.
The Output tab shows that only one pushback is mined, that is pit 8 top down. The bench RL column
slowly decreases as the period increases until the end of year 5. The negative waste tonnes are also
seen in years 4 and 5.
The only output from the Bench Schedule node is a csv file containing the previous results.
Schedule Graph
9. Run both the Schedule Graph and the Bench Schedule nodes.
The bench schedule output (below) shows 8 pushbacks (pits 1 8) mined consecutively. Each pushback
is mined from the top bench to the bottom bench, and then returns to the top bench for the next
pushback.
Schedule Graph
A comparison of the two graphs shows the difference in the timing of waste mining. The graphs have
been resized so that the Y axis scales are similar. This highlights the large amount of waste that the
Worst case schedule mines in the first year.
Using a final pit of pit 8, these two schedules represent the worst and best extremes. Now you will start
using specified schedules with a final pit of pit 8 but with 1 starter pit.
Assessment
Exercise 2 Assessments
What is a best sequence?
What is a worst sequence?
Describe a fixed lead sequence
Where is the Discount Rate entered in Whittle?
Which of the best and worst sequence is usually
achievable?
Which has the highest NPV?
Exercise 2 Assessments
Assessment
Milawa
Milawa
Scheduling
will make a
difference
Need to find a
practical
schedule
Schedules
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Milawa
Milawa
Pit
by
Pit
1
2
3
4
etc
Specified schedule
Milawa
Milawa Algorithm
Ph 1
Ph 2
Ph 3
Milawa
Milawa Algorithm
Allows full user control
Max
benches
per
period
Max lead
Min lead
Milawa Rules
Mining in the next cutback can never advance ahead of the current
cutback
Applies any user defined controls over the separation between the
cutbacks and the number of benches that can be mined per period.
You can mine part of a bench in a period
All mining in a cutback is assumed to occur at the same rate
Milawa
The algorithm then EITHER calculates NPVs of balances for
possible schedules and searches the domain of feasible solutions.
Milawa NPV
Milawa
Milawa
Milawa Balanced
Throughput Rates
Tonnage (m)
40
30
Processing
Mining
20
10
0
1
9 10 11 12 13 14 15
Period
Milawa
NPV vs Tonnage
NPV and Tonnage
1000
500
Waste
Ore
Best
Worst
900
800
400
300
500
400
200
Tonnage (m)
600
300
200
100
100
0
0
1
11
16
21
26
31
36
Pit Number
3.0
2.5
O re
35
Cus
30
2.0
25
1.5
20
15
1.0
10
0.5
5
0
0.0
1
Ye a r
10
11
12
13
14
15
16
Coppe r He a d Gra de
(%)
W as te
40
Tonne s (m )
NPV (m)
700
Milawa
B est S equence
3.0
40
W as te
35
O re
30
Cus
2.5
2.0
25
1.5
20
15
1.0
10
0.5
Coppe r He a d Gra de
(%)
Tonne s (m )
45
5
0
0.0
1
10
11
12
13
14
15
16
Ye a r
3.0
Tonne s (m )
2.5
O re
35
Cus
30
2.0
25
1.5
20
15
1.0
10
0.5
5
0
0.0
1
10
Ye a r
11
12
13
14
15
16
Coppe r He a d Gra de
(%)
W as te
40
Milawa
M ila w a B a la n c e d S e q u e n c e
3.0
40
W as te
35
O re
30
Cus
2.5
2.0
25
1.5
20
15
1.0
10
0.5
5
0
0.0
1
10
11
12
13
14
15
16
Ye a r
M ila w a N P V S e q u e n c e
50
W as te
45
O re
Tonne s (m )
40
35
30
25
20
15
10
5
0
1
7
Ye a r
10
11
12
Coppe r He a d Gra de
(%)
Tonne s (m )
45
Milawa
M ila w a B a la n c e d S e q u e n c e
Tonne s (m )
90
80
W as te
70
O re
60
50
40
30
20
10
0
1
10
11
12
Ye a r
Conclusions
Milawa
Milawa
Milawa is a linear algorithm that utilises three routines.
The first routine takes a set of variables and generates a feasible schedule from them.
The second is an evaluation routine which calculates the NPV or balance for an individual schedule.
The third routine searches the domain of feasible schedules for the one which has the highest NPV or
best balance. The routine also has logic built in to decide when to stop searching.
The Milawa algorithm does not generate and evaluate all feasible schedules, because the number of
feasible schedules in any analysis is extremely large. Instead, it strategically samples the feasible domain
and gradually focuses the search (without necessarily narrowing it) until it converges on its solution.
Milawa NPV
Use this mode to specify that you want Whittle to find a sequence that improves the NPV of the project
within the mining and processing constraints.
In this case, both a mining and processing limit exists and the Milawa NPV algorithm might not reach one
limit (similar to the worst and best case scenarios)
Milawa Balanced
Use this mode to specify that you want Whittle to find a sequence with improved throughput balance. In
this mode, the Milawa algorithm seeks to maximise the usage of production facilities early in the life of the
mine instead of maximising NPV. This option will only be of use if you specify at least two of the possible
mining, processing, or selling limits.
For a quick and simple schedule, you will use the Fixed Lead algorithm with a lead of 0 to produce a
schedule for pits 4 and 8.
1. Copy the Pit 8 Best schedule graph and bench schedule nodes, and rename the copy to Pit 4
and 8 fixed (no lead).
2. Fill out the Schedule tab as shown below. Ensure that you have added the first pit (pit 4) under
the Specified Case Pushback Definitions section.
Milawa
To save time you will also test which pit produces more desirable results as a starter pit.
3. Before running the new nodes, copy them again to produce a series of fixed lead schedule
graphs and bench schedules using pits 3 and 8; 5 and 8; and 6 and 8, similar to that shown
below.
4. When all of the nodes have been added, run all nodes.
The graphs show that as the starter pit number increases, the waste is mined later in the total mine life.
Compare the NPVs of the various scenarios. Pits 4 and 8 have a similar NPV of $43.3 million as Pits 5
and 8. The cashflows are the same, however the total tonnage mined each year is very different.
Page 250 of 308
Milawa
Despite the very similar cashflows, the pit 4 and 8 option is much more attractive in terms of mining
capacity. All of the schedules so far have fully utilised the mill while there is spare mining capacity in
many years.
You will now use Milawa to generate the highest NPV schedule for pits 4 and 8.
5. Copy a Schedule Graph and Bench Schedule branch, and call it Pits 4 and 8, Milawa NPV.
6. Complete the form as shown, and run the branch.
7. Compare the bench schedule and the graphs of the fixed lead and NPV schedules for pits 4 and
8.
For this project, the top-down mining of pits 4 and 8 is the highest NPV option.
8. Now create a Milawa Balanced schedule for pits 4 and 8, and compare the NPV and the timing
of mining pit 8.
Milawa
The Balanced option tries to fill both mining and milling capacity each year. Because there is a large
mining capacity, the mining limit has not been reached in any year but all of the waste mining has been
brought forward.
The bench schedule shows that pit 8 (pushback 2) is mined in every year in an attempt to reach the
mining limit.
Milawa
Finally you will investigate the effect of adding a lead to a fixed lead schedule.
9. Copy the pit 4 and 8 fixed (no lead) branch and change the lead to 4 as shown below:
With no lead, pit 4 is completely finished before the waste stripping of pit 8 occurs. With a lead of 4
benches the waste of pit 8 is brought forward into year 1 and the bottom of pit 4 is pushed back from year
2 to year 3. This change in schedule reduces the project NPV by $0.5 million.
Milawa
You have explored the basic scheduling options using the chosen pits of 4 and 8. Next you will
look at changing the mining capacity to achieve a more realistic result. In this case, you will
assume the mill capacity is fixed because the mill is already built and the company does not want
to spend any money at the start of the mine life. The following process could just as easily be
used to test the financial impact of increasing the milling capacity.
10. Select the Operational Scenario node ($ symbol) and make a copy of the branch.
11. Rename the new Operational Scenario to 5 Mtpa mining, 800ktpa processing as shown:
Milawa
Investigate the changes caused by limiting mining to 5 Mtpa. Firstly comparing all of the fixed schedules
with no lead, the pit 4 and 8 combination still gives the highest NPV. The 5 Mtpa scenario makes about
$150 k LESS than the 9 Mtpa scenario. The output of the 4 and 8 fixed schedule shows that the mill is not
quite filled in year 2 due to the mining limit being reached. Whittle will mine within a period until any one
of the limits has been reached, and then moves to the next period.
As an interesting aside, do you know how much it costs for your mill to sit idle? An actual Australian hard
rock mine had a loss in revenue recently of $6000/hour! With this in mind, leaving a mill idle is usually not
the preferred option. Many operations are processing limited.
The fixed schedule with no lead is no longer the highest NPV scenario with the NPV scenario giving $80k
more cashflow. The graphs look very similar as shown below.
Milawa
The bench schedule for the NPV scenario shows that a small amount of waste from pit 8 is mined in year
1 which exposes more ore to be mined in year 2 as shown below:
In the 5 Mtpa scenario, the Balanced option gives an NPV very close to the NPV option (<$20 K
difference). The Output tab shows that the mining limit is reached in the first 2 years (although the milling
tonnage is close to the 800 ktpa limit) while the milling limit is reached in years 3 and 4.
In this way you could add more scenarios to gain the best balance between mining limit, achieving milling
limit, and NPV.
Reporting
Reporting
So far you have used the schedule graph, pit by pit graph and bench schedule nodes to choose a starter
and final pit. You can also use these nodes for reporting purposes. Typically you need a certain format for
the csv output which you then import into a spreadsheet program for further processing. This section will
focus on the Milawa NPV schedule for pits 4 and 8 using 5mtpa as the mining limit.
1. Select the Output tab of the Milawa NPV Schedule Graph.
Reporting
The possible selections are divided into groups depending on their activity basis:
You might want to report ore tonnes and grade by rock code, waste tonnes and strip ratio. You can
choose ore tonnes from Material processed.
Reporting
4. Fill in the Data Selector as shown, and click Add to selection list.
Brackets (<>) indicate that Whittle needs additional input. In this case, there are 4 rock types.
You must choose one or all of them.
5. Instead of selecting each individually, select the Select all items box to add all the possible
choices.
6. Now for the grade of the 4 rock types, choose Material processed, Grade of <element> mined
in <rock type> input into processing, and click Add to selection list.
Note:
Be careful because the first items listed in the Element and Grade category are
total contained units rather than grades.
Reporting
7. Click OK.
Alternatively, instead of selecting from the lists in the Data Selector, you can use the Search field at the
bottom of the form. When you type the word grade, all the items whose description include the word
grade are listed in the middle section of the form.
Reporting
8. For waste tonnage mined, go to the Material mined as Waste category, and add Tonnage of
waste rock to the selection list.
Reporting
Reporting
Reporting
What is the difference between tonnage in ground selected for processing and tonnage input to
processing?
In the Data Selector, there are two items that are similar for reporting ore tonnes tonnage in ground for
processing (TG spreadsheet code) and tonnage input to processing (TI). In our case TG is less than TI
although its possible for the reverse to occur as well. These items vary due to mining dilution and mining
recovery. The tonnage of material in the ground above cut-off is around 761 ktpa. However you have
used a 5% dilution which increases the tonnage put into the crusher for processing (waste included as
dilution) to 800 ktpa.
If you had given a value for mining recovery of 0.95, for example, and no mining dilution, ore tonnage in
ground is 842 ktpa while the processed tonnage is 800 ktpa due to ore accidently taken to the waste
dump.
What is the difference between the various options for reporting bench number or RL?
If you search for the word bench in the Data Selector, you see the following results.
Reporting
There are only 4 items above that actually report bench number, and they are only applicable to specific
nodes. The table below summarises which item is used in which node.
Description
Code
Pit by pit
graph
Schedule
graph
Bench
schedule
BENCH/FI
Yes
No
No
BENCH/ST
No
Yes
No
BENCH/EN
No
Yes
No
BENCH
No
No
Yes
Final bench
Period
bench
used
(bench report only)
Report templates
When you have a format that you wish to use again, it is possible to save the definition as a template and
use it in other projects.
1. Highlight the Schedule Graph node of the Pit 4 and 8 fixed, Milawa NPV node, and right click.
2. Select Other > Convert to Template.
3. In the Template description field, type Standard physicals, and click OK.
4. Click OK in the Maintain Graph Templates form.
Reporting
5. Under the Operational Scenario node, add a Pit by Pit Graph for a Milawa NPV schedule of
pits 4 and 8 with the same definition as the schedule graph above.
6. Edit the Definition tab so that it contains the following data.
Note:
Reporting
Still at the Pit by Pit graph, you will add some financials to the report.
8. Change the definition to include total discounted cost of mining, total discounted cost of
processing and discounted open pit value.
Process time
12 x 10 x 8
1.5 min
8 x 10 x 8
5.5 min
8 x 10 x 4
17 min
4. At the Adjust Blocks tab, combine the blocks to give a block size of 8 x 10 x 8, and only allow 1
parcel per rock type per block.
You can add many nodes in a branch, and then run all nodes rather than add each node and
process at each step. The easiest way to add the nodes is to copy an existing branch. However,
in this case, you cannot add the slope node and other nodes in the rest of the first branch without
running the reblock node first.
5. Run the reblock node and check for error messages.
How many blocks exist in the reblocked model?
6. Select the slope node under the reblocked block model node above, right-click, and choose
Copy Branch.
7. Select the 8 x 10 x 8 (1 parcel) block value node, and choose Paste.
It is now possible to run all of the nodes in the copied branch, but it might be better to stop after
the slopes node to check the slope error.
8. Delete the 9Mtpa Operational Scenario branch because it is not needed.
9. Copy the grade tonnage graph from the old reblocked block model node to the new one.
The new block size represents a 5% increase in tonnes and a 16% decrease in contained metal
so you will change the mining dilution fraction to 1.
10. On the Pit Shells node, on the Mining tab, change the Mining Dilution Fraction to 1.
11. Run all nodes.
The following table shows how tonnes or ore, grade, and ounces of gold differ depending on the
block size (using a cut-off grade is 0.75).
Item
Tonnes
Grade
Grams
4x5x4
3,522,312
2.15
7,575,320
8 x 10 x 8
3,730,928
1.74
6,508,896
Variance
208616
- 0.41
- 1,066,424
5.6%
-19%
-16%
Investigate the effect of using the smaller block model. Which two pits are you going to select?
Page 269 of 308
res file (created at the pit shells node and includes pit number)
msq file (created at the schedule graph node and includes pushback and period mined)
You will perform all 3 options of exporting results to Surpac to see what data they contain.
3. In the Export file box, type a name for the dxf file.
4. In the Selected pitshells box, type 8.
Note that although it is possible to create a single dxf file with all 19 pit shells, the resulting file is
not useful, or usable, in Surpac.
Do NOT select the Block centroids option which would cause the dxf to be formed using
centroids. If the Show topography box is ticked, the dxf is extended to the edge of the model
space. If it is not selected, the dxf ends where the pit reaches the surface and intersects the AIR
blocks. If the Contour Plot option is selected, the dxf consists of lines representing contours
rather than the usual surface that converts to a dtm.
Note that the dxf filename must have the extension of dxf in the previous form so that Surpac can
recognise the file as the correct file type and allow the import.
5. Export pit 8 again using the Block centroids option, and save to a different file name.
The following image shows the difference the brown dtm is the block outline option and the
blue dtm is the centroids option.
You can open the dxf files in Surpac, and then save them to a dtm file. Further processing might include
creating contours of the pit shell (using Contour Extract) or writing values to a block model using the pit
shell dtm.
2. Choose the directory for the res file to be written and enter names for both the res and par files.
3. In Surpac, change the working directory to the directory where you exported the res file.
4. Highlight the block_val.res file in the navigator and drag into Graphics.
In this case, the res file will NOT be read into an existing block model but a brand new Surpac block
model (mdl file). The Import Whittle file form is displayed. The Parameter file name field has the same
value by default as the res file you are opening. Enter a different file name if the parameter file has a
different name to the res file. The Model file name field is used only if the pit number is being read into
an existing block model. Leave the Model file name field blank to create a new block model.
The block size, origin and other information from the parameter file is displayed. Check that these are
correct for the Whittle res file. The reference block coordinates fields are used if the res file block with
IJK of 1,1,1 is not located at the origin coordinates that are displayed.
If the Retrieve Parcel Information box IS selected, the form is displayed as follows.
Note that the X origin has changed from the original Surpac block model of 9950 E because of
the reblocking. The data that is stored as block lines (rather than parcel lines) is converted
directly to attribute information mcaf, pcaf and block tonnage. Storing the block tonnage as an
attribute is different to most Surpac block models, so you will have to define reports carefully.
The IJK value of each block is also stored to allow easy reference between Whittle and Surpac.
The pit number is also unique for each block in Whittle and has been directly transferred to the
Surpac block model. The attributes relating to grades and tonnages are split by rock code.
In the original res or mod file, there were up to 999 parcels per block. In the resulting Surpac
block model, all of the parcels of the same rock code within a block have been combined to allow
usable data. It is important to remember that it is possible that some of the parcels from the same
rock code in the same block could be sent to the mill (that is, they are ore parcels) while other
parcels in the same block are sent to the waste dump. The imported res file will NOT show what
destination the parcels went to and will not allow reporting of processed and non-processed
statistics.
10. Use the identify block value function to look at the values of a block outside the largest pit (the
top right block in plan view in this case).
The pit number is 20 for this block. Looking at the results of the Pit Shells node in Whittle, the highest pit
number was 19. This block is outside of all pits found in the Pit Shells node. If a block had a pit number
of 4, it would be included in pits 4 and above but not pits 3 or less.
The sg of the block is 0.73 which indicates that it included parcels of AIR and parcels of rock (HCAP for
this block).
Note that the IJK value (60090038) is concatenated and only contains 8 digits when the expected number
of digits is 3 x 3 or 9. To interpret the IJK value, read from right to left and know that any leading zeroes
are dropped. The above value is an IJK of 60,90,38.
11. Add a graphical constraint of pit_num<=8 to display the blocks in pit 8.
You could extract sections, as string files, from the block model to use in pit design.
The tonnage of pit 8 on the Output tab of the Pit Shells node was 16,400,467 tonnes.
4. Click OK.
5. In Surpac, drag the msq file into Graphics to create a new mdl file.
Page 281 of 308
The pushback and period information could have been read into an existing mdl file similar to the
res file and with the same restrictions.
There are a series of grade and tonnage attributes that are a combination of rock code and
destination For example, the oxid_rejected_tonnage attribute reports OXID tonnes sent to the
waste dump. Unlike the res file import, the msq file import does not combine all OXID parcels.
Instead the import process in Surpac combines all parcels of the same rock code AND
destination. This allows reporting by ore/waste split and also reporting by destination.
In the previous example, there are 4 attributes related to when the block is mined, grouped in
pairs of fraction and value. The mined period fraction is the amount of the block mined during the
period (because blocks can be mined in multiple periods). A value of 0.43 indicates 43% of the
block is mined and that means that 43% of each parcel is mined as well as 43% of the total
tonnage. The mined period value is the actual period number the block is mined in. For example,
a value of 2 means the block is mined in the second period.
The a pair and b pair indicate that no block took more than 2 periods to mine. For example, a
particular block might have been mined first in period 3 and again in period 4. So the
mined_period_a_value would be 3 and the mined_period_b_value would be 4.
7. Display the block model
Note that there are blocks larger than the Whittle block size of 10 x 8 x 8. The original msq file
only contains blocks inside the largest pushback used. The blocks outside of the largest pit have
all been assigned the background values and have been superblocked by Surpac. The res file
did not behave this way because it contains all the blocks from the mod file so the IJK was
unique for each block and did not allow superblocking.
8. Use Identify Block Values to look at the values of a superblock.
In the following example, the block size is 160 x 128 x 128 which is obviously larger than the msq
file block size of 10 x 8 x 8.
The lower levels of the inner pit (pushback 1), and the upper levels of the cutback (pushback 2), have
been mined.
Page 284 of 308
The resulting reports are shown below with the corresponding tonnages from the schedule graph node.
To report grades from an imported msq file, you must add an extra attribute to the model.
2. Add a calculated attribute which multiplies tonnage by grade for fresh material sent to the mill.
This attribute defines total metal for the fresh/mill combination.
3. Create a block model report using the format file shown below.
The third line is calculating a new value on the fly which divides total metal by total tonnage to give the
average grade for fresh/mill material.
The average grade reported from Whittle for fresh material sent to the mill is 2.093 g/t.
You could create reports in a similar way for tonnages and grades by rock code, destination, RL, and
pushback using the appropriate constraints. Reports by period would require an additional set of
calculated attributes. This procedure is not covered in this document.
Pit Design
After the pit shells or schedule are imported into Surpac, you can complete a pit design. Due to time
restrictions, pit design is not covered in this course but is part of the Surpac for Open Pit Engineers
course.
The final step in the strategic mine planning process is to use the pit designs (based on the Whittle pit
shells) for further scheduling and analysis in Whittle. This requires a method of exporting pit designs from
Surpac and importing into Whittle. There are two possible methods:
Use the Import Surfaces node in Whittle to import DTMs (or other surfaces) that represent the
pit design.
Page 288 of 308
Export the pit designs from Surpac as a .PIL file (that is, a pit list file) and, in Whittle, merge the
pit list file with the .mod file.
1. Select the 8 x 10 x 8 block model node in Whittle, right-click, and choose Add > Import
Surfaces.
2. On the New Import Surfaces node, select the Surfaces tab.
3. Click Choose surfaces.
6. Click Close.
7. Clear the check box for pit 2 so that only pit 1 is displayed.
8. Scroll down in the left pane so that you can see the Display Settings, and Parameters sections.
9. Click each of the Display Settings, and Parameters sections to expand them.
In Display Settings, you can control the display of topography. The Parameters section displays
details of the block model.
10. Drag the Topography transparency slider
11. Use the mouse to rotate and pan in the Surface Chooser.
12. Click OK to import the two surfaces into Whittle.
Note: The order in which you selected the surfaces determines the pit number of each pit.
13. Click Accept, and run the New Import Surfaces node.
You have now specified that Whittle will use these two surfaces, pitdes1 and pitdes2, as
pushbacks. You can now enter costs and limits in an Operational Scenario node, and add a Pit
by Pit Graph to determine the NPV, tonnage, or the other measures or movements that you
would achieve if you mined those pushbacks as specified on the Operational Scenario node.
14. Copy the Operational Scenario node from the Pit Shells node to the New Import Surfaces
node.
15. Copy the Pit by Pit Graph from the first Operational Scenario node to the new Operational
Scenario node.
16. Run the new Operational Scenario node and Pit by Pit Graph node.
17. Select the Output tab of the Pit by Pit Graph node.
18. Select the New Import Surfaces node, and use the 3D Viewer button
surfaces in 3D.
to display the
In this exercise, the optimisation block model has an extra attribute added called pit which you will fill
using the two pit designs provided (pitdes1.dtm and pitdes2.dtm).
First, you will output the reblocked 8 x 10 x 8 block model to Surpac because 8 x 10 x 8 is a practical
mining width
1. In Whittle, select the 8 x 10 x 8 block model, right-click, and choose Other > Export Files.
2. In the Export files dialog box, set the options to export an export8.par and an export8.mod to
the 4_Design folder.
3. In Surpac, choose Block model > Block model > Import > Whittle file.
4. Select the export8.mod block model, and click Apply.
7. In Surpac, display the two pit designs (pitdes1.dtm and pitdes2.dtm) from the 4_Design
directory.
Note that no part of pitdes1 is outside pitdes2 at any time. If it were, the data would be invalid,
and you should correct the data before performing the PIL export.
Page 296 of 308
10. Choose Block model > Estimation > Assign Value, and fill the pit attribute with the values of 2
for the final pit (pitdes2), and 1 for the starter pit (pitdes1). Ensure that the blocks above the
topography have a value of 0.
11. Check that the pit attribute has the values you intended.
Tip: You can use Colour by Attribute to colour the block model by the value of the pit attribute.
The block model is ready for exporting as a .pil file.
12. Under the Block Model menu, select Export > to Whittle and fill out the form as shown.
13. Ensure that the Generate Mod File box under the Mod File tab is NOT selected.
14. Click Apply.
15. Constrain the export to pit > 0.
An example of a .pil file is shown in the following image.
You can merge a .pil file with a .mod file in Whittle to produce a .res file. You can then use the
.res file to generate an operational scenario node and perform all of the scheduling and analysis
that is possible under an operational scenario. The .pil file must be merged with a .mod file of
the same block size and dimensions.
16. In Whittle, select the 8 x 10 x 8 block model node and, in the Messages tab, find the name of the
.mod file (FXPE__35.mod).
Note: The name of the corresponding .par file is the same (except that the file extension is .par).
You must click Edit to fill in the Model and Parameters boxes.
19. Click Check Data to check that the data is valid.
20. Click Run.
Page 301 of 308
The resulting node has a blue version of the Pit Shells icon. It contains tabs that are the same as
those of the Import Block Model node.
You can now use the Imported Reblocked Pitshells node in the same way as a Pit Shells
node.
21. Run the Imported Pit Shells node.
22. Use the 3D Viewer to show data for the Pit attribute in the XY plane.
Review Exercise
Review Exercise
Heres an example to work through by yourself. Imagine that I am your boss on site and I have asked you
to perform a very quick, first pass assessment of a project. You can use the validated block model from
Surpac for the geological data. The rest of the known information is shown in the following table:
Item
Type
Value
Unit
Metallurgical Recovery
95
Metallurgical Recovery
Fresh
92
Au price
All
850
$/oz
Slopes
40
degrees
Slopes
Fresh
46
degrees
Mining cost
Hardcap/oxide
0.9
$/t mined
Mining cost
Tran
1.15
$/t mined
Mining cost
Fresh
1.30
$/t mined
All
2.5
Discount Rate
All
Mill throughput
All
1 million
tpa
Hcap
12.50
$/t proc
Oxide
13
$/t proc
Tran
14
$/t proc
Fresh
16
$/t proc
I require a report showing NPV, mine life in years, ore tonnes and grade by rock type, waste tonnes by
rock type, total discounted mining cost, total discounted processing cost, discounted time costs, total
discounted cost in $/oz, mining capacity, break even RL and first year with positive cashflow.
Although we have a contractor for mining purposes, they will increase costs if the annual mined tonnage
varies more than 25%.
Index
Review Exercise
Index
*
* 170
3
3D Viewer 206
9
999 parcels per block 144
99999 blocks 84
A
adding
attributes 99
constraints 90
slopes to Surpac block model 113
Adjust Framework tab 182
adjusting framework 55
air blocks 101
density 107
displaying 99
algorithms 36
arcs 185
Assign Value command 98
Assign Value vs Block Maths 101
assigning value to attribute 98
asterisk 170
attributes 85
adding 99
assigning values to 98
deleting 96
viewing 105
au_cut15 attribute 85
audit trail 143
averse to risk 18
B
batters 186
bearing and slope 58
bearings 187
bench phases 203
bench schedule 234
berms 186
best case sequence 224
Block line in mod file 49
Block Maths vs Assign Value 101
block model
attributes 85
audit trail 143
dimensions 168
in Whittle 49
report 91
validating import 171
block size 53
12 x 10 x 8 182
D
Data pane 153
Data Selector 258
Def Mass 172
deleting attributes 96
density of air blocks 107
dilution 61, 200, 229
dimensions 168
directories 159
displaying
air blocks 99
block model audit trail 143
Index
Review Exercise
dongle 158
DXF export 270
processing 215
system 163
litho attribute 85
marginal cut-off 45
Max Lead 241
maximisation of reserves 19
Maximum Benches per Period 241
maximum parcels per block 184
MCAF 136
formula 140
METAL 66
Milawa 238
Milawa Balanced 239
Milawa NPV 239
Min Lead 241
mine planning 25
minimum mining width 54
mining
dilution 61
recovery 62
mining cost adjustment factor 136
mining costs 116
table 87
mining dilution 200
mining sequence files 281
mod files 49
exporting 144
model offset 182
modules licensed 165
msq files 281
reporting 286
F
Floating Cone 36
folders 159
Format Headers 92
framework 55
fresh 87
fxb file 159
fxp file 159
FXPE_1 160
G
Geometric Grouping 92
geotechnical attribute 86
grade
negative 102
non-zero 102
grade attribute 86
grade tonnage graph 179
graphical constraints 89
NASCO 17
Navigation pane 154
nested shells 70
neutral to risk 17
nodes 154
np 234
NPV 14
NPV curve 21
null values
checking for 99
in weathering attribute 99
hardcap 87
hardlock 158
I
IJK 50
Import Surfaces node 289
importing pit designs 288
installation directory 159
J
JORC Code (2004) 27
K
Keep Blocks Partially in the Constraint option
93
L
Lerchs-Grossman 37
licensing 158
limits
mining 215
Page 306 of 308
Index
Review Exercise
ramps 186
reblocking 182
recovery 62
RECOVERY 66
reference block 67
reference mining cost 140
reference processing cost 142
removing constraints 102
report templates 265
reporting 257
from imported msq file in Surpac 286
from imported Res file in Surpac 279
res files
exporting to Surpac 273
reporting 279
tabs 155
templates 265
time costs 120
where to record 215
Index
Review Exercise
version 165
viewing block attributes 105
waste 45
waste dumps position 198
weathering attribute 85
workflow 76
working directory 160
worst case sequence 223
V
validating block model import 171
VALUE 66
value calculation block size 268
X
XYZ vs YXZ 144