Help Stope Optimiser

2 Slice Method Page 1 of 30
2 SLICE METHOD
2.1 Stope-Shape Frameworks General
The Slice Method frameworks are sub-categorised as:
 Vertical; XZ or YZ orebody orientations.

 Horizontal; XY or YX orebody orientations.
 Transverse; special case of XZ or YZ orebody orientations, that optimise shapes in both the transverse and vertical directions.
The Slice Method XZ|YZ|XY|YX framework orientations can also be mapped or visualised with a generic naming convention related to the axis direction:
 U-axis is the primary "strike" direction of the stope-shapes (i.e. length),

 V-axis is the secondary "vertical" direction of the stope-shapes (i.e. height for vertical/width for horizontal),
 W-axis is the tertiary "transverse" direction (i.e. width for vertical/thickness for horizontal). This is the stope-shape axis that the annealing process optimises for the Slice
Methods.
The first letter of the framework orientation is the U direction and the second is the V direction. So for example with the XZ orientation, X is the U-axis and Z is the V-axis.
Slice frameworks for Vertical and Horizontal orientations are further sub-categorised as:
 Regular frameworks, where both axes use fixed intervals (the U-axis intervals and V-axis intervals are fixed) - like having fixed section spacing and fixes level spacing for
vertical orientations or fixed strike sections and fixed wall spacing for horizontal orientations.
 Irregular frameworks; where either one of the U or V-axis intervals is variable. For example, like having 4x25m levels and 1x15m sill pillar level for a mine block covering
115m vertical extent which is repeated, or like having 15m primary and 20m secondary stopes along strike (for vertical orientations).
 User Defined frameworks; where coordinates are provided to define long section (U, V) dimensions of the stope-shape geometry. The coordinates can represent either
rectangular shapes (orthogonal), or trapezoidal or quadrilateral shapes (non-orthogonal).
Irregular Slice Method frameworks can also use control strings for defining variable level elevations (gradients for vertical orientations) or variable side-wall spacing – likened to
topographic contours (for horizontal orientations).
The framework and stope geometry are described in the following terms:
 stope orientation plane, a two dimensional plane defined by the framework orientation (XZ, YZ, XY or YX)
 stope face, the (fixed) U and V dimensions of the stope-shape
 The tube, a three dimensional volume defined by the stope face and the framework extents in the W direction (transverse). The seed-slice, seed-shape and stope-shape are
defined within the extents of the tube. The Slice Method optimisation is applied within the individual tubes, independently, and each tube is assigned a unique quadrilateral or
QUAD number.
Figure 2-1 Tube Definitions
Figure 2-1 Tube Definitions depicts the tube shape within regular and irregular stope-shape frameworks for the Slice method. The controls for stope-shape geometry for each case
will be explored in more detail in subsequent sections.
Figure 2-2 Slice Method Stope - Shape Framework Hierarchical Structure summarises the stope framework hierarchical structure for the Slice methods. The controls for stope-
shape geometry for each case will be explored in more detail in subsequent sections.
mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-2 Slice Method Stope - Shape Framework Hierarchical Structure
2.2 Stope Frameworks

The Slice Method can be applied in all four stope-shape framework orientations, vertically for (XZ|YZ) and horizontally for (XY|YX). The correct framework orientation to apply is
dictated by the orebody orientation within the block model.
The stope sections (U-axis) may be regularly spaced or irregularly spaced. At the same time, the stope levels (V-axis) may be regularly spaced, irregularly spaced or irregularly
spaced with variable gradient.
The faces of the stope-shapes produced are sectional outlines defined by four points. For orebodies with vertical orientation this will be two on the floor and two on the back. For
orebodies with horizontal orientation this will be or two on each of the stope end faces.
The points lay in the stope-shape UV-axis plane and the projection of the face is either a rectangle or a trapezoid where the opposite sides are parallel. Rectangular and
trapezoidal shapes are special cases of 4 sided polygons. These are commonly referred to as quadrilaterals in SSO. The quadrilaterals form a tube-shape when extruded in the
transverse direction representing the stope-shape W-axis.
2.2.1 Vertical Orientation Frameworks
Examples of the Slice Method vertical orientation frameworks are depicted in Figure 2-3 and Figure 2-4 Slice Method Framework - Vertical YZ
Figure 2-3 Slice Method Framework - Vertical XZ
Figure 2-4 Slice Method Framework - Vertical YZ

2.2.2 Horizontal Orientation Frameworks
The Slice Method horizontal orientation frameworks are depicted in Figure 2-5 and Figure 2-6. The selection of YX over XY will be dictated by the ability to have greater control
over roof and floor angles along the V-axis.
Figure 2-5 Slice Method Framework - Horizontal XY
Figure 2-6 Slice Method Framework - Horizontal YX
2.2.3 Transverse Section Orientation Frameworks
The Slice Method transverse section orientation frameworks are depicted in Figure 2-7 and Figure 2-8. The sublevel height and location are optimised independently between
sections. This method is suited to gentle/shallow dipping, thicker orebodies.
Figure 2-7 Slice Method Framework - Transverse Section XZ

Figure 2-8 Slice Method Framework - Transverse Section YZ
2.3 Framework Optimisation

The selection of start location for levels or sections and spacing for levels and/or sections are key design choices for mine planners. The following describes the options available
to optimise these.
2.3.1 Framework Iteration
The stope-shape framework can be floated in the stope orientation plane (using defined step increments) to optimise the start location for both the level and section without
changing the dimensions of the level interval (V-axis dimension) or section spacing (U-axis dimension). The stope-shape framework can then be further refined by also changing
the level and/or section spacing dimensions.
For framework optimisation the stope-shape minimum and maximum size and increment are supplied for the axes specified by the stope orientation plane, and the step size for the
origin shift along the same axes.
The initial framework specification defines the extent of the volume to be considered. To minimise the number of combinations to be considered, the stope-shape size increment
should be a sub-multiple of the stope dimension. The origin shift should be a sub-multiple of the stope sizes. For example, the stope sizes might be 20-35 in increments of 5 and
the origin step size is set to 5. As the run-time for a large set of increment combinations might be prohibitive, a single test scenario should be run to estimate the likely runtime for all
combinations. E.g. nominate a single stope size and no origin step to ascertain the run-time for a single iteration.
The “output_best_case” option will only output the stopes for the single framework that generates the best set of stopes. The “output_all_cases” option outputs all combinations
of the stope size and framework increment to the output files (if supported by the vendor data store). The “output_all_cases” and “output_best_case” options are mutually exclusive
(i.e. one or the other may be selected, but not both at the same time).
For the Transverse Section method there is a more restricted optimisation procedure whereby the user provides a set of alternative stope heights, and the best combination of
sublevel intervals and stope heights is chosen.
2.3.2 Zone Iteration
If a zone iteration field is defined in the block model and each zone is spatially separate, then the framework position and extent will be optimised for each zone. The set of
possible zones is nominated by a list. These zones can be different lodes, pods or even independent mine areas.
Zone Iteration allows the flexing of the level start location, level interval, and level offset and/or section start location, section interval, and section offset between multiple zones.
The “iterate” function evaluates all combinations of level spacing and offsets of the level start location, and reports the results as a sorted list to guide the design choice. The user
has the option of outputting the best case.
Where the orebody is made up of discrete ore zones, each is evaluated independently, and the sorted list is reported by ore zone. This feature can be useful when there are a
significant number of zones to be evaluated, for example where the orebody has many pods.
The runtime for a large number of combinations can be prohibitive. A single scenario with all required zones should be run to estimate the likely runtime for all combinations. (I.e. a
run without zone iteration)
Figure 2-9 shows the result of optimising (in the sectional orientation) the level location for two different lodes with a fixed sublevel interval. This technique could be extended to
allow a range of sublevel intervals.
Figure 2-9 Zone Iteration - Sectional Orientation Example

Figure 2-10 shows the result of using zone iteration to optimise the section start location for multiple lodes. The iteration is in the horizontal orientation and there is a fixed section
interval.
Figure 2-10 Zone Iteration - Horizontal Orientation Example
2.4 Control Strings

Control-strings provide additional constraints on the shape or position of a stope. This can be dictated in one of two ways:
(i.) Where the string projection in the UV plane is used to define the face shape, or
(ii.) Where the string position in the W dimension controls the position of the stope.
An example of (i.) would be to allow gradients on levels, and an example of (ii.) would be to control the position of stopes and pillars.
The best location for these control-strings will only be known after a preliminary run without control-strings has been completed.
A multiple optimisation run approach should be used to guide the optimal placement of the control-strings. In the first pass, a regular optimisation can be completed without the
control-strings. The first pass result is then used to guide the location of the control-strings, and a second optimisation run is conducted to define the stope-shapes that honour the
control-strings. Alternatively in cases where level development data already exists, control-strings may be created from that information.
Control-strings can be supplied for each of the framework orientation cases as detailed below.
2.4.1 Vertical Orientation Frameworks
Control-string functions for XZ|YZ orientation frameworks that extend along the U-axis.
2.4.1.1 GRADIENT
The gradient control-string is used to define the gradients of levels along the orebody strike axis (variable V-axis). The strings would typically be used for orebodies with an
extensive strike length, such that the difference in elevation from the level access point to the level strike extremity is significant. The gradient control-string will change the
framework geometry, as shown in Figure 2-11.
A stope face is created where a pair of gradient strings intersects adjacent sections. Ideally the control-strings should have the same orientation as the stope-framework U-axis.
Care is required to ensure that adjacent strings extend to intersect all sections. This will avoid unintended trapezoidal shapes, or unintended gaps, as shown in Figure 2-12 OreDev
Control-String Example.
Figure 2-11 Gradient Control-String Example
2.4.1.2 OREDEV
The OreDev control-string is used to define level layouts on fixed elevations (horizontal gradient) using development centrelines. A stope cannot be created unless its’ floor is
located on a control-string, and a stope floor cannot be located on more than one control-string.

OreDev strings do not change the framework, and should be added as an option to a second-pass run. The strings are typically used to control the location of stopes and
pillars from section to section. They define practical level layouts by constraining the transverse lateral extents of stope-shapes for parallel lodes (i.e. W-axis direction). The strings
can also be used to constrain the strike extents (U-axis) to say remove strike outliers.
The seed / annealed stope-section-ends lying at floor level must straddle the string in order to be accepted. The strings ideally run along the floor of the stopes (but if above, they
will be projected to the floor position) and assume horizontal levels. The intent of this control-string is to remove transverse pillars and transverse outliers. They may be used for
dropping out irregular pillars. This function may combine transverse stopes, creating more practical but less optimal stope-shapes. An example is provided in Figure 2-12
Figure 2-12 OreDev Control-String Example
2.4.1.3 OREDEV-GRADIENT
The OreDev-Gradient control-string combines both the GRADIENT and OREDEV functions. Note however this function does not currently allow strings with different gradients for
the same long-section strike extent – hence only one string can be used. Consequently, only one stope in the W-axis is possible.
2.4.1.4 SECTION
The section control-string is used to define gradients of levels for the Transverse Section XZ|YZ cases (special case of sectional frameworks). It should be used as a second pass
process to apply gradients to levels and would be used after the initial transverse section optimisation identifies the approximate sublevel position and height for that section.
2.4.2 Horizontal Orientation Frameworks
Control-string functions for XY|YX orientation frameworks, like the vertical orientation frameworks, extend along the U-axis.
2.4.2.1 CONTOUR
This control-string is equivalent to GRADIENT for XY|YZ – it defines the stope-shape side-wall positions (variable V-axis).
2.4.2.2 HORIZON
This control-string is equivalent to OREDEV for XY|YX – the floor and roof must straddle the string (in the vertical sense i.e. W-axis) to be included.
2.4.2.3 HORIZON-CONTOUR
This control-string is equivalent to OREDEV-GRADIENT for XY|YX - where a string can define both the gradient and the development centreline. The stope “connects” these strings
to locate the roof and floor.
2.5 Rotated Frameworks

2.5.1 General Introduction
Frameworks, like the geological block models, can be defined as Un-rotated or Rotated. For most cases, users deal with un-rotated frameworks which are easier to understand and
faster to process due to simpler geometric calculations.
An example of where a rotated framework may be utilised is where an orebody may have a strike orientation of 300 o/120o azimuth and dip sub-vertically 70o to the east. The most
logical un-rotated stope-shape framework orientation to select, in this example case, would be the Slice Method Vertical XZ framework, such that stope annealing occurs in the
transverse W-axis direction (i.e. in the north-south azimuth with bearing of 0/180o). However, this stope-shape framework may be re-oriented to be more parallel with the primary
strike-axis of the orebody such that W-axis annealing occurs at right-angles to the orebody (e.g. azimuth bearing of 30/210o). This example could also be further complicated
whereby the block model framework has also been rotated such that its orientation is more favourable for geological interpolation purposes.
The block model rotations may be the same or different to the stope-shape framework rotations. The rotated block model is defined in local coordinates and world coordinates via
rotation parameters. The following describes the various cases.
2.5.2 Rotated Framework Combinations
The orientation of the stope-shape framework is defined and operates independently of the block model framework orientation.
It is always advised that the location and extents of a rotated framework be checked visually against the mineralised portion of a block model to ensure that stopes can be
generated where they are expected.
The stope-shape framework and block model combinations (as illustrated in Figure 2.13) are reported in the run-time logfile (with verbose output) as:
 Neither the block model nor the stope-shape framework are rotated (MODROT=1).
 The block model and stope-shape framework are both rotated, and have the same rotation definitions (MODROT=2).
 The block model and stope-shape framework are both rotated, and the axes are parallel but offset (MODROT=3). Note that this case also requires the rotations of model and
framework to be identical.
 One r both of the block model and stope-shape framework are rotated, but do not have equal rotation angles (MODROT=4).
Block model and stope-shape framework rotations have impacts on runtime performance (especially for the MODROT=4 case).
Figure 2-13 Model - Framework Combinations

For un-rotated cases, the stope-shape frameworks implicitly use world coordinates [X, Y, Z] i.e. both the level coordinates and section coordinates are defined in the world
coordinates.
For un-rotated cases, the world coordinates define the framework extents (minimum and maximum) and the step sizes are entered directly to the user interface. This defines the x,
y and z extents and the number of steps based on the step size.
Rotated frameworks use local coordinates [X’,Y’,Z’] and world coordinates [X,Y,Z] to specify the framework origin point (both local and world). The coordinates for the local
framework origin are typically set as [0, 0, 0] or even [0, 0, “mine_level_elevation”], and use of the world coordinates for the local framework origin should be avoided for simplicity.
A specific requirement is that the rotation point and the origin of the framework must be coincident.
There is no standard among mining software suppliers for conventions for specifying block models. Specification of origins, rotations, angle conventions, cell and sub-cell size and
indexing, and even absent data values and modelling of undefined cells differ. Each mining software supplier provides a software API (Application Programming Interface) to AMS
to directly and efficiently read the proprietary model formats.
The specification of the framework is equivalent to specifying a model origin, extent, cell size and rotation for a block model.
The AMS Stope Shape Optimiser internally adopts the conventions of the CAE Studio software, and each vendor convention is mapped to the CAE Studio convention to run the
engine. The goal of the vendor interface is to accept framework parameters in conventions that are familiar to the user of that mining software and convert these to the conventions
used by the Stope Shape Optimiser. Inspection of the XML parameter files supplied to the Stope Shape Optimiser will show that a conversion has taken place to make the
appropriate transformations between the mining software conventions and those of the Stope Shape Optimiser.
The stope-shape framework reference point in the XML is defined in world coordinates at the bottom left-hand-side corner of the stope-shape framework [x0, y0, z0]. The origin and
dimensions are then redefined in local coordinates [xr0, yr0, zr0]. The order of the axes selected for rotation and the rotation angles are defined in Table 2.1.
The stope-shape framework orientation [U,V,W] axes correspond to the world [X,Y,Z] for the un-rotated case. However, for the rotated case, the stope orientation [U,V,W] axes
correspond to the local framework grid coordinates [X’,Y’,Z’], not the world framework coordinates [X,Y,Z]. The stope-shape framework is specified in the [U,V,W] grid.
With an irregular framework, parameters like section location and sublevel elevation should be provided in local coordinates. If no stopes are output it is likely that the coordinates
were supplied in world rather than local coordinates.
If strings or wireframes are supplied to the Stope Shape Optimiser, then these will need to be in world coordinates. All output from SSO is in world coordinates, but internally the
Stope Shape Optimiser works in local coordinates.
The key elements to consider for rotated models are illustrated in Figure 2-14. The world coordinates are identified as [X,Y,Z] and the local coordinates as [X',Y',Z']
Figure 2-14 Rotated Framework Parameters - Vertical XZ Orientation Plane

2.5.3 Rotation Parameters
The parameters entered for rotated frameworks are summarised in Table 2.1.
Table 2.1 Rotation Parameters
Parameter Description of Parameter
The first rotation angle clockwise in degrees, around axis ROTAXIS1. It must lie
ANGLE1
between -360.0 and +360.0. A value of zero indicates no rotation.
The second rotation angle clockwise in degrees, around axis ROTAXIS2. It must lie
ANGLE2
between 360.0 and +360.0. A value of zero indicates no rotation.
The third rotation angle clockwise in degrees, around axis ROTAXIS3. It must lie
ANGLE3
between -360.0 and +360.0. A value of zero indicates no rotation.
The axis around which the first rotation angle will occur. 0 for no rotation, 1 for X
ROTAXIS1
axis, 2 for Y axis, 3 for Z axis.
The axis around which the second rotation angle will occur. 0 for no rotation, 1 for X
ROTAXIS2
The axis around which the third rotation angle will occur. 0 for no rotation, 1 for X
ROTAXIS3
The X coordinate of known point in both systems, in the un-rotated coordinate

X0
system.
The Y coordinate of known point in both systems, in the un-rotated coordinate
Y0
system.
The Z coordinate of known point in both systems, in the un-rotated co-ordinate
Z0
system.
XR0 The X coordinate of known point in both systems, in the rotated coordinate system.
YR0 The Y coordinate of known point in both systems, in the rotated coordinate system.
ZR0 The Z coordinate of known point in both systems, in the rotated coordinate system.
Below is an example of the XML syntax for a rotated framework, located in the XML <Control> Section. This specifies an XZ stope-framework that is rotated -30o and has the effect
of orientating the stope-framework’s primary strike axis azimuth to 60o :
<stope_shape_framework>
<cell_specification xmorig="0" ymorig="0" zmorig="0"
xinc="25" yinc="1410" zinc="20"
nx="101" ny="1" nz="40" />
<rotation_specification rotaxis1="3" rotaxis2="0" rotaxis3="0"
angle1="-30" angle2="0" angle3="0"
x0="26829" y0="10733" z0="280" />
<model_evaluation_orientation plane="xz" />
</stope_shape_framework>
Note in particular that the cell_specification number of cells [nx, ny, nz] can only be set using whole integer values, decimal or fractional vales are not permitted.
It is good practice to visually confirm that the stope-framework has been oriented as intended relative to the deposit block model using your specific vendor mining package. For
CAE Studio, this can be done by creating an empty test block model using the framework specification parameters, and viewing this test model in conjunction with the deposit block
model.
2.6 Stope-Shape Framework Extents Guidelines

Table 2.2provides guidelines for defining the stope-shape framework parameters for the different orthogonal stope-framework orientations. The extent of the stope-shape
framework volume is defined by [(nx*xinc), (ny*yinc), (nz*zinc)] from a local origin [xmorig,ymorig,zmorig].
For irregular stope-shape frameworks, where the stope dimension vary in the U and/or V direction, the extent of the stope-framework volume is determined by the values supplied
for [(nx*xinc), (ny*yinc), (nz*zinc)], even though the notional intervals [xinc,yinc,zinc] for a regular framework will be superseded by an irregular definition when defining the tube
shapes (as illustrated in Figure 2-1). In the direction where the stope-dimension is variable (i.e. the transverse anneal direction), specify NX=1, NY=1 or NZ=1.
N is the number of intervals in the primary “strike” direction (U, the stope-shape U-axis), M in the number of intervals in the secondary “vertical” direction (V, the stope-shape V-
axis), and P is the number of intervals in the transverse “width” direction (W, the stope-shape W-axis).
Table 2.2 Stope-Shape Framework Orientation and Extents Summary
Framework Orientation U-axis “strike” V-axis “height” W-axis “width” NX NY NZ

Vertical XZ X axis Z axis Y axis N 1 M
Vertical YZ Y axis Z axis X axis 1 N M
Horizontal XY X axis Y axis Z axis N M 1
Horizontal YX Y axis X axis Z axis M N 1
Transverse Section XZ X Z Y axis N P 1
Transverse Section YZ Y axis Z axis X axis P N 1
Prism XZ X axis Z axis Y axis N P M
Prism YZ Y Z X P N M
2.7 Stope-Shape Geometry

2.7.1 Relationship between Stope Frameworks and Stope-shapes
For the Slice Methods, the stope-shapes are constrained to the tube geometry (as illustrated in Figure 2-1), when extruded in the transverse direction (W-axis). The face or
projected wall shape will be defined by the four corners in the wall plane (i.e. hangingwall/footwall or near/far wall for vertical orientation, roof/floor wall for horizontal orientation).
The shape of the tubes can be rectangular or trapezoidal (and later releases will be able to defined with full quadrilaterals).
The stope-shape tube volumes are defined by the framework options and optional control-string combinations. These are generally described in the following sub-sections.
2.7.2 Vertical Orientation Stope-Shape Tubes
The vertical orientation is best suited to narrow vein, tabular and lenticular geometry with narrow to moderate thickness and moderate to sub-vertical dip. For an overview diagram
of deposit morphology, see Figure 1-2 . Option codes are used to identify the various stope-shape types, a summary of vertical codes appears in Table 2.3 below.
Table 2.3 Vertical Orientation Stope-Shape Tubes Summary
Stope-Shape Code Stope-Shape Framework Tube End-Face

VR (VR-O, VR-D) Vertical Regular Rectangular
VIL, VIS Vertical Irregular (Level, section) Rectangular
VIVSG, VIFSG Vertical Irregular Trapezoidal
VUR, VUT Vertical Irregular User Rectangular or Trapezoidal
2.7.2.1 (VR) Vertical Regular Framework (Rectangular Stope-Shape Tubes)
The regular stope-shape tubes are rectangles having a fixed section spacing (U-axis) and a fixed level interval (V-axis). The option code for regular stope-shape tubes is VR.
Special case optimisation options (with option code VR-O) are available that can:
 Optimise the stope framework origin (Optimise Origin - U|V origin) whereby the U|V origin is floated on axes step increments.
 Optimise the section spacing (Optimise Section _U dimension). The U dimension (a single optimised dimension) can also be optimised independently by ZONE (e.g. a
different V dimension for each independent lode, pod or mine area).
 Optimise the level interval (Optimise Level - V dimension). The V dimension (a single optimised dimension) can also be optimised independently by ZONE (e.g. a different V
dimension for each independent lode, pod or mine area).
 Combinations of all of the above.
The control-string case (with option code VR-D) can be used to constrain the stopes in the W and U axes. It is typically applied as a secondary run refinement to remove impractical
stopes such as strike outliers and/or impractical stopes in the transverse extents i.e. W-axis direction (e.g. where the mineralisation is erratic for parallel lodes which results in
stopes jumping across from one lode to the other or randomly bridging the parallel lodes). This control-string option does not change the U|V dimensions, and cannot be used with
the optimisation options.
The regular framework rectangle stope-shape tube options (with option codes) are depicted in Figure 2-15below:
Figure 2-15 (VR) Vertical Regular – Rectangular Stope-Shape Tubes
2.7.2.2 (VIL VIS) Vertical Irregular Framework (Rectangular Stope-Shape Tubes)
The irregular rectangle framework can have either:
 Framework with regular section spacing (U-axis) and irregular level intervals (V-axis - option code VIL), or
 Framework with regular level interval (V-axis) and irregular section spacing (U-axis - option code VIS).
The control-string cases (with option codes VIL-D and VIS-D) are used to constrain the stopes in the W and U axes. It is typically applied as a secondary run refinement to remove
impractical stopes such as strike outliers and/or impractical transverse extents in the W-axis direction (e.g. where the mineralisation is erratic for parallel lodes which results in
stopes jumping across from one lode to the other or randomly bridging the parallel lodes). This control-string option does not change the U|V dimensions.
The irregular framework rectangular stope-shape tube options (with option codes) are depicted in Figure 2-16below:
Figure 2-16 (VIL, VIS) Vertical Irregular - Rectangular Stope-Shape Tubes

2.7.2.3 (VIVSG, VIFSG) Vertical Irregular Framework (Trapezoidal Stope-Shape Tubes)
Control-strings with variable gradient (and lateral extents in the U-axis direction) are used to form the irregular level intervals (in the V-axis). The irregular trapezoid framework can
either have:
 Framework with regular section spacing (U-axis) and irregular level intervals (V-axis - option code VIFSG), or
 Framework with regular level interval (V-axis) and irregular section spacing (U-axis - option code VIVSG).
In addition to irregular level intervals (using the variable gradient control-strings) the control-strings can optionally be used to remove impractical stopes such as strike outliers
and/or impractical transverse extents in the W-axis direction (e.g. where the mineralisation is erratic for parallel lodes resulting in stopes jumping from one lode to the other or
randomly bridging the parallel lodes). Note that the OreDev-Gradient control-string for Version 2 only allows defining one long section trace per level in the stope orientation plane
projection. As such, for multiple parallel orebodies it can only produce one stope in the W-axis direction (unlike the OreDev control strings for option codes VR-D, VIL_D or VIS_D
which can allow multiple parallel lodes but for a horizontal plane only (i.e. multiple stopes in the W-axis for each horizontal level).
The irregular framework trapezoidal stope-shape tube options (with option codes) are depicted in Figure 2-17below:
Figure 2-17 (VIVSG, VIFSG) Vertical Irregular - Trapezoidal Stope-Shape Tubes
2.7.2.4 (VUR, VUT) Vertical Irregular Framework (User Defined Rectangular or Trapezoidal Stope-Shape Tubes)
Specification of the stope-shape dimensions can be individually specified within the extents of the framework. This may be of particular use for orebodies that require irregularly
located/shaped crown, rib or sill pillars or mines that require exclusion zones around say an irregular surface feature such as a meandering river, variable depth lake, and irregular
pit excavations or require exclusion zone around a key piece of infrastructure such as shafts, workshop, etc.
Stope-shapes can be specified with coordinates for a set of Rectangular or quadrilateral shapes, as depicted in Figure 2-22. A rectangle stope-shape is defined by coordinates
(umin, umax, vmin, vmax). A trapezoid stope-shape has two opposite parallel sides and is defined by the coordinates (umin, umax, v1min, v1max, v2min, v2max).
Both methods define the stope-shape in long section, and the projection of this shape in the W-axis direction (i.e. transverse direction) forms the face of the stope-shape "tube".
The irregular framework user-defined rectangular or trapezoidal stope-shape tube options (with option codes) are depicted in Figure 2-18below:
Figure 2-18 (VUR, VUT) Vertical Irregular – User Defined Rectangular or Trapezoidal Stope-Shape Tubes
2.7.3 (HR) Horizontal Orientation Stope-Shape Tubes
The horizontal orientation is best suited to narrow vein, tabular and lenticular geometry with narrow to moderate thickness and moderate to sub-horizontal dip. For an overview
diagram of deposit morphology, see Figure 1-2. Option codes are used to identify the various stope-shape types, a summary of horizontal codes appears in Table 2.4 below
Table 2.4 Horizontal Orientation Stope-Shape Tubes Summary

HR (HR-O, HR-D) Horizontal Regular Rectangular
HIW, HIS Horizontal Irregular (Wall, Section) Rectangular
HIFSC, HIVSC Horizontal Irregular Trapezoidal
HUR, HUT Horizontal Irregular User Rectangular or
Trapezoidal
2.7.3.1 (HR) Horizontal Regular Framework (Rectangular Stope-Shape Tubes)
The regular stope-shape tubes are rectangles having a fixed section spacing (U-axis) and a fixed width interval (V-axis). The option code for regular stope-shape tubes is HR.
Special case optimisation options (with option code HR-O) are available that can:
 Optimise the stope framework origin (Optimise Origin _ U|V origin) whereby the U|V origin is floated on axes step increments.

 Optimise the section spacing (Optimise Section - U dimension). The U dimension (a single optimised dimension) can be optimised independently by ZONE (e.g. a different
AU dimension for each independent lode, pod or mine area).
 Optimise the wall interval (Optimise Wall - V dimension). The V dimension (a single optimised dimension) can also be optimised independently by ZONE (e.g. a different V
dimension for each independent lode, pod or mine area).
 Combinations of all of the above.
The control-string case (with option code HR-D) can be used to constrain the stopes in the W and U axes. It is typically applied as a secondary run refinement to remove impractical
stopes such as strike outliers and/or impractical steps in the W-axis direction (e.g. where the mineralisation is erratic for parallel lodes and resulting in stopes jumping up and down
from one lode to the other or randomly bridging the parallel lodes). This control-string option does not change the U|V dimensions, and cannot be used with the optimisation options.
The regular framework rectangle stope-shape tube options (with option codes) are depicted in Figure 2-19 below:
Figure 2-19 (HR) Horizontal Regular – Rectangular Stope-Shape Tubes
2.7.3.2 (HIW, HIS) Horizontal Irregular Framework (Rectangular Stope-Shape Tubes)
The irregular rectangle framework can have either:
 Framework with regular section spacing (U-axis) and irregular wall intervals (V-axis – option code HIW), or
 Framework with regular wall interval (V-axis) and irregular section spacing (U-axis – option code HIS).
The control-string cases (with option codes HIW-D and HIS-D) are used to constrain the stopes in the W and U axes. It is typically applied as a secondary run refinement to remove
impractical stopes such as strike outliers and/or impractical steps in the W-axis direction (e.g. where the mineralisation is erratic for parallel lodes and results in stopes jumping up
and down from one lode to the other or randomly bridging the parallel lodes). This control-string option does not change the U|V dimensions.
The irregular framework rectangle stope-shape tube options (with option codes) are depicted in Figure 2-20 below:
Figure 2-20 (HIW, HIS) Horizontal Irregular - Rectangular Stope-Shape Tubes
2.7.3.3 (HIFSC, HIVSC) Horizontal Irregular Framework (Trapezoidal Stope-Shape Tubes)
Control-strings with variable strike direction (and lateral extents in the U-axis direction) are used to form irregular wall intervals (in the V-axis). The irregular trapezoid framework can
either have:
 Framework with regular section spacing (U-axis) and irregular “contour-like” intervals defined by variable strike-direction control-strings (V-axis – option code HIFSC), or
 Framework with irregular section spacing (U-axis) and irregular “contour-like” intervals defined by variable strike-direction control-strings (V-axis – option code HIVSC).
In addition to irregular wall intervals (using the variable strike-direction control-strings) the control-strings can optionally be used to remove impractical stopes such as strike outliers
and/or impractical steps in the W-axis direction (e.g. where the mineralisation is erratic for parallel lodes resulting in stopes jumping up or down from one lode to the other or
randomly bridging the parallel lodes). Note that the Horizon-Contour control-string for Version 2 only allows defining one plan trace per level in the orientation axis plane. As such,
for multiple parallel orebodies it can only produce one stope in the W-axis direction (unlike the Horizon control-strings for option codes HR-D, HIW_D or HIS_D which can allow
multiple parallel lodes but for a constant strike plane only (e.g. multiple stopes in the W-axis for each constant strike plane).
The irregular framework trapezoidal stope-shape tube options (with option codes) are depicted in Figure 2-21 below:
Figure 2-21 (HIFSC, HIVSC) Horizontal Irregular - Trapezoidal Stope-Shape Tubes
2.7.3.4 (HUR, HUT) Horizontal Irregular Framework (User Defined Rectangular or Trapezoidal Stope-Shape Tubes)
Specification of the stope-shape dimensions can be individually specified within the extents of the framework. This may be of particular use for orebodies that require irregularly
located/shaped pillars or require an irregular exclusion zone around say a shaft, workshop, and development access or require irregular stopes with say variable section end -walls
so that they are at right-angles to the “contour”.
Stope-shapes can be specified with coordinates for a set of Rectangular or quadrilateral shapes, as depicted in Figure 2-22. A rectangle stope-shape is defined by coordinates
(umin, umax, vmin, vmax). A trapezoid stope-shape has two opposite parallel sides and is defined by the coordinates (umin, umax, v1min, v1max, v2min, v2max).
Figure 2-22 Definitions for Rectangular and Trapezoidal Shapes

Both methods define the stope-shape in plan and the projection of this shape in the W-axis direction (i.e. transverse direction) forms the face of the stope-shape "tube".
Figure 2-23 (HUR, HUT) Horizontal Irregular – User Defined Rectangular or Trapezoidal Stope-Shape Tubes
2.7.4 Transverse Section Orientation Stope-Shape Tubes
The transverse section orientation is best suited to moderate plunging orebodies with lenticular, pipe or massive geometry and moderate to wide thickness that are amenable to
benching type mine methods where the stopes walls for adjacent sections abut. It is typically a “special case” vertical orientation option. For an overview diagram of deposit
morphology, see Figure 1-2. Option codes are used to identify the various stope-shape types, a summary of transverse codes appears in Table 2.5
Table 2.5 Transverse Section Orientation Stope-Shape Tubes Summary

TIL, TISL Transverse Irregular Rectangular
TIFS, TIVS Transverse Irregular Trapezoidal
2.7.4.1 (TIL, TISL) Transverse Section Irregular Framework (Rectangular Stope-Shape Tubes)
The rectangle transverse section framework can have:
 Regular width (equal) transverse-section spacing (U-axis) with the position of the horizontal levels (V-axis) optimised within each transverse section and the transverse
section length (W-axis) also optimised (potentially with angled walls but typically vertical). The option code for this option is TIL.
 Irregular width transverse-section spacing (U-axis) with the position of the horizontal levels (V-axis) optimised within each transverse section and the transverse section
length (W-axis) also optimised (potentially with angled walls but typically vertical). The option code for this option is TISL.
Figure 2-24 (TIL, TISL) Transverse Section – Irregular Rectangular Stope-Shape Tubes
2.7.4.2 (TIFS, TIVS) Transverse Section Irregular Framework (Trapezoidal Stope-Shape Tubes)
The trapezoid transverse section framework can have:
 Regular width (equal) transverse-section spacing (U-axis) with the position of the levels (V-axis) defined by section control-strings (with gradient) within each
transverse section. The transverse section length (W-axis) is not optimised. Its length is defined by the lateral extent of the section control-string or the step-size
specified for the W-axis. The step-size could typically be the blast hole ring burden. The option code for this option if TFIS.
 Irregular width transverse-section spacing (U-axis) with the position of the levels (V-axis) defined by section control-strings (with gradient) within each transverse
section. The transverse section length (W-axis) is not optimised. Its length is defined by the lateral extent of the section control-string or the step-size specified for the
W-axis. The step-size could typically be the blast hole ring burden. The option code for this option is TIVS.
Figure 2-25 (TIFS, TIVS) Transverse Section – Irregular Trapezoidal Stope-Shape Tubes

2.7.5 Shape Types
Three different shape types are recognised:
(1) Full-shapes (often referred to as "full-stope"). This is the entire framework trapezoid / rectangle shape in the U and V axes. A full stope-shape is defined as one that covers
one full interval in the U-axis and V-axis (i.e. one section-spacing and one level spacing for the vertical orientation frameworks, or one strike-interval and one contour- interval
for horizontal orientation frameworks).
(2) Sub-shapes (often referred to as "sub-stope"), a regular or irregular sub-interval of the full-shape.
(3) Development-shapes, with dimensions of width and height (typically equivalent to the ore development profile).
In the stope optimisation run, full-shapes are processed first in a tube, and then sub-shapes (from the remaining ore not mined with a full-shape), and lastly development-shapes
(from the ore that cannot be taken in either full-shapes or sub-shapes).
Any combination of these shapes can be selected in an optimisation run, and some overlap in definition is possible. A sub-shape having the same dimension as the full-shape
could be specified, but there would not be any value in this if the full-shape had also been selected, as the full-stope would have already formed stopes of that dimension.
2.7.5.1 Sub-Stopes
Full-shapes (full-stopes) can be sub-divided in either and/or both of the shape’s U-axis and V-axis, to create sub-shapes (sub-stopes). Defining sub-stopes helps maximise resource
recovery into stopes, especially at the orebody strike extents or where the mineralisation may be patchy or irregular.
A maximum of 25 sub-shapes can be specified in any one run (i.e. 25 different sub-shape U|V dimensions).
Sub-stope geometries can be regular, irregular rectangular or irregular trapezoid to identify full and partial sub-stope shapes, and optimal combinations of sub-stopes.
Note that sub-stoping requires a model with cells defined in waste/pillar areas in order to honour stope geometry rules for the minimum pillar width between stopes and sub-stopes,
and between sub-stopes (in the W-axis). The “minimum pillar width” function requires model cells to be present in order to flag what has already been mined in earlier passes of
stopes and sub-stopes. This is necessary so that the minimum pillar width can be maintained between "mined" material and the new sub-stope.
Sub-shapes are defined by one of several techniques:
a. Divisor - specifying the number of sub-shapes along the U and/or V axes. The sub-shapes are strictly non-overlapping.
b. User Defined – The user can specify a list of sub-shapes using proportions as fractions. Later entries in the list can be overlapping and/or non-overlapping.
c. Automatic Function - this will automatically generate a list of over-lapping sub-shapes relative to a face position based on a selected divisor. There are three options being;
[edge, forward edge, and backward edge].
The following two examples show the parameters for two equal half stopes along strike for a sectional orientation where U is along the primary horizontal strike -axis and V is along
the vertical axis. The first example uses the (a) integer divisor approach and the second uses the (b) user defined fractional proportion approach.
Example 1
Divisors would be
unumber = 2 vnumber = 1
Example 2
Fractions would be (umin, umax, vmin, vmax)

(0.0, 0.5, 0.0, 1.0) and (0.5, 1.0, 0.0, 1.0)
More complex sub-stopes could be half of the section and half of the level spacing – potentially forming 4 sub-stopes. Figure 2-26 depicts examples of situations with 2 sub-stopes
in both the strike (U-axis) and elevation (V-axis) directions for (1) a regular U|V framework and also for (2) a regular U and irregular V framework.
Figure 2-26 Sub-Stopes for Regular and Irregular Vertical Frameworks Example
2.7.5.2 Sub-Stopes with Varying Strike Length
The sub-divisions to form sub-stopes may be regular divisors (e.g. 2 for halves, 3 for thirds, 4 for quarters, 5 for fifths, etc.) or they can be variable divisors based on fractional
proportion (e.g. {0.3 and 0.7}, {0.2 and 0.4 and 0.4}, etc.)
The following set of values would create 3 sub-stopes with a full-stope height of 20m but with varying strike lengths of 10m, 15m, and 25m based on a full-stope dimension of 50m
strike by 20m height. This is achieved by using fractional proportions of 0.2, 0.3 and 0.5 for the U-axis, as illustrated in Figure 2-27.
(0.0, 0.2, 0.0, 1.0)
(0.2, 0.5, 0.0, 1.0)
(0.5, 1.0, 0.0, 1.0)

Figure 2-27 Sub-Stope Fractional Proportion Example
The simplest sub-stope could be half-of either the section or level spacing (vertical orientation framework) – potentially forming 2 sub-stopes (either in the strike direction or in the
vertical direction). Both sub-stopes could be produced, especially if the orebody geometry is erratic but still carries acceptable grade within the full-stope interval. Alternatively,
either sub-stope could be produced, especially if the orebody extents or mineralisation extents do not cover the full-stope extents (e.g. mineralisation that “falls-short” of a level
interval or strike interval).
Variable divisors (using fractional proportion) can be used to replicate repeating regular patterns, such as for the Room-and-Pillar mining method layout (using a Horizontal
framework).
Up to 25 sub-stopes can be specified and will be processed in the sequence supplied. The sub-stope-shapes considered can overlap. If you wanted SSO to attempt the creation
of a stope-shape using 80% of the section spacing (starting from the lower-coordinate section), then attempt 60% if the 80% doesn’t succeed, then 40% if the 60% doesn’t succeed,
then 20% if the 40% doesn’t succeed (e.g. used for maximising mining at the lode strike extremities) with the (umin,umax,vmin,vmax) fraction values would be:
(0.0, 0.8, 0.0, 1.0)
(0.0, 0.6, 0.0, 1.0)
(0.0, 0.4, 0,.0 1.0)
(0.5, 0.2, 0,.0 1.0)
The SSO can be visualised as working in sequence from top to bottom in Figure 2-28 until a successful stope-shape was found.
Figure 2-28 Sub-Stope UV Fractions
2.7.5.3 Automatic Sub-Stope Edge Sequences
The 2.7.5.1 Sub-Stopes (c) “automatic function” option enables irregular shapes to be defined (e.g. the stope and drive shapes, or the primary and secondary stope-shapes) and
would be used without full shapes being specified. A number of automatic configurations of sub-shapes can be supplied where the interval dimension is given on the U or the V-
axis (but not both at the same time), and where sub-shapes abut the framework cell boundary. The largest sub-shape is chosen first. These automatic methods have the goal of
identifying how stope sub-shapes might be used to find a sub-shape contiguous with an adjacent full shape (or sub-shape).
The following automatic function examples (depicted in Figure 2-29) assume the U divisor number is 4 and the sub-shape dimensions are umin, umax, vmin, vmax.
1) Edge sequence (sub-stope-shape must abut with the adjacent shape)
(0.00, 0.75, 0.0, 1.0)
(0.25, 1.00, 0.0, 1.0)
(0.00, 0.50, 0.0, 1.0)
(0.50, 1.00, 0.0, 1.0)
(0.00, 0.25, 0.0, 1.0)
(0.75, 1.00, 0.0, 1.0)
2) Forward sequence (abuts with adjacent shape wall advancing from near side)
(0.00, 0.75, 0.0, 1.0)
(0.00, 0.50, 0.0, 1.0)
(0.00, 0.25, 0.0, 1.0)
3) Back sequence (abuts with adjacent shape wall retreating from far side)
(0.25, 1.00, 0.0, 1.0)
(0.50, 1.00, 0.0, 1.0)
(0.75, 1.00, 0.0, 1.0)
Figure 2-29 Automatic Sub-stope Edge Sequences

2.7.5.4 Overlapping Sub-Shapes
By default, sub-shapes are processed in the order supplied (either from a user supplied list or an automatically generated list), and ore will be mined in the first economic sub -shape
encountered in the list. In the case where there are multiple orebody lodes/lenses and potentially multiple stopes transversely across strike, if one lode/lens is successful with a full
stope, but not the others, the others will still be considered for sub-stopes.
With overlapping sub-shapes it is logical to supply the largest first, otherwise a smaller sub-shape may remove the opportunity to mine a larger shape in the same position.
The sequential order of evaluation does not guarantee the best combination of sub-shapes is found.
Note that the Stope Shape Optimiser does not check the combination of full-stopes and sub-stopes simultaneously for the set that optimises the objective (grade / value / calculated
value). E.g. It could be the case that two solutions exist, one with a full-shape and another with sub-shape(s). Even though the sub-shape solution may have a higher
grade/value/calculated value, the full-shape solution will always be generated as the evaluation sequence terminates with the earliest solution found.
2.7.5.5 Sub-Stope Optimisation
Sub-stope Optimisation evaluates all the sub-stopes, and finds the best combination of non-overlapping sub-stopes. The processing time overhead should not be any higher than
evaluating a list of sub-stopes.
If the option to optimise sub-stopes is selected, then an optimal choice of sub-shapes is made to maximise the value (or metal). This option can only have an effect when the
supplied list of sub-shapes is overlapping. In the earlier Edge sequence example (Figure 2-29) the optimal combination might be one that leaves intermediate waste in a different
position to that found by exploring the sub-shapes in the sequence given e.g. with a result (0.00, 0.25, 0.0 1.0) and (0.50, 1.00, 0.0, 1.0) rather than (0.00, 0.50, 0.0, 1.0) and (0.75,
1.00, 0.0, 1.0) to better isolate the higher grade ore.
If sub-stopes are optimised, a set of non-overlapping sub-stopes will be selected, and these are the only shapes that will be used on each lens. Without sub-stope optimisation each
shape in the list is run in turn, and any lens that can form stopes from that shape will do so at the first opportunity. In the sub-stope optimisation case, SSO is looking for the best
overall solution with a subset of the shapes, whereas in the non-optimisation case it is like a greedy algorithm that will apply a sub-stope at the first available opportunity, and
consequently the sub-stopes from one lens might overlap the sub-stopes from another adjacent lens (which cannot occur by definition in the sub-stope optimise case).
If a full stope-shape is included in the sub-shape list then optimisation can assess whether a sub-stope captures more value than a full-stope (that is otherwise forced to mine more
included waste).
If you evaluate a list of overlapping sub-stopes on parallel lenses (without sub-stope optimisation) you might get overlapping stope-shapes from one lens to the next and also
potentially get a higher value. Note also that if there are multiple lenses and multiple stopes transversely across strike, the lenses are not checked individually for the choice of full
stope and sub-stope combinations – the full stope-shape is applied to all lenses, and where a full stope does not create an economic stope, a sub-stope will be considered.
2.7.5.6 Pillar Locations
Irregular U or V axes “pillars” can be defined by specifying the position co-ordinate and size. Typical pillar examples are:
 Defining primary and secondary (pillar) stopes,

 Defining rib pillars between stopes along strike
 Defining sill pillars between mining blocks on dip.
 Analysing frameworks to define the optimum U, V dimensions combination e.g. assessing the best combination of level interval combined with section interval.
An example of alternating 10m and 20m definitions is provided in Figure 2-30:
Figure 2-30 Irregular Rib Pillar or Stope Example
2.8 Seed-Slice Interval Selection

Although the seed-slice interval definition is not specifically a stope-shape geometry parameter, its selection has implications regarding the accuracy of the stope-shapes and
processing speed.

The slice interval ideally should be an integer divisor of:
 the minimum mining width,

 half the pillar width between transverse stopes, and
 the dilution widths for near/far or hangingwall/footwall surfaces.
The seed-slice interval should typically be set to get a minimum of 3 to 5 intervals in the minimum stope width.
If the slice interval is not an integer divisor then inaccuracies may occur at the seed stage because the seed-shape and pillar shape can only be multiples of the seed-slice
intervals. If ore lenses are widely spaced, the choice of slice interval will be less critical. However if the ore lenses are closely spaced and the optimal pillar width is close to the
minimum pillar width, then the choice of slice interval will be more critical. If there is a single lens then the choice is less critical.
The number of seed-slice intervals across the orebody can be up to a maximum of 4096 in Version 2. The seed-slice generation process gets proportionately slower as the number
of slice intervals increases. Hence careful selection of seed-slice interval, minimum stope width, minimum pillar width between stopes and dilution skin intervals is required to not
exceed this limit and/or to keep the processing time reasonable. The maximum number of seed-slice intervals allowed is not constrained by the W dimension of the framework, but
it is constrained by the maximum distance along the tube (W-axis) that would contain the slices.
As an example, a 1m slice interval would be appropriate for:
 5m minimum mining width,

 6. minimum pillar width,
 2m hangingwall dilution,
 1m footwall dilution
In Figure 2-31, the results of using different seed-slice intervals are illustrated where the orebody has 2 lenses and the lenses are either both high grade, or a combination of high
and marginal grade. In case (A) a1.0m interval is modelled which is optimal for the stope and pillar geometry. In case (B) a 1.5m slice is used and while a larger seed-shape must
be modelled so that multiples of the interval satisfy the stope and pillar geometry, two stopes are still generated. In case (C) the lower grade lens does not meet cut-off for the
larger seed-shape and only one seed-shape is (incorrectly) found.. A slice interval of 1.5 is not an integer divisor, and hence the seed-shapes chosen differ. If both seed-shapes
remain economic then nothing is lost because the seed-shapes will be refined in the annealing process, and both slice intervals should return the same final stope-shapes. If the
seed generation returned only one seed-shape (because the other was found to be sub-economic) then a seed-slice interval of 1.5 would have been a poor choice.
Figure 2-31 Influence of Slice Interval on Stope Seed-Shape and Pillar Geometry
As a second example, a 0.75m slice interval would be appropriate for:
 7.5m minimum mining width,

 9m pillar width,
 1.5m hangingwall dilution,
 0.75m footwall dilution
In the second example, if the lenses above cut-off were always greater than 11.25m (i.e. 9+1.5+0.75) apart, or the seed-shapes were further apart than 11.25m, then a choice of
1.0 would also suffice. If the dilutions were small, say 0.25m each and the seeds were greater than 9.5m (i.e. 9+0.25+0.25) then using 0.25m as the seed-slice interval would be
the most accurate, but would add significant runtime overhead to seed generation.
If in doubt, then running a test area for a range of slice intervals would be sensible.
2.8.1 Use of Control Surface for Lenses with Different Orientations
If the ore lenses have different orientations across the orebody, then no single default dip and strike would be satisfactory for generating slices for seed generation. This is
illustrated in Figure 2-32. A vertical slice is a compromise orientation but for the right-hand lens, sampling the orebody with vertical slices yields sub-economic slices because of the
waste carried, so no seed-shape is generated, and consequently no stope-shape is generated. If a Stope Control Surface that is roughly oriented to the expected stope-shape is
added (as indicated by the red lines on the section), then the slices carry little waste. The slices will be economic and a seed-shape will be generated for both lenses. Note that the

seed-shape includes the hangingwall and footwall dilution parameters.
Figure 2-32 Impact of Control Surface
2.9 Shape Parameterisations

The following sub-sections describe the key parameters to control the Slice Method stope-shape. Figure 2-33 illustrates the a number of the key geometry issues.
Figure 2-33 Key Geometry Parameters (Sectional Orientations)
2.9.1 Minimum Stope Width
The minimum mining width parameter is defined as distance in the horizontal plane on the framework section along the W-axis (and consequently measures the apparent width).
If the orebody dip is moderate or the strike deviates from the framework axis, then it would be appropriate to make a correction to the width specified to better approximate the
intended true width. As an example, if the minimum stope width in the true-width dip-direction was intended to be 10m and the orebody was dipping at 45o, then setting the
minimum stope width to 14.1m (horizontal distance) would approximate the intended minimum stope width. See Figure 2-34. Note that the true width is a function of both strike
and dip orientation in three dimensions for the general case.
Figure 2-34 Minimum Mining Width

If the stope wall angle ranges are the same for both the hangingwall and footwall, or roof and floor, then the minimum stope width is checked at the stope corners.
If the stope wall angle ranges are different, then the minimum stope width is checked at the wall centre, because the optimal seed-shape is measured at the wall centre, and the
annealing shape must be measured in the same manner to ensure that a feasible annealing shape is available at the start of annealing.
2.9.2 Maximum Stope Width
The maximum mining width parameter is defined as distance in the horizontal plane (in a similar manner to Figure 2-34) on the framework section along the W-axis (and
consequently measures the apparent width).
An example use for the maximum stope width is to restrict the transverse dimension for geotechnical purposes (e.g. not to exceed the stable hydraulic radius for the crown face or
the strike-face walls).
There is also the option in post-processing to split the stope width into smaller intervals without pillars (see Post-Processing Options Split 2.12.2). The maximum stope width should
be interpreted as maximum stope width between pillars. The post-processing approach is preferred over the now discouraged approach of specifying a small (non-zero) pillar
width, and a maximum stope width equal to the interval sought, as was used with Version 1 by some users.
2.9.3 Minimum Pillar Width
A pillar will separate seed-shapes or stope-shapes if the maximum stope width would otherwise be exceeded, or low grade/waste can be isolated from stope shapes.
Waste cells (representing mineralisation below cut-off, or rock without mineralisation) surrounding the ore cells are required for runs with sub-stopes, as the location of the mined-
out cells is used to force the pillar width between stopes and sub-stopes, and between sub-stopes and sub-stopes.
If the stope wall angle ranges are the same for both the hangingwall and footwall, or roof and floor, then the minimum pillar width is checked at each corner. If the stope wall angle
ranges are different, then the minimum pillar width is checked at the wall centre.
Note that the pillar width parameter is defined as the distance in the horizontal plane i.e. the apparent pillar width (in a similar manner to Figure 2-34).
2.9.4 Strike Angle
The strike angle is the angle of any one of the four stope wall edges (measured relative to the U axis of the stope orientation plane).
Ideally the strike angles would be loosely defined (i.e. using broad tolerance range) in a preliminary test SSO run in order to give a reasonable upper limit on the number of stopes
produced or to maximise the stope dimensions. The strike angle parameters would then be progressively refined as required.
One example application would be where stope-shapes are formed in a criss-crossing pattern between “parallel” lenses which have discontinuous mineralisation, and the user
wanted to force the stopes to not criss-cross between the lenses.
Another example would be the formation of stope-shapes that have rapid or chaotic changes in wall angles, giving the appearance of being “malformed” (but are actually not).
The above examples may be considered to be impractical stope-shapes to implement, and hence the wall strike angle changes are “smoothed” out to better approximate a
mineable set of stope shapes.
The following sub-sections describe the strike angle parameters in more detail.
2.9.4.1 Strike Angle Range
This defines the strike angle range of either edge (i.e. top or bottom) of either wall of the stope-shape (i.e. near/far wall or hangingwall/footwall wall) relative to the framework’s strike
axis (i.e. the U-axis). The range can be independently defined as positive and/or negative relative to the stope shape framework strike axis.
2.9.4.2 Strike Angle Maximum Change
This defines the maximum allowable stope-shape “twist” relative to the top and bottom wall edges.
2.9.5 Side Length Ratio
The ratio is defined by the end-face wall lengths and the axis direction pairing being considered, described further in the following sub-sections.
Ideally the side length ratios would be loosely defined (broad range) on a preliminary SSO run to maximise the number of stopes produced or to maximise the stope dimensions.
The side length ratios would then be progressively refined as required.
An example use of the side length ratio is to force walls (i.e. near/far walls or hangingwall/footwall walls) to be parallel to each other (i.e. a sectional parallelogram) so that all
production hole drilling is parallel for a narrow tabular orebody. This is achieved by using a 1:1 ratio, but this ratio should only be used in a final run to ensure that all the required
shapes are generated in the annealing phase. Likewise, in the U-axis direction plan view parallelograms can also be specified.
2.9.5.1 Side Length - vertical_side_length_ratio, top_bottom_maximum
For either top-edge divided by bottom-edge or bottom-edge divided by top-edge) and applies to either end-wall face (i.e. front or back). The upper limit for the ratio (longer/shorter)
of the top and bottom edges of the front and back strike-face of a stope-shape as defined in Figure 2-35.
Figure 2-35 Top-Edge to Bottom-Edge Ratio (V-axis pairing)

2.9.5.2 Side Length - vertical_side_length_ratio, front_back_maximum
The upper limit for the ratio (longer/shorter) of the front and back edges of the top and bottom face of a stope-shape as defined in Figure 2-36.
Figure 2-36 Front Edge to Back Edge Ratio (U-axis pairing)
2.9.6 Waste Inclusion Control
The maximum waste fraction of stope-shapes can be defined (i.e. proportion of rock with mineralisation values below specified cut-off included within the stope-shape).
The waste inclusion is defined as
A default fraction value of 1.0 means any waste proportion is acceptable.
It is good practice to:
 Start with a test value of 1.0 and gradually refine this in subsequent runs to monitor the impact.
 If you have back-fill in the “voids” located within stope-shapes, then do not associate this material with the “report_exclusion_field”, as it should have a density and grade,
and effectively be in one of the three waste categories (as described in Section 4.5.3 Waste Reporting).
2.9.7 Exclusion Control
The exclusion control can be used to avoid the creation of stope-shapes within say deleterious-processing material or in poor rock-mass zones as a few examples of its use. The
material in the model is flagged with a (numeric or alphanumeric) field and a value.
Up to two exclusion fields can be defined in a run. Each exclusion field permits a certain maximum tolerance of inclusion within a stope-shape.
The exclusion control is defined as the maximum fractional proportion of material (by volume) in the stope-shape that is flagged as either “exclusion1” or “exclusion2” material.
The tolerance for each exclusion field “exclusion1” and “exclusion2” is set independently, and allows for a maximum fractional proportion of the respective exclusion material to be
incorporated into the stope-shape. As an example, you may be mining a secondary pillar stope and the shape of the adjacent primary backfilled stopes may bulge into the
secondary, so for practical purposes you may allow say up to 5% (0.05 fraction) of backfill material within the secondary stope-shape.
2.10 Stope-Shape Framework Angles

The following discussion regarding angles relates only to the Slice Method frameworks. For rotated stope-shape frameworks the strike and dip parameters are measured in the
local co-ordinate system.
2.10.1 Angle Conventions
All angle values are specified in degrees. Angles are defined relative to the selected framework orientation (i.e. XZ|YZ|XY|YX). Table 2.6summarises the permitted values.
The Strike angle conventions are the same for the various Slice frameworks (i.e. XZ|YZ|XY|YX). Strike is measured positive clockwise from the primary strike-axis (U-axis positive
direction) of the selected stope-framework orientation/plane (i.e. 0 degrees = looking along the strike-axis in the positive co-ordinate direction, +90 degrees = looking clockwise at
right-angles from the positive strike-axis plane, -90 degrees = looking anti-clockwise at right-angles from the positive strike-axis plane). The strike angle range is [-90 to + 90]
degrees.
The Dip angle for XZ|YZ frameworks is measured as 0 degrees from the left-hand-side horizontal axis as you look along the primary strike-axis (U-axis) and increases
anticlockwise to +90 degrees vertically down and +180 for the right-hand-side horizontal axis. The dip angle range is [0 to 180] degrees.
The dip angle for XY|YX frameworks is measured positive downwards from the horizontal (and negative upwards) on both the primary axes (the first axis in XY|YX orientation i.e.
U-axis) and the secondary axes (the second axis in XY|YX orientation i.e. V-axis) and are termed the ‘strike dip’ angle, and the ‘transverse dip’ angle respectively. The dip angle
range is [-90 (upwards) to +90 (downward)] degrees.
Table 2.6 Strike and Dip Angle Conventions
All Angles in Degrees

XZ, YZ XY, YX
Strike [-90 ..0.. +90] [-90 ..0..+90]
Dip [0..+180] [-90..0..+90]
Strike and dip convention examples are illustrated in Figure 2-37 to Figure 2-46 for each framework orientation.

Figure 2-37 Vertical XZ Strike Convention (Plan View)
Figure 2-38 Vertical YZ Strike Convention (Plan View)
Figure 2-39 Rotated Vertical XZ Strike Convention
Figure 2-40 Vertical XZ Dip Convention

Figure 2-41 Vertical YZ Dip Convention
Figure 2-42 Transverse Section XZ Dip Convention
Figure 2-43 Transverse Section YZ Dip Convention
Figure 2-44 Horizontal YX Dip Conventions

Figure 2-45 Horizontal XY Dip Conventions
Figure 2-46 Vertical Rotated XZ Dip Convention
2.10.2 Stope Wall Angles (Equal or Different)
Stope walls are referred to as the near or far, and hangingwall or footwall sides for XZ|YZ oriented stopes and roof or floor sides for XY|YX oriented stopes.
There are two wall-angle range cases as follows:
 Equal - using the same angle range for both walls (e.g. 45-90o).
 Different - using independent angle ranges for each wall. For example, this allows setting say a minimum rill dip for the footwall side (e.g. 45-90o) and allowing a flatter
minimum dip for the hangingwall side (e.g. 30-90o) – a typical requirement for flat-dipping orebodies.
Note that the default dip value for seed-slices cannot fall outside of the near wall and far wall (hangingwall and footwall) range. This requirement also applies to the default strike
angle and the strike angle range.
2.10.3 Geological and Geometric Stope Wall Definitions

2.10.3.1 Geometric (previously called Local)
It is possible for the hangingwall and footwall to tip over and back again over the vertical extent of an orebody. This behaviour is typical of “banana-shaped” lodes (viewed as a
vertical transverse section). The geometric-dip orientation defines whether it is the hangingwall or the footwall.
The stope wall-angles, dilution intervals and structure orientation are variable using the geometric hangingwall and geometric footwall dip orientation definitions in the geometric
case.
2.10.3.2 Geological (previously called Global)
In the Geological case, the overall hangingwall and footwall sides are fixed over the vertical extent of the orebody. The stope wall angles, dilution interval and structure orientation

are constant “geological hangingwall” side and constant “geological footwall” side using Geological case definitions.
The Geological definitions are near for the lower co-ordinate and far for the higher co-ordinate, along the W-axis relative to the stope shape framework origin in the local coordinate
system.
The geological and geometric stope-shape wall definitions for vertical frameworks are depicted in Figure 2-47.
Figure 2-47 Geological and Geometric Wall Definitions – Vertical Frameworks
The geological (global) and geometric (local) stope-shape wall definitions for horizontal frameworks are depicted in Figure 2-48.
Figure 2-48 Geological and Geometric Wall Definitions – Horizontal Frameworks
2.11 Additional Control Surfaces

Control surfaces are wireframe surfaces that are used to guide the orientation and shape of stopes.
2.11.1 Structure Surface
The structure function is applicable to both full-stopes and sub-stopes.
The structure function results in the stope-shape either snapping-to the structure surface (e.g. include “waste” that would “normally” fall into the stope-void due to the presence of
the structure) or standing-off from the structure surface (e.g. leaving a skin of “ore” against the structure for dilution control i.e. ore loss). The Stope Shape Optimiser assesses
both options in generating the seed-shape, and applies the same rules to the annealed shape. If the snapping-to shape is sub-economic, then the Stope Shape Optimiser will still
consider the “standing-off” option. This will depend upon the relative position of mineralised material and the structure surface.
Where a stope-shape wall corner falls within a nominated minimum distance of the structure wireframe, the stope projects the stope-shape to that structure surface. If one or more
corners are within the minimum distance, the remaining corners are tested against the maximum distance.
The stope wall is snapped-to the structure position if it falls within the set criteria (minimum - target, maximum - range). This can result in a dip angle that is flatter or steeper than
set in the stope geometry parameters.
Where snapping to the structure wireframe causes angle tests to fail, these stopes would not be output unless the “output_subeconomic_stopes” parameter is set. Stopes that do
not meet the stope geometry settings are reported as a failed stope (i.e. with RESULT=0).
Figure 2-49depicts the operation of the structure surface function.
Figure 2-49 Structure Function Example

2.12 Post-Processing Options

The following post-processing options are only applicable to the SLICE METHOD(section 2).
2.12.1 Smooth
The stope-shapes from the SSO are optimised on a tube-by-tube basis independently, and consequently, the abutting stope walls will typically not match exactly in position. This
may be ideal for abutting stopes that can be mined independently of each other (e.g. primary and secondary long-hole stopes) but commonly this does not occur for geotechnical
reasons and/or for the mining method practicalities. A typical example would be a continuous retreat long-hole benching mining method.
Smoothing is an additional step using the same annealing algorithms. This optimises the shapes not just in a single tube, but taking into account the adjacent tubes. The “gaps”
between the corners of adjacent stope-shapes will be eliminated / minimised from one tube to the next in order to:
 Create a smooth transition vertically (V-axis) and/or horizontally (U-axis) for vertical framework orientations, or
 Create a smooth transition in the U and/or V axes for the roof/floor for horizontal framework orientations.
The additional smoothing is applied subject to:
 A maximum allowable gap tolerance between corners of abutting stopes,

 The smoothed stope-shape not falling below the designated cut-off
 The stope geometry rules not being violated.
Smoothing does not “average”. It first tries to minimise the gap while maintaining each of the user supplied stope geometry constraints. For example, if your stope-wall dip range
was tight, then it may not necessarily “close the gap” if the dip limit was violated. It cannot make an economic stope sub-economic. It then tries to improve the value of the
resultant stope-shapes without significantly increasing the gaps. So in some runs, only portions of the stope wireframes may be smoothed. Any gaps between stopes that are
greater than the maximum allowable tolerance will not be changed. Gaps below the maximum may be partially or completely reduced to zero gaps.
Sub-stope corners are adjusted at the corners adjacent to a full-stope corner, however sub-stopes are not smoothed with other sub-stopes.
Prior to smoothing the gaps are analysed and stopes with more gaps to adjacent stopes and sub-stopes are processed first. i.e. stopes at the centre of the orebody are likely to be
smoothed first and stopes at the extremities will be smoothed last. All tubes are processed in sequence in the first pass. As each tube is smoothed, up to 8 other adjacent tubes
(with stopes or sub-stopes) will be adjusted. The complexity of the annealing in this additional step can be an order of magnitude greater, and consequently the smoothing run can
often take 5-10 times longer than an unsmoothed run, and sometimes more, particularly if there are multiple lenses or parallel orebodies.
The first pass is usually enough to give a good result. There are no parameters to control the time spent on the first pass. Any additional passes are optional, and under user
control. The parameters to control additional runs are either by number of passes or by time limit (in hours).
Smoothing can be controlled by either comparing the "gaps" between stope corners, or alternatively the "ratio" of stope edges. The first method has been found to be the most
effective.
An example of the smoothing function is illustrated in Figure 2-50. This is plan view where the original stope-shapes (white) are smoothed (orange).
Figure 2-50 Smoothing Function Example

2.12.2 Split
This function subdivides stope-shapes according to user defined rules. It is generally applicable for stope-shapes that are wider than their maximum stable wall span(s) or for sub-
setting very-wide ore bodies. It may also have some conceptual design application for drift and fill and mechanised cut and fill type mining methods by setting split widths to
development width and using a minimum and maximum tolerance on the split width. It may also be applicable for establishing shapes that correspond with blast ring increments
such as for the transverse SLC mining method.
Various rules are provided for splitting stope-shapes:
 Split in transverse direction (W-axis).

 Split in longitudinal direction (U-axis and/or V-axis). Note that longitudinal splitting is equivalent to applying U-axis (and/or V-axis) sub-stopes from a “full” stope.
 Split on a regular stope-framework grid (with optional annealing) to create a checkerboard pattern (for abutting open stoping layouts).
 Offset the split wall positions between adjoining stopes (i.e. adjacent in the U-axis sense) by using an offset from the stope-framework grid for a staggered checkerboard
pattern.
 Split from a centred stope. It places the first stope central to the transverse width i.e. not on either side of the centre. The "centre" is defined as being at the mid-stope
height).
 Split equally.
 Split from the near wall with the final split adjusted to tolerance settings.
 Split from far wall with the final split adjusted to tolerance settings.
 Split from hangingwall with the final split adjusted to tolerance settings.
 Split from footwall with the final split adjusted to tolerance settings.
 Vertical or angled internal split walls. The internal walls are either split vertically or proportionally parallel to the near/far or hangingwall/footwall walls.
 Vertical or angles end walls (at the W-axis extremities).
Some notes on the parameters used follow:
 The dimensions of split stopes is determined by an "interval", a "minimum" and a "maximum" length. In most cases the un-split stope length will not be an exact multiple of
the "interval", and so the "minimum" and "maximum" values will need to be selected to take into account all possible split sizes. Two cases will need to be considered
depending on whether the "interval" value is also the "maximum" value, or the "maximum" can exceed the "interval". As a guideline the “minimum” should be less than half
the “maximum”.
 The “transverse_offset” only functions for grid splitting.
 In the detailed output report, the average width field “SAVGWID” is calculated as an apparent width using the stope volume divided by the UV face projection area (i.e. the
projection area of the quadrilateral/rectangle tube representing the hangingwall/footwall faces). If the face walls dip, there will be a slight difference from the geometric or true
width.
 Splitting is applicable to sub-stopes.
 The increments, minimum and maximum distances are referenced from the stope centroid position (i.e. not from the floor, roof/back/crown, or either sidewall).
 To identify each split stope in the stope naming function you can concatenate the output file fields QUAD, PASSNUM and STOPENUM to create
QUAD_PASSNUM_STOPENUM (i.e. individual split stopes in a primary pillar-line or secondary pillar-line). The QUAD will be the same for both primary and secondary as it
is a subdivision of a “full” stope, the PASSNUM identifies which sub-stope it is, either primary=2 or secondary=3 for example, and the STOPENUM provides the split order
(i.e. transverse counter) from lower co-ordinate to higher co-ordinate (in this case footwall/near side to hangingwall/far side).
Examples of various split cases are depicted in Figure 2-51 and Figure 2-52. Figure 2-51 shows plan view examples for an orthogonal framework with regular section spacing.
Figure 2-52 shows “split_equal” result in plan and section view for a rotated framework with irregular section spacing.
Figure 2-51 Split from HW, Split from Centre and Split Equal

Figure 2-52 Split Equal Applied to Rotated Framework with Irregular Section Spacing using Sub-Stope Function
2.12.3 Merge
The “merge” function is only applicable for full-stope runs in the current implementation of SSO Version 2.
Merge may be used for the following reasons:
 To define a maximum that is determined by geotechnical stability (e.g. not to exceed a hydraulic radius criteria).
 To define a minimum that is determined by economics or mining practicalities (e.g. combining stope-shapes that required small intervals due to variability of the orebody)
 To define a regularised extraction sequence for stope-shapes (e.g. vertical stacking or primary and secondary stopes)
Details of the two available merge options follow. Note that after applying a merge option, the QUAD number remains the same as prior to the merge. The stope name and stope
number are changed to reflect the agglomerated stopes.
2.12.3.1 Merge on Grid
A grid interval, being a multiple of the full stope dimension in the U-axis direction, is selected where the framework U origin provides the reference “U” value.
With the “grid” interval option, the interval is the grid spacing with stopes grouped to be aligned on a grid. This is useful if merged (grouped) stopes must be aligned from level-to-
level. The “grid” option requires stopes to be bounded by grid lines (unless the remaining pieces can be attached within the stope strike-length min/max range).
The maximum length for the grid option must be less than twice the grid interval.
2.12.3.2 Merge on Interval
For the “interval” option, the interval is not constrained to be on a grid. i.e. shapes will be grouped level by level but not conforming to a grid spacing, just to achieve the target
interval. The “interval” option aims to group the stopes at the interval size (regardless of the grid position), but otherwise still work within the min/max stope strike-length limits.
With regards to minimum and maximum merge lengths, a whole integer multiple makes good sense as no other outputs can be generated. For example if you have 5m stopes and
you want to group into 15m stopes, your only options are 5m, 10m, 15m for minimum, and 15m, 20m, 25m for maximum. If you want the most stopes output you would have
minimum of 5 and maximum of 25 to pick up all the pieces that do not form full 15m aggregations.
Figure 2-53 depicts 10m strike-length full stopes, while Figure 2-54 and Figure 2-55provide examples of the grid and interval approaches, respectively.
Figure 2-53 Full stopes (10m strike) prior to applying any Merge function
Figure 2-54 Updated stopes after applying 50m Grid Merge (using 30-90 Range)

Figure 2-55 Updated stopes after applying 50m Interval Merge (using 30-70 Range)
2.13 Advanced Optimisation Controls

Advanced options are provided to handle some special cases.
2.13.1 Do Not Force Full-stope Sub-stope Pillars
Use this in situations where you would want to mine sub-stopes that are adjacent to full stopes in the W-axis (where the normal practice is to require a pillar). An example of this
function is depicted in Figure 2-56.
Figure 2-56 “Do Not Force a Pillar between the Full-stope and Sub-stope” Example
2.13.2 Do Not Force Stope Split Sub-stope Pillars
In the Split post-process operation, there are no pillars between split stopes, but at the end of a split sequence the pillar is still required. This switch removes the latter requirement,
much like the previous switch.
2.13.3 Stope Sequence Subset Optimisation
The complexity and hence the runtime for annealing can increase dramatically when many lenses are optimised concurrently in the transverse direction. This option allows the
annealing to be decomposed into a number of smaller annealing steps if a separation distance can be specified. Seed-shapes that are further apart than this separation distance
are annealed independently from others groups of seeds. It is important to note that the pillar width separation test is only enforced within each subset of seed-shapes, so the
separation distance should exceed the minimum pillar width.
2.13.4 Disable Structure Test on Seed
If expanding the seed-shape to the structure surface makes the seed-shape sub-economic, you don't otherwise know where a potential stope might have been (unless you do a full
run without structure post-processing), so this restricts the expand-to-structure test to only the anneal phase.
2.13.5 Other Optimisation Controls
The use of these advanced selections may speed up the SSO processing but they disable optimisation features that enhance the robustness of the optimisation results.
The other optimisation controls are:
 No seed cut-off reduction,

 No seed cut-off increase,
 No control surface averaging,
 No seed corner optimisation,
 No corner volume optimisation.
Seed cut-off reduction and increase are techniques to flex the cut-off grade when no seed-shapes are produced to cope with poor choice of parameters eg slice interval for narrow
orebodies. Control surface averaging is used to reduce the rapid change in dip and strike with complex surfaces. Seed and corner volume optimisation are techniques to improve
results in generating seed shapes for highly variable orebodies.
Using these options can speed up processing but they can degrade the quality of the results, and their use is not recommended without specific instructions from AMS.
2.14 Good Practice - Tips and Procedures

Every orebody is different, and optimising stope-shapes to those orebodies to produce realistic shapes is a complex geometric optimisation problem.
The Stope Shape Optimiser provides a tool to produce optimal stope-shapes, with careful selection of parameters, in a rapid and repeatable fashion. The procedure is not fully

automatic and sensible selection of parameters and controls will assist to improve the quality of results for complex situations.
A quick summary is provided here and then more detailed guidelines and examples are provided on specific topics:
(1) Block Model
Check the model to ensure that missing or absent data is identified and appropriate defaults are set. A default of zero for a value field is probably not appropriate because the
missing blocks or values are probably intended to be treated as waste (e.g. have negative values due to processing cost). Flag areas in the model where stopes should not be
optimised e.g. near the surface, close to infrastructure or in poor geotechnical ground. Use model fields like oretype or resource category to allow later classification of stopes.
Processing a model with waste cells adds a significant overhead, but this is necessary if creating full-stopes and sub-stopes. Remove barren waste cells from the input model if
only full-stopes are required.
(2) Framework
For a first run use a restricted framework size on a small test area to verify parameter selections are appropriate, and that the results generated are those expected. Alternatively
use the full framework but select an (x,y,z) coordinate as a validation test cell so that only one quad is optimised, or perhaps select several (x,y,z) coordinates to test several
areas. Using a cut down model will also speed up this initial testing.
Always do a visual check when using rotated frameworks. In particular verify that the framework extent and orientation are the expected ones and that the framework extends
beyond all the mineralisation.
(3) Manual Designs
Compare results with manual designs, or have other methods for checking the outputs, especially when there is a lack of familiarity with the data. It is too easy to do optimisation
work on computers. The modern day challenge is to have techniques to "prove" that the optimised result is correct.
(4) Refining and Repeating Optimisation Runs
There are many new features for adding controls and fine-tuning the stope optimisation process. The first run should have the simplest setup. These should then be followed with
more detailed optimisation runs, where functions are added incrementally to understand the implications of each change (e.g. maximum waste inclusion).
For example prior to doing a design with gradients on the levels, do a run with fixed sublevels.
Don't do post-processing (smooth, split and merge) on the first run. Smoothing typically improves the look of the result, but doesn’t significantly affect the overall tonnages that are
optimised.
(5) Stope Control Surface and Slice Interval
The Stope Shape Optimiser will only produce stope-shapes if seed-shapes are created. Seed-shapes define the number and approximate location of stope-shapes. A Stope
Control Surface is used to locally define the strike and dip orientation of the stope-shapes for the mineralised economic component of the orebody.
Seed-shapes are formed by aggregating seed-slices to model the stope width, pillar width and dilution. The choice of slice interval will affect the accuracy of the seed-shape
optimisation.
While a geological wireframe can be a good proxy for the Stope Control Surface, in some cases these wireframes can be large, or have local inflections that can confuse the Stope
Shape Optimiser when generating slices. The geological wireframe may be well understood by the geologist but should not automatically be adopted by the mining engineer. It is
better to digitise a set of strings section by section to define the local dip, and create one or more wireframe surfaces. Make sure that these surface(s) extend past the mineralised
zone. One technique for reducing the number of triangles is to use a “decimate” wireframe function in the Vendor environment.
A Stope Control Surface is mandatory for all but the simplest orebodies.
The Stope Control Surface need only have a few thousand triangles, and certainly not 20-100,000 triangles.
A stope slice interval where the minimum stope width is 3-5 times the slice interval is usually a good guide.
The dip and strike orientation taken from the Stope Control Surface, or the Default Dip and Default Strike parameters, must fall within the minimum and maximum dip and strike
ranges set for the stope geometry.
(6) Model Discretisation Plane and Discretisation
In almost all cases the model discretisation plane and the stope orientation plane for the framework will be identical, but this is not the case if the rotations for model and stope-
shape framework are different. Beware the case where one is rotated by 90 degrees relative to the other so that the axes look to be parallel but the XZ plane of one is the YZ plane
of the other.
A good rule for model discretisation number (in U and V) is that the number is twice the number of sub-stope intervals. The goal is to ensure that there are a minimum of two
discretised cell centres in U and V for each stope or sub-stope shape. The default of 4x4 is suitable for regular sub-stope splits of 2x2. If the model discretisation number is too
small, a more suitable choice will be automatically assigned and noted in the log file.
(7) Stope Geometry
In the first pass, nominate broad tolerance ranges for dip and strike of stope walls, and the geometric ratios of end wall or roof/floor dimensions. If the final strike limits are +/-10
degrees, then still consider a first run with a wider tolerance, even +/-45 deg. This will typically generate the maximum tonnage and then these values can be tightened up if this is
appropriate for the stoping method. Beware of setting tight dip and strike ranges if there is a lot of local variability in the orebody dip and strike.
In the second and subsequent runs tighten the tolerance ranges and parameters. The tighter the stope geometry constraints, the more complex and slower will be the
optimisation. If the side length ratios are set to one to generate parallel sides this can be a very severe constraint for the optimisation. Do several runs with diminishing side length
ratios to check if stopes are lost, and establish the threshold where stope-shapes begin to disappear.
(8) Different Evaluation Techniques
Because the Stope Shape Optimiser works off the Block Model, the sub-cell modelling and its approximation to wireframe surfaces will be a key determinant in the quality of the
output stope-shapes and also the accuracy of the resultant tonnes and grade. Different mining software packages have different capabilities and approximations in sub-cell
modelling. A full cell model with percentages for the mineralised fraction cannot be used with the Stope Shape Optimiser - overlay a sub-cell model to model these fractions and
derive the spatial position.
Engineers often align stope sections to the axis of the orebody, but geologists do not have the same expectation. This will mean that the stope-shape framework will often be
rotated to the block model. For a first pass an un-rotated stope-shape framework should yield a similar result to a rotated one, with small discrepancies at the end of ore zones.
The Stope Shape Optimiser has three evaluation modes (“Fast”, “Precise” and “Fast with Precise reporting”). The “Fast” and “Precise” methods will generally yield different results.
The “Fast” method is typically quicker, and it is often more accurate because it tends to adapt well to the sub-cell approximations in the block model. While the tonnages will be
very similar, the metal content can vary between the two methods by 5-10%. Significant differences between "Fast" and "Precise" methods will be reported in the runtime
log file for grade, tonnes and metal. The differences are more exaggerated in narrow orebodies.
(9) Stope Not Generated Where Expected
First check the log file (after running with the verbose output option) to look for reasons as to why the stope failed or was rejected. Missing stopes might be more likely for
narrower orebodies.
Some simple checks to consider:
 Does the Stope-shape Framework extend to the location of mineralization? Do a visual check for rotated models and/or rotated frameworks in particular.
 Review the wireframe verification file to examine the slices, slice orientation, and the seed-shape (noting that only slices above cur-off are output). Perhaps the slice
orientation is not correct for the orebody and more detail is required in the Stope Control Surface. Perhaps the slice interval is too corase. The message output to the log file
for failed stopes is also output as an attribute in the verification wireframe file for those stopes _ the "output subeconomic stopes" option must be selected and a colour needs
to be defined for failed stopes.
 Narrow orebodies with inferior sub-ceiling will prove difficult to optimise because too much waste will be carried in the stope-shape. The problem is the geological model, not
the Stope Shape Optimiser.
 If using Irregular cases of the rotated stope shape framework, particularly where individual sections or level coordinates are nominated, check to ensure the values are in
local coordinates (rather than world coordinates).

 In rare cases a seed-shape is produced that fails to meet the head-grade in the annealing process.
 If all else fails, pass a test model on to Technical Support with the Vendor, who will liaise with AMS to resolve the problem.
2.14.1 Discretisation
The default method of stope-shape evaluation is Fast (cell-centreline) evaluation. This method is fast and accurate. Two parameters are required:
(1) The model discretisation plane
(2) The model discretisation interval (along U and V axes)
A key consideration regarding discretisation is the cell filling plane. If the cell filling plane is not generally aligned with the orebody strike then there is potential for anomalous results
to occur. This is particularly true if the cell dimensions are disproportionately larger in the fill plane (i.e. at right angles to fill plane) when compared to the orebody width direction.
A case example where tall cell columns represent the orebody geology is illustrated in Figure 2-57. If the default discretisation parameters are used (4x4) then depending on the
evaluation method applied, Fast Evaluation or Precise Evaluation, the outcomes can be significantly different. The issue for Fast Evaluation, in this case example, is the influence
each cell column has as the cell’s centroid falls within or outside of the stope-shape during annealing. The best result would be obtained if subcells are orientated horizontally, not
vertically, for this case.
Figure 2-57 Discretisation Plane vs Cell Fill Plane Issue
The possible solutions are:
 Increase the level of discretisation to reduce the disproportionate dimensions,
 Change the orientation of the discretisation plane. The discretisation orientation plane is defined for rotated frameworks. The user can enable this option for un -rotated stope-
framework cases by enabling rotation parameters, but with zero rotations.
 Use a model that is filled perpendicular to the orebody strike plane.
2.14.2 Stope Control Surface
The easiest, most intuitive and most accurate method of specifying the strike and dip angles for the initial stope-seed-shapes is to provide a Stope Control Surface wireframe over
the full extent of the orebody where stope-shapes are to be generated. The Stope Control Surface has priority over the default dip and strike (Vertical XZ|YZ, Section XZ|YZ) and
default strike-dip and transverse-dip (Horizontal XY|YX) angles. Note that vendor software can generally display the true-dip and strike of the Stope Control Surface if required to
better understand the dip and strike orientations of the Stope Control Surface.
Stope Control Surfaces can typically use the geology surfaces depending on complexity. If the surface is overly complex with reversal of dip and/or strike orientations over short
distances the localised surface orientation may be at odds with the practical stope orientation and hence seed orientation. It is generally good practice to decimate the surface to
simplify it. An excellent alternative is to generate the wireframe from widely spaced sectional strings that are oriented to the expected stope orientation.
Multiple control surfaces are possible as the slice-seed is interpolated between the two surfaces. For example, if one surface is at 60 degrees and the other is at 50 degrees - a
slice-seed surface midway between the 2 surfaces would assume 55 degrees and as it got closer to the 60 degree surface it would approach 60 degrees and vice versa for the 50
degree surface. Outside of the 50 degree surface the seed-slice orientation would be 50 degrees and likewise for the 60 degree surface.
Crossover of control surfaces will have the same function as for separate sub-parallel control surfaces. However, always remember to check the end result.
2.14.3 Stope String Types
The following stope string types are produced:
 "Plan" (i.e. floor outline at base of stope).

 "Cross-section" (i.e. the stope end walls at the primary axis extremities) with optional intermediate stope cross-sections.
 "Section" (i.e. the stope near and far/hangingwall and footwall wall outlines).
Each string type can be uniquely coloured. Note that strings are not produced for the verification files.
2.14.4 Wireframe Types

2.14.4.1 Stopes
The Stope Shape Optimiser creates wireframe shapes from two 4-point cross section strings. Because the wall faces are not planar, an additional point is introduced on the walls
at the centre of the wall using bilinear interpolation (effectively an average of the 4 wall corner points). If one intermediate cross-section is output for each stope-shape, then this
wireframing procedure can be replicated with vendor wireframing software.
Without this additional string (from Version 2) there can be a significant source of reporting difference between SSO reports and reports from general mine planning packages that
only use the stope end walls for forming the stope wireframe.
It is possible to output and uniquely colour annealed shapes that failed to make it as a stope (failed stope) due to not meeting cut-off or other geometric / exclusion criteria (e.g. the
stope head-grade fell below the cut-off after dilution skins were added). It is sometimes very useful to identify the location (e.g. bridging low grade zones) and the extent of “failed
stopes” (e.g. failing due to excessive dilution) regarding setting of the SSO parameters.
These annealed shapes (i.e. stope seeds) will have the field record with “RESULT=0” in the case log-file and the report results file. They are generally yellow in colour
(COLOUR=4) by default.
2.14.4.2 Verification
The verification wireframe output is optional. It can output one or more of the following types:
 Seed-Slice (i.e. the slice intervals based on teh default strike and dip or Stope Control Surface),
 Seed-Shape (i.e. the aggregated slices),

 Failed Seed-Shape (i.e. aggregated slices that failed to meet stope criteria),
 Undiluted_stope (i.e. the annealed stope-shape prior to skin dilutions),
 Near dilution (i.e. the near side skin dilution shape for the Global definition case),
 Hangingwall dilution (i.e. the hangingwall side skin dilution shape for the Local definition case),
 Footwall dilution (i.e. the footwall side skin dilution shape for the Local definition case).
By filtering, individual features can be extracted for further analysis; a good example of this is evaluating the annealed stope-shape prior to the dilution skins. Each verification
wireframe type can be uniquely coloured.
2.14.5 Gradient Control Strings
The level gradient strings for sectional orientations are digitised in long section (i.e. along the framework primary strike U-axis). They should be non-overlapping. They should also
intersect every pair of stope sections to locate a stope floor and roof.
Need more help?

Go to www.maptek.com/support to access the Maptek Client Support page.
Have feedback on this help topic? Share it with us at [email protected]. Copyright© 2018 Maptek Pty Limited

Help Stope Optimiser

Uploaded by

Copyright:

Available Formats

Help Stope Optimiser

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Help Stope Optimiser

Uploaded by

Copyright:

Available Formats

2 Slice Method Page 1 of 30

 Vertical; XZ or YZ orebody orientations.

 U-axis is the primary "strike" direction of the stope-shapes (i.e. length),

Figure 2-1 Tube Definitions

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-2 Slice Method Stope - Shape Framework Hierarchical Structure

2.2 Stope Frameworks

2.2.1 Vertical Orientation Frameworks

Figure 2-3 Slice Method Framework - Vertical XZ

Figure 2-4 Slice Method Framework - Vertical YZ

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

2.2.2 Horizontal Orientation Frameworks

Figure 2-5 Slice Method Framework - Horizontal XY

Figure 2-6 Slice Method Framework - Horizontal YX

2.2.3 Transverse Section Orientation Frameworks

Figure 2-7 Slice Method Framework - Transverse Section XZ

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-8 Slice Method Framework - Transverse Section YZ

2.3 Framework Optimisation

2.3.1 Framework Iteration

2.3.2 Zone Iteration

Figure 2-9 Zone Iteration - Sectional Orientation Example

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-10 Zone Iteration - Horizontal Orientation Example

2.4 Control Strings

2.4.1 Vertical Orientation Frameworks

Figure 2-11 Gradient Control-String Example

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-12 OreDev Control-String Example

2.4.2 Horizontal Orientation Frameworks

2.5 Rotated Frameworks

2.5.2 Rotated Framework Combinations

Figure 2-13 Model - Framework Combinations

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-14 Rotated Framework Parameters - Vertical XZ Orientation Plane

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

2.5.3 Rotation Parameters

Table 2.1 Rotation Parameters

Parameter Description of Parameter

The X coordinate of known point in both systems, in the un-rotated coordinate

2.6 Stope-Shape Framework Extents Guidelines

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Table 2.2 Stope-Shape Framework Orientation and Extents Summary

Framework Orientation U-axis “strike” V-axis “height” W-axis “width” NX NY NZ

2.7 Stope-Shape Geometry

2.7.2 Vertical Orientation Stope-Shape Tubes

Table 2.3 Vertical Orientation Stope-Shape Tubes Summary

Stope-Shape Code Stope-Shape Framework Tube End-Face

Figure 2-15 (VR) Vertical Regular – Rectangular Stope-Shape Tubes

2.7.2.2 (VIL VIS) Vertical Irregular Framework (Rectangular Stope-Shape Tubes)

The irregular rectangle framework can have either:

Figure 2-16 (VIL, VIS) Vertical Irregular - Rectangular Stope-Shape Tubes

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

2.7.2.3 (VIVSG, VIFSG) Vertical Irregular Framework (Trapezoidal Stope-Shape Tubes)

Figure 2-17 (VIVSG, VIFSG) Vertical Irregular - Trapezoidal Stope-Shape Tubes

2.7.3 (HR) Horizontal Orientation Stope-Shape Tubes

Table 2.4 Horizontal Orientation Stope-Shape Tubes Summary

Stope-Shape Code Stope-Shape Framework Tube End-Face

mk:@MSITStore:C:\PROGRA~1\Maptek\VULCAN~1\help\Underground.chm::/Un... 25 jun. 2020

Figure 2-19 (HR) Horizontal Regular – Rectangular Stope-Shape Tubes