Pit Optimiser Tutorial

Overview
An optimal pit for a deposit is one that maximises profit and satisfies both cost
and slope constraints. Designing an optimal pit is a challenging and complex
problem that is difficult to solve. As a result, several mathematical algorithms
have been developed to solve this problem in a reasonable time frame. The
Pit Optimiser developed by the Surpac Minex Group is based on a variation of
the Lerchs Grossman algorithm that was first proposed by Koenigsberg in
1985.

The Pit Optimiser works on a block model of the deposit where each block
must have a net value that represents the economic value that will be
returned if that block is extracted in isolation. The Pit Optimiser then considers
each of these blocks in turn to work out which combinations of blocks should
be mined in order to return the highest possible total value given mining
constraints for a particular sale price. The result is a 3D surface that
represents the base / limit of the pit that maximises the total value of the mine
and any further extension of the pit will not increase the total return.

Concepts
Optimisation Theory

Pit optimisation works on a block model of the deposit to be mined and the
surrounding material. With any block model, only a subset of blocks will
actually contain material of sufficient grade to make them worthwhile to mine
once they are exposed at the surface. In order to reflect this information, each
block must be assigned a value that essentially shows the net cash flow that
would result from mining that block in isolation. This value must be positive for
an ore block and can be calculated as the sale price minus the costs of mining
and milling. For waste and air blocks, this value will be zero or a negative
value representing the cost of mining that block.

To illustrate this, the diagram below shows a cross section of a deposit. For
simplicity, all air blocks in the block model for this ore body will be given a
value of $0, the mining cost associated with extracting an ore or waste block
will be set at $3, while the sale price for each ore block will be $20.

1
The diagram below shows a cross section of the block model for this deposit.
Each block has been assigned a net value indicating the cost of extracting the
block if it were already exposed.

Once all the blocks have been assigned a net value showing the cost of
extracting just that block as if it were already exposed, each block is then
considered in turn and all the blocks that are needed to be mined in order to
uncover it are identified. From these blocks, it is then possible to calculate the
total net value of a block in relation to its position in the model. This will be the
net value of the block in question minus the cost of mining all the blocks that
need to be mined before it can be extracted.

2
In the above example, we have assumed that in order to mine a block, the
block directly above it must be extracted first. Therefore, to calculate the total
net value of a block, you would take the net value of the block and subtract
the net value for all the blocks above it.

After the total net value for each block is calculated, it is then a matter of
finding the combination of blocks to extract that will result in the maximum
return for that deposit. As shown below, the maximum value for a vertical pit
(90 degree slope angle) for this deposit is $40. This is computed by adding up
the total net values of the blocks along the base of the pit outline.

If more blocks were extracted, the total return of the pit would actually
decrease as shown below. When an extra column of blocks is extracted, the
value of the pit is $39.

3
Likewise with a smaller pit, the total return is only $38. This is illustrated
below.

An optimal pit is one which returns maximum revenue. A slightly smaller pit
will leave revenue on the pit walls, while a slightly bigger pit is unprofitable at
its limit.

In pit optimisation, wall slope angles are an important consideration when

calculating the optimum pit because these angles determine which blocks
need to be mined before a certain block can be extracted. This in turn affects
the total net value for each block and will result in different blocks forming the
optimal pit. The above simplified case assumed a slope angle of 90 degrees

4
which is unrealistic. If on the other hand, a slope angle of 45 degrees is
assumed, the following results would be achieved.

The maximum value for a 45 degree slope pit would give a value of $13 and is
once again calculated by adding up all the total net values of the blocks along
the base of the optimum pit. Regardless of the slope angle used, a slightly
smaller or larger pit would result in less return as shown in the two diagrams
below.

5
As can be seen, finding the optimal pit for a deposit is not a simple process.
There are many different combinations and sizes of pit outlines that need to
be assessed before it is possible determine which pit will return the most profit
for a given sale price.

This process can be repeated by increasing the sale price to calculate more
profitable pits closer to the surface that are shorter term (2 to 3 years). In
effect, a series of nested pit shells can be produced and used as a rough
guide for determining the schedule of the mine or estimating the maximum net
present value of the mine.

The two algorithms that the Surpac Minex Group pit optimiser uses are the
Floating Cone and Lerchs Grossman techniques to determine an optimal pit
for a deposit. These two algorithms are discussed below.

Floating Cone Algorithm

Refer to Jon Barber’s paper

Lerchs Grossman Algorithm

Refer to Jon Barber’s paper

Prerequisites

To use the pit optimiser, it is necessary to be familiar with the basic principles
of Surpac Vision, the block model module and the geological database
module in order to estimate quality data into the block model. If you are not
familiar with these modules, it is recommended you go through the tutorials
(MC – add link) on these topics before commencing this tutorial.

6
Data Requirements and Block Model Preparation
The minimum data requirements for the Pit Optimiser are:

 Block model containing quality data and an attribute indicating which

blocks are ore and waste blocks in the model or alternatively, just an
attribute representing the net value for a block. The net value for a
block is the net cash flow generated after the block has been extracted
and processed, and can be calculated as follows:

Block Net Value = Sale Price – Mining Costs – Milling Costs

where

Sale Price is the price of the material that it can be sold at. For ore
material, this will be a positive value as it can be sold and for waste
and air material, this will be zero because they cannot be sold.

Mining Costs refer to the cost of extracting the block as if it were the
only block in the model. It does not include the cost of extracting the
immediate blocks above it. Both ore and waste blocks will have a
mining cost associated with it. Air blocks cost nothing to extract and
therefore have no mining costs.

Milling Costs represent the cost of processing ore material through

the mill. Only ore blocks will have a milling cost component and hence,
waste and air blocks have no milling costs.

From the above calculation, ore blocks will always have a positive net
value while waste blocks will have a negative net value and air blocks
have a zero net value.

The net value must be expressed in either dollars per mass unit (eg.
$/tonne) or dollars per volume unit (eg. $/m3) of ore. This will be
discussed in more detail later.

 Topography DTM surface

Once all the necessary attributes have been added to the model and the
topography surface has been prepared, the pit optimisation parameters can
be entered and the pit optimiser will produce a pit shell or series of nested pit
shells representing the base of the optimal pit at given sale prices.

Exercise

Objective

7
Generate an optimal pit for a ore deposit where the net value of each block in
the model must be calculated automatically by the Pit Optimiser from given
sale prices and costs. As an additional requirement, the resultant optimal pit
must have wall slope angles of no greater than 45 degrees around the entire
wall of the pit.

The data files required for this exercise are:

gold.mdl
topo1.dtm
topo1.str

These files are located in the dem/training/pitopt directory where Surpac

Vision has been installed.

1. Set your current working directory to be the directory containing the files to
be used for this exercise. To do this, left click with the mouse on the
directory in the navigator and then right click to pop up a menu. Select Set
As Work Directory from this menu.

2. Display the block model menu by right clicking with the mouse at the end
of the main menu bar (to the right of the Help menu). When discussing any
block model functions in this tutorial, they will be referred to from this
menu.

3. Open the block model file gold.mdl by selecting this file from the navigator
with the left mouse button and while holding down this button, and
dragging the mouse pointer into the graphics viewport. The block model
status item should appear at the bottom of the Surpac window as shown
below:

4. To find out information about this block model, select Block Model –
Summary from the block model menu. This will display a form
summarising the properties of the block model and should appear as
follows:

8
 The important features of the block model to note are the user block size
of 20m x 20m x 20m and that four levels of sub-blocking have been
allowed making the minimum possible block size 5m x 5m x 5m.

In addition, it can be seen from the list of attributes that we are dealing
with an ore deposit where the background value of gold is -99.00 and
there is also an ore_type attribute that classifies the different ore materials
in the model by an integer value.

From the information in the block model, it can be seen that there is no
attribute containing the net value for each block and so the Pit Optimiser
will have to automatically calculate it.

Press Apply on the form to close it.

5. In order to visualise the ore deposit that is to be mined, select Block

Model – Display from the block model menu to display the entire block
model. The following form should appear:

9
 Fill in the form as displayed above and press Apply to draw the model
with block faces. The model should be centred in the graphics viewport as
the Rescale option has been checked and should look similar to the image
below:

6. To get a better indication of where the ore deposit is located in the block
model, a graphical constraint can be added to display on the blocks where
the gold value is greater than zero. To do this, select Constraints – New
Graphical Constraint from the block model menu.

Fill in the constraint form as follows:

10