Lerchs-Grossman Algorithm

An ultimate pit limit method

 Block model.

 The deposit can then be represented by a set of blocks.

 With geostatistical modelling, it is possible to know the grade of each
block. Using known parameters such as ore price, cut-off grade, mining
cost, block mass etc., the economic value of a block can be calculated.
Feasible Pit.
• There is a slope angle constraint in a surface mining pit. We assume that the
slope angle is 45◦.
• In a 2-dimensional pit optimisation problem, to extract an underlying block,
you have to make sure that the blocks extracted above satisfy the slope angle
constraint.
• Example: To extract block A, you have to extract at least 3 blocks (B, C, D) in
the row above to satisfy the angle constraint. For blocks B, C and D, the same
principle applies, and so on until reaching the surface. You cannot extract
block J since only extracting the two blocks above (E and F) won’t satisfy the
constraint (90 degrees slope on the left!).
E F G H I

J B C D

A A
Optimal pit contour.

 The purpose of block modeling is to find the optimal pit contour and the
sequence of extraction.
 The optimal pit contour is composed of a set of blocks forming a feasible
pit, such that the selected parameter (usually the profit) is maximized.
This is called pit limit optimization.
 For instance, when adding the weight of the blocks forming the optimal
pit, the sum obtained should be the highest of all feasible pits.

E F G H I
Example: In this block model, EFGB, FGHC,
GHID, EFGHBC, or EFGHIBCDA etc. are all
B C D feasible pits. However, one of them has
the highest sum of weights and this one
A would be the optimal pit contour.
Graphic representation.
 After obtaining the block model, the first thing to do is to eliminate
the blocks on the side that can’t form any feasible pit (in blue).
E F G H I

B C D

 The blocks can then be represented by a set of nodes.

E F G H I

B C D

A
Graphic Representation.
 Edges are straight lines that connect nodes to the nearest overlying
neighbors.
 Arcs, as opposed to edges have a direction (arrowhead). An arc denoted
by (A,B) means that the flow goes from node A to B. The arc are always
directed to the surface.
 A set of nodes connected by arcs is called a directed graph G=(X,A).

E F G H I

B C D

A
Definitions
 Path: Sequence of arcs such that the terminal node of each arcs is the initial
node of the following one
 Circuit: Path in which the initial node is the same as the terminal node
 Chain: Sequence of edges in which each edge has one node in common with
the following one
 Cycle: Chain in which the initial and final node coincide
 Tree: Connected and directed graph with no cycles. A tree is denoted Tn. A
rooted tree is a tree with a special node, called the root
 Root: Node selected from a tree, usually denoted as X0. A tree can only have
one root.
 Dummy arc: Arc connected to the root node.
 Branch: When a tree is cut into two part, the part of not containing the root
is called a branch.
Solving the graph.
Step 1 : Assign weight to the nodes and build the initial Tree T0 by
connecting all nodes to the root node Xo with arcs.

E F G H I 2 -1 -3 -2 -6

B C D 8 -11 1

A 10

X0
Step 2 : Label nodes using the following information's :

Direction Cumulative weight Label

plus + Strong
plus - or 0 Weak Plus: directed away from the root.
minus + Weak Minus: directed towards the root.
minus - or 0 Strong

Initially, all nodes in the initial tree are “plus” as they are all directed away
from the root.

2 -1 -3 -2 -6
Distinguish two groups in this rooted initial S+ W+ W+ W+ W+
tree : The group TA that contains all strong-
plus (S+) nodes connected to root and the
8 -11 1
group TB that contains all weak-plus (W+) S+ W+ S+
nodes connected to root.

10
S+

X0
Important note
 The cumulative weight of an arc is the sum of all the nodes supported by
this arc.
 Example:

A2 is directed away from

the root.
A1 is directed away from
the root.

 The cumulative weight of arc A1 is -3+5+4=6.

 The cumulative weight of arc A2 is 4.
Important Note.
 The goal of the Leachs-Grossman algorithm is to find all possible
connections between group TA (S+ nodes) and group TB (W+ nodes)
 All strong arcs not directly connected to the root should be treated:
• All strong plus (S+) arcs should be connected directly with the root x0 .
• A strong minus (S-) arc should be connected directly with the root x0.
Step 3 : Iterations.
 Always start with the second row from surface. Take a S+ node and connect it with an
upper W+ node (Iteration #1)
 RULE # 1: If the cumulative weight is positive after adding the W+ node for the same
S+ node, keep the S+ node arc connected to the root, and remove the W+ node arc
connected to the root (Iteration #1).

 Iteration #1:
Positive Cumulative weight = 8+(-1)= 7

2 -1 -3 -2 -6 2 -1 -3 -2 -6
S+ W+ W+ W+ W+ S+ W+ W+ W+ W+

8 -11 1 8 -11 1
S+ W+ S+ S+ W+ S+

10 10
S+ S+

X0 X0
 In Iteration #1, the cumulative weight was positive. We can connect the same
S+ node with another upper W+ node (Iteration #2).

 The cumulative weight after adding the second W+ node is still positive. (See
Rule #1).

 Iteration #2:
Positive Cumulative weight = 8+(-1)+(-
3)= 4

2 -1 -3 -2 -6 2 -1 -3 -2 -6
S+ W+ W+ W+ W+ S+ W+ W+ W+ W+

8 -11 1 8 -11 1
S+ W+ S+ S+ W+ S+

10 10
S+ S+

X0 X0
 In iteration #2, the same S+ node doesn’t have any possible connection left
with a W+ node. We proceed at the next S+ nodes of the same row.
 RULE #2 : If the cumulative weight is negative or 0 after adding a W+ node for
the same S+ node, keep the W+ node arc connected to the root and remove
the S+ node arc connected to the root. The S+ node becomes a W- node.
Therefore, no further connection is possible for this node. (Iteration #3).

Negative Cumulative weight =1+(-2)=-1

 Iteration #3:
2 -1 -3 -2 -6 2 -1 -3 -2 -6
S+ W+ W+ W+ W+ S+ W+ W+ W+ W+

8 -11 1 8 -11 1
S+ W+ S+ S+ W+ W-

10 10
S+ S+

X0 X0
 There is no more S+ nodes in the second row. We now proceed in the third row (
row below).
 We follow the same steps/rules as the two last iterations.
 Reminder : We have to make the most possible connection between the TA group
(S+ nodes) and the TB (W+ nodes).

 Iteration #4:

Negative Cumulative weight =10+(-11)=-1

2 -1 -3 -2 -6 2 -1 -3 -2 -6
S+ W+ W+ W+ W+ S+ W+ W+ W+ W+

8 -11 1 8 -11 1
S+ W+ W- S+ W+ W-

10 10
S+ W-

X0 X0
Step 4: Ultimate pit.
After finishing the third row, we notice that all possible connections
between TA group (S+ nodes) and the TB (W+ nodes) have been made.
The ultimate pit will be all the nodes connected to the S+ nodes tree.
The final pit has a weight of 8+2-1-3=6.

2 -1 -3 -2 -6
S+ W+ W+ W+ W+

8 -11 1
S+ W+ W-

10
W-

X0
Assignment
• Find the ultimate pit limit and the corresponding closure for the
following block models.
1. First block model:

2 -1 3 -2 -1 2 1 -1 -1
4 2 -3 3 1 -1 -2
1 -1 -3 1 5
5 1 -2
2. The grades in g per ton of each block of a gold deposit is given for the
block model below. The weight of a block is 1000 tons. The mining and
processing cost are respectively 6$/t and 27 $/t. The recovery is 85% .
• Determine the ultimate pit limit for three iterations. In first iteration, use:
50 $/g (parametrization factor of 0.80; 50*0.8= $40 and 40*0.8=$32)

0.12 0.68 0.12 0.65 0.62 0.70 1.15 1.17

0.63 1.2 1.29 0.64 3 0.68 1.13 1.57
1.03 1.01 0.66 1.4 0.67 0.65 0.3 0.4
0.2 0.4 1 0.71 0.64 1.3 0.4 1.16

𝑃𝑟𝑜𝑓𝑖𝑡 = 𝑝𝑟𝑖𝑐𝑒 × 𝑟𝑒𝑐𝑜𝑣𝑒𝑟𝑦 × 𝑡𝑜𝑛𝑛𝑎𝑔𝑒 × 𝑝𝑎𝑟𝑎𝑚𝑒𝑡𝑟𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑓𝑎𝑐𝑡𝑜𝑟 × 𝑔𝑟𝑎𝑑𝑒

− 𝑚𝑖𝑛𝑖𝑛𝑔 𝑐𝑜𝑠𝑡 + 𝑝𝑟𝑜𝑐𝑒𝑠𝑠𝑖𝑛𝑔 𝑐𝑜𝑠𝑡 ∗ 𝑡𝑜𝑛𝑛𝑎𝑔𝑒