Alex Miranda and Beck Nader.

Mining
Mineração
Direct sequencing of blocks in
stochastic models with multi-
http://dx.doi.org/10.1590/0370-44672018720181 mines and multi-destinations
Alex Miranda1,2 Abstract
https://orcid.org/0000-0001-8606-7705
Beck Nader1,3 Mining Scheduling is the one that maximizes profit from mining over time. By
https://orcid.org/0000-0002-7215-7379 means of computational methods, the deposit is discretized in blocks and algorithms
are used to consummate this objective. The methods that are widely known nowadays
1
Universidade Federal de Minas Gerais - UFMG, for mining scheduling optimization of a discrete block model were based on graph the-
Escola de Engenharia, Departamento de Engenharia ory, and among those most used by the mining industry is a solution found by Lerchs
de Minas, Belo Horizonte - Minas Gerais - Brasil. and Grossmann (1965) which, together with other methods, was consolidated as the
process of traditional mining planning. For both supply and multi-destination models,
E-mails: [email protected], [email protected], there is a limitation of the current methodology, since it consists of the optimization
[email protected] of each mine separately which may not be a global optimization solution. This article
proposes an optimization of the benefit in a stochastic model through the DBS (Direct
Block Scheduling) for a copper mining complex with two mines, a pre-existing copper
stockpile and two treatment streams, comparing several scenarios and analyzing the
best alternative for the proposed problem.

keywords: mining planning, direct block sequencing, mining.

1. Introduction

Linear programming techniques this method impossible. The algorithm of this model is as follows: • b ∈ B: set
for mine scheduling were pioneered of Lerchs and Grossmann (1965), for of all B blocks.
by Johson (1968). Johson developed having its computational implementa- • t ∈ T: set of periods in a time
a mathematical model to solve the tion performed by Whittle (1993), has frame T.
problem of long-term programming been used by various mining planning • vbt: value associated with extrac-
using the Dantzig-Wolfe - originally packages ever since. The process of tion of the block b in period t.
developed by American mathematicians scheduling the mining process involves • cb: resource consumption associ-
George Dantzig and Phil Wolfe - (1960) the removal of at least two types of ated with block extraction b (ton).
principle that uses the algorithm for the materials: ore and waste. The mining • C , C : minimum, maximum re-
decomposition of the complex master problem, then, can be modeled in terms source tied to any period (ton).
problem into subproblems. The com- of entire programming (HOCHBAUM e • ybt: 1 if block b is extracted in
putational time for processing made CHEN, 2000). The typical formulation period t, 0 otherwise (variable).
REM, Int. Eng. J., Ouro Preto, 72(4), 661-666, oct. dec. | 2019 661
Direct sequencing of blocks in stochastic models with multi-mines and multi-destinations

Max vbt ybt

(1)
b∈B t∈T

subject
ybt ≤ 1∀ b, (2)
t∈T

C≤ cb ybt ≤ C ∀ t, (3)
b∈B

t
,
ybt ≤ ybt ∀ b, b ∈ Bb, t, (4)
T=1

ybt ∈ {0,1} ∀ b,t. (5)

The illustrated method is a simple net present value as the optimization crite- optimize the mining production program
example of modeling, and it is possible to rion. Ravenscroft (1992) suggested a new by means of a fixed-stages Stochastic-
include stockpiles, blending, etc. Over the approach when proposing that produc- Integer Programming (SIP) model that
years several algorithms have been devel- tion scheduling should be planned in con- successfully optimizes open-pit mine
oped to solve the problem of the mining ditionally simulated alternative models scheduling under geological uncertainty.
sequencing. Gershon (1983) and Barbaro to verify the sensitivity of the production In this research, the general objec-
and Ramani (1986) have proposed mixed schedule in relation to the uncertainty of tive is the application of a methodol-
integer programming (MIP) to optimize the contents. Smith and Dimitrakopoulos ogy for the sequencing of mining with
long-term production scheduling models. (1999) proposed an approach in which simulated models, using computational
Seymour (1994) and Tolwinski (1998) several optimal sequencings are obtained, tools based on surface mining to solve
presented approaches based on dynamic one for each simulated model. the problem, evaluating the influence of
programming methods for optimizing the Ramazan and Dimitrakopoulos some variables (such as the inclusion or
production sequence using the maximum (2012) propose a general formulation to not of stockpiles).

Surface based mine scheduling

A mine sequencing can be under- in SIP solution formulations because variable xzct = 1 if mining in period t and
stood as a set of surfaces in the space it is not necessary to do precedence of 0, otherwise. A fundamental aspect of
that divides the orebody into parts to block extraction. this approach is the need to associate
be extracted during different periods The centroids of the blocks be- blocks with surface cells and to make
(ALMEIDA, 2013). This surface- come part of a network of measured comparisons of their elevations in
based approach facilitates the manage- points of a topographic surface and can space. The formulation of this model is
ment of slopes, improving efficiency therefore be represented by a binary detailed below:
S T M Z

Max 1 z
Vc,t,s ( xc,tz - x c,t
z
-1) (6)
S
S=1 t=1 c=1 z=1

This function is a variation on the the number of surfaces to be considered; z: level index, z = 1, 2, ..., Z. V zc,t,s is the ac-
model of Ramazan and Dimitrakopoulos t: index of periods, t = 1, 2, ..., T. M is the cumulated discounted value of the block (c,
(2012)unlike equation (1), there is inserted amount of cells on each surface, where z) and of all the blocks above the scenario
the term S referring to the number of M = x.y represents the total of blocks in s and period t. x zc,t is the binary variable
simulations of the mineral body , s: index both x and y dimensions; c: location index that assumes value 1 if the block (c, z) is
of each simulation, s = 1, 2, ..., S. T is of each cell / block, c = 1, 2, ..., M. Z is the the last block being drawn in period t over
the number of periods that also defines number of levels in the orebody model; c and 0 (zero) otherwise.

2. Materials and methods

SimSched software was used as graphic software Studio OP (Datamine) to reduce the processing time of the two
a tool to optimize the sequencing. As was used. The block models were granted mines, only 15 Cu grade simulations were
for the manipulation of the models and by Datamine containing oxide and sulfide used (randomly chosen only to illustrate
construction of the benefit function, the ore with 40 Cu grade simulations, but the case study).

2.1 Block Model

The block model of mine 01 (Fig- regularization, the two block models created to distinguish one mine from the
ure 1) was regularized, since it was were merged in order to perform the other (MINA field).
composed of irregular blocks. After its study with a single model and a field was Mine 02 (Figure 2) consists of oxide
662 REM, Int. Eng. J., Ouro Preto, 72(4), 661-666, oct. dec. | 2019
 Alex Miranda and Beck Nader.

and sulfide ore, but with estimates of sented the 2D and 3D visualization of the
only the oxide ore grade. Below are pre- model with copper grade for this mine.

Figure 1
Detail of the "box" of the mine 01
block model with a vertical cut (Cu grade).

Figure 2
Detail of the "box" of the mine 02
block model with a vertical cut (Cu grade).

2.2 Benefit Function

The benefit function was elaborated the basis of some feasibility studies done the values used to make the value of each
according to parameters determined on by the author himself. The Table 1 with block is presented below.

Economic Parameters
Value Unit
Copper price 1.7 US$/lb
Mining cost 1.95 US$/t
Rehandling cost* 0.05 US$/t
Flotation cost 3.5 US$/t
Leach cost 3.4 US$/t
Selling cost 10.0 US$/t
Table 1
Economic parameters used in the model. *stock pile

For each Cu grade simulation, there treatment circuit (or route) considered ac-
was calculated the mass recovery in each cording to the Table 2 below:
Mass recovery by process
Range grade (Cu) Value Unit
0 > %Cu < 0.2999 R = 2.33 x %Cu R = 2.16 x %Cu
0.3 > %Cu < 0.5999 0.70 0.65
Table 2
Mass recovery according Cu grade. %Cu > 0.6 0.82 0.65

2.3 Configuring multiple destinations and scenarios

As defined a priori, the treatment alternative are: for later feeding.
process consists of two destinations cor- • Scenario 01: Direct feed of ore in • Scenario 03: It considers a pile
responding to the two possible processes each plant, whereby the oxide ore can go of pre-existing oxide ore (with a certain
for concentrating the R.O.M mass from to the flotation or leach circuits and the capacity and only Cu grade average) that
the mines. Three (3) scenarios were con- sulfide ore can only go to the leach circuit; can feed the plant directly at any time of
sidered for comparing NPV values in each • Scenario 02: Same as scenario 01, the mining sequence. The flowchart of this
alternative. The characteristics of each but the sulfide can be stored in a stockpile scenario is presented (Figure 3):
REM, Int. Eng. J., Ouro Preto, 72(4), 661-666, oct. dec. | 2019 663
Direct sequencing of blocks in stochastic models with multi-mines and multi-destinations

Figure 3
Simplified Scenario 03 Flowchart.

3. Results

For scenario 01, 170.8 Mt of ore and the accumulated strip ratio was total life of mine was 20 years. However,
(oxide and sulfide) was mined and the 1.70 t/t, and also the useful life of mine there were already 2 Mt of ore available,
accumulated strip ratio was 1.65 t/t was 19 years. In scenario 03 (Figure 6), therefore, only 177 Mt were mined in the
and the useful life of mine reached 19 179 Mt of ore was mined with an accu- mine. A summary chart of each scenario
years. For scenario 02 (Figure 5), ap- mulated strip ratio of 1.71 t/t. With the comparing the R.O.M. and the strip ratio
proximately 175.4 Mt of ore was mined, strategy of having an initial stockpile, the of each scenario is presented (Figure 4).

Figure 4
Comparison between the studied scenarios.

4. Discussion

Each scenario attempted to meet compared, and it was concluded that 3 and Figure 5). The table with the NPV
the need for the subsequent scenario. the most favorable scenario from the and the variation of the minimum and
After evaluating each strategy, the financial return point of view would be maximum value in each alternative is
NPV variations of all scenarios were scenario 02 (using the stock pile - Table presented below:
664 REM, Int. Eng. J., Ouro Preto, 72(4), 661-666, oct. dec. | 2019
 Alex Miranda and Beck Nader.

Accumulated NPV ( US$ x 106 )

Scenario Min Max Expected % D (max-min)
01 86,615.3 104,354.0 91,466.0 20%
Table 3
02 100,374.0 120,123.0 105,292.0 20%
NPV values (with their
03 92,630.4 110,955.0 97,626.0 20% variations) for each studied scenario.

Figure 5
Comparison of NPV
variation in different scenarios.

For scenario 01 (Figure 6), the tra- was done considering the average of the 18th period (last year sequencing by
ditional sequencing was done considering 15 simulations (E-type) and the use of traditional methodology). shows that
the average of the 15 simulations (E-type) the algorithm of Lerchs and Grossmann there is a difference of 11% between the
and the use of the algorithm of Lerchs (LG). The chart below shows that there NPV values accumulated until the end of
and Grossmann (LG). The chart below is a difference of 11% between the NPV the 18th period (last year sequencing by
(Figure 6), the traditional sequencing values accumulated until the end of the traditional methodology).

Figure 6
Comparison of NPV
between scenario 01 and E-type (LG).

5. Conclusions

It was observed that DBS is an excel- • Obtaining the results in a single with deterministic models, it is also pos-
lent tool to solve the problem of mine se- step of work, without the need to first find sible to work with stockpiles and maximize
quencing with multiple simulation grades the best economic envelope and sectoriza- the project value;
and that, in addition to working with these tion of pit in portions and then perform the • In the software used (SimSched)
simulations, it was shown that it has the mine sequencing; it was also possible to include some op-
ability to process different scenarios with • With the currently existing com- erational restrictions such as vertical rate
lower and higher complexity degrees. puter tools, you can do multiple analyzes advance and minimum width for mining
Among the main advantages of the ap- with different inputs in different scenarios and bottom, which makes the solution
proach with multiple models and multiple and SimSched software was robust enough closer to the operational reality of mining;
destinations that currently are imposed on to generate these scenarios; • As the operational details and
mining projects, we can highlight: • Analogous to packages that work blending restrictions are incorporated into
REM, Int. Eng. J., Ouro Preto, 72(4), 661-666, oct. dec. | 2019 665
Direct sequencing of blocks in stochastic models with multi-mines and multi-destinations

the sequencing, the most intense is the can be processed together. Given that the scenarios. However, the ranges of variation
delay in finding the solution. studied scenarios were able to reach their in each scenario allow to affirm that, with
Despite the advantages mentioned objectives (without exceeding the capacity small adaptations in supply (geological
above, there is still the drawback with limit of the mine and feeding the plant's model) and demand (metal price) policies,
the processing time for more demanding feed flows), it was observed that scenario it will lead to a variety of alternative sce-
processes in terms of complexity and the 02 - Figure 5 (strategy of using stockpiles) narios and opportunities for greater gains
need for the models to be composed of managed to obtain a higher economic that can be exploited in future implementa-
blocks with the same size so that they return expected among the three studied tions through sensitivity analysis.

References

ALMEIDA, A. Surface constrained stochastic life-of-mine production scheduling.

Montreal: MCGill University, 2013. p. 191.
BARBARO, R. W., RAMANI, R. V. Generalized multiperiod MIP model for produc-
tion scheduling and processing facilities selection and location. Mining Enginee-
ring, p. 107-114, February 1986.
DANTZIG, G. B., WOLFE, P. Decomposition principle for linear programs. Opera-
tions Research 8. p. 101-111, 1960.
GERSHON, M. E. Mine scheduling optimization with mixed integer programming.
Mining Engineering, p. 351-354, April 1983.
HOCHBAUM, D. S., CHEN, A. Performance analysis and best implementations of
old and new algorithms for the open-pit mining problem. p. 894-914, 2000.
JOHNSON, T. B. Optimum Open Pit Mine Production Scheduling. Berkeley, 1968.
p. 120.
LERCHS, H., GROSSMANN, L. Optimum design of open-pit mine. Montreal:
1965. p. 17-24.
RAMAZAN, S. D., DIMITRAKOPOULOS, R. Production scheduling with uncer-
tain supply: a new solution to the open pit mining problem. Optimization and
Engineering, 2012.
SMITH, M., DIMITRAKOPOULOS, R. Influence of deposit modelling on mine
production scheduling. [S.l.], p. 173-178. 1999.
SEYMOUR, F. Finding the mining sequence and cutoff-grade schedule that maximizes
net present value. 1994. p. 1880-1884.
TOLWINSKI, B. Scheduling production for open pit mines. London: 1998.
p. 651-662.
WHITTLE, J. Four-D Whittle Open Pit Opimisation Software. Melbourne: Whittle
Programming Pty Ltd, 1993. (User Manual).

Received: 10 December 2018 - Accepted: 16 July 2019.

All content of the journal, except where identified, is licensed under a Creative Commons attribution-type BY.

666 REM, Int. Eng. J., Ouro Preto, 72(4), 661-666, oct. dec. | 2019