Vous êtes sur la page 1sur 5

M INING

SCIENCE AND
TECHNOLOGY
Mining Science and Technology 20 (2010) 06910695
www.elsevier.com/locate/jcumt

Mineral processing plant location using the analytic hierarchy


processa case study: the Sangan iron ore mine (phase 1)
SAFARI Mohsen, ATAEI Mohammad*, KHALOKAKAIE Reza, KARAMOZIAN Mohammad
Faculty of Mining Engineering, Petroleum & Geophysics, Shahrood University of Technology, Shahrood, Iran

Abstract: Locating the mineral processing plant near a mine is the most important parameter that affects the whole process. Many
factors, and their preferences, should be considered in this stage. The factors include economical, geological, technical, environmental
and tectonic parameters. A multi-criteria decision making method is necessary to rank the alternatives. In this paper we describe
how plant location is selected by using the Analytic Hierarchy Process (AHP). This method, with eight criteria, was used to select a
location for the mineral processing plant at the Sangan iron ore mine (phase 1). Three alternatives for the processing plant were
evaluated. The main criteria were distance from the mine, access to heavy machinery transport, the amount of excavation required
for grading, bed mixture capacity, belt conveyor length, distance from the tailing dam, distance from the waste dumps and surface
water diversion requirements. Finally, the alternatives were ranked and the best location was proposed.
Keywords: mineral processing; multi-criteria decision making; Sangan iron ore mine; AHP

1 Introduction reported yet.


Numerous factors are involved in mineral proces-
The selection of a location for mineral processing sing site selection most of which are presented in Fig.
is a critical factor during the design stage. This is one 1. The factors with the greatest impact should be
of the most important parameters to have an impact chosen for analysis during site selection.
on the whole process. Proper site selection decreases
costs and the destructive effects on the environment
for the entire life of the plant. In most cases all the
factors involved in site selection have not been taken
into consideration so the optimum alternative can not
be chosen. Processing-plant site selection is a
function of many parameters; therefore it can be
considered as a decision making process. This
problem involves finding the best option among all
the feasible sites, or a ranking of the options. Over the
past decades many methods, such as simple additive
weighting, order preference by similarity to the ideal
solution and the analytical hierarchy process, have Fig. 1 Important factors for processing plant site selection
been developed to deal with the multiple decision
problem[1-3]. Among these methods, the Analytical
2 Analytical Hierarchy Process (AHP)
Hierarchy Process (AHP) is the most powerful and
the simplest method for use in solving multiple The AHP is a multi-criteria decision method that
criteria decision problems. This method has been uses hierarchical structures to represent a problem.
used for a variety of specific applications in decision Priorities for alternatives are then developed based on
making involving mining operations[4-6]. Its applica- the judgment of the user[7]. The AHP procedure
tion to mineral processing site selection has not been involves six essential steps[8]: defining the unstruc-
tured problem, developing the AHP hierarchy,
Received 15 January 2010; accepted 05 April 2010 pairwise comparison, estimating the relative weights,
*Corresponding author. Tel: 98 273 3395509 checking the consistency and, finally, obtaining the
E-mail address: ataei@shahroodut.ac.ir
doi: 10.1016/S1674-5264(09)60264-7
overall rating.
692 Mining Science and Technology Vol.20 No.5

2.1 Define the unstructured problem 2.4 Estimating the relative weights
In this step the unstructured problem and its Some technique, like the eigenvalue method, is
characteristics should be recognized and the objec- used to calculate the relative weights of elements in
tives and outcomes stated clearly. each pairwise comparison matrix. The relative
2.2 Developing the AHP hierarchy weights, W, of matrix A are obtained from:

The first step in the AHP procedure is to decom- ( A max I ) W = 0 (2)


pose the decision problem into a hierarchy that con- where max is the biggest eigenvalue of matrix A and
sists of the most important elements of the decision the unit matrix.
problem[9-11]. In this step the complex problem is de-
composed into a hierarchical structure with decision 2.5 Consistency checking
elements (the objective, the attributes i.e., criteria and In this step the consistency of the matrices is
alternatives). Fig. 2 represents this structure. checked to ensure that the judgments of decision
makers are consistent. The Inconsistency Index (I.I)
is calculated as:
n
I.I = max (3)
n 1
The inconsistency index of a randomly generated
reciprocal matrix shall be called the Random
Inconsistency Index (R.I.I) with reciprocals forced.
Table 2 shows the random inconsistency index for
matrices of the order from 1 to 10.
Fig. 2 Hierarchical structure of the decision problem
Table 2 Random inconsistency indices[8]
2.3 Pairwise comparison No. of
1 2 3 4 5 6 7 8 9 10
criteria
For each element of the hierarchy structure all the R.I.I 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49
associated elements in low hierarchy are compared in
pairwise matrices as: The Inconsistency Ratio (I.R) is the last ratio that
has to be calculated. The formulation of the I.R is:
w1 w1
1 w2

wn I.R=
I.I (4)
R.I.I
w2 1 
w2
The closer the inconsistency index is to zero is the
A = w wn (1) greater the consistency. If it is high, that means the
1
    input judgments are not consistent: hence they have
w wn
n  1 to be elicited again. In general an inconsistency ratio
w1 w2 of about 10% or less is usually considered acceptable
and then the derived weights may be used[8].
where A is the comparison pairwise matrix; w1 the
weight of element 1; w2 the weight of element 2 and 2.6 Obtaining the overall rating
wn is the weight of element n. In the last step the relative weights of the decision
To determine the relative preference for two ele- elements are aggregated to obtain an overall rating for
ments of the hierarchy in matrix A an underlying se- the alternatives as follows:
mantic scale is employed having values from 1 to 9 m

(Table 1). Wi s = WijsW ja (i = 1, 2, ..., m) (5)


j =1
[8]
Table 1 Scales for pairwise comparison
where Wi s is the total weight of alternative i; Wij s
Preferences
expressed in Preferences expressed in linguistic variable the weight of alternative i associated to attribute j;
numeric variable
W j a the weight of attribute j; m the number of
1 Equal importance
attributes and n the number of alternatives.
3 Moderate importance
5 Strong importance 3 Sangan iron ore mine
7 Very strong importance
9 Extreme importance The Sangan iron ore mine is located 16 km north of
Intermediate values between adjacent scale Sangan and 300 km southeast of Mashhad in
2, 4, 6, 8
values Khorasan Razavi province, Iran (Fig. 3). Its location
SAFARI Mohsen et al Mineral processing plant location using the analytic hierarchy process 693

is 6016 longitude and 3424 latitude. The total Table 3 Criteria for processing plant site selection
geological reserve of the Sangan iron ore mine is es- Criterion Operation
timated to be near 1.2 billion tons with an average C1 Bed mixture capacity
grade of 50 percent Fe3O4. The annual iron ore pro-
C2 Access to heavy machinery transport
duction from this mine is 4.5 million tons and the
mineral processing plant is designed to produce 2.6 C3 Belt conveyor length

million tons of iron pellets per year in phase one of C4 The amount of excavation for grading the ground
the project. C5 Distance from waste dumps
C6 Distance from tailing dam
C7 Surface water diversion requirements
C8 Distance from the mine

5 Processing plant site selection using the


AHP method

The AHP procedure involves these steps:


First develop a hierarchy by breaking the problem
down into components. The three major levels of the
hierarchy are the goal, the objectives and the alterna-
tives. The hierarchy for the selection of the Sangan
processing plant is illustrated in Fig. 5. As can be
seen from this figure selection of the appropriate
Fig. 3 Location of the Sangan iron mine
processing plant site from three alternatives after
considering the eight criteria is the goal of this study.
4 Alternatives and criteria of processing
plant site selection

The first step in the process of site selection is col-


lecting and evaluating information. Exact considera-
tion suggested three alternatives for the processing
plant, which are shown in Fig. 4 as letters A, B and
C. The important criteria for site selection, which
were well thought out in this research, are summa-
rized in Table 3. Eight criteria were identified. Due to
the relative closeness of the alternative sites some of
Fig. 5 Hierarchical structure of the processing
the factors were considered common for all of them. plant site-selection
In the process of deciding on the selection of the
processing plant site each alternative is evaluated The next step is to establish a pairwise comparison
against the corresponding criteria based on technical matrix, which is set up by comparing pairs of criteria
and experimental experience and also by asking the or alternatives. For the Sangan iron ore mine project
decision makers and experts. the pairwise comparison matrix was established based
upon the judgment of experts using the nine point
scale shown in Table 1. Once the pairwise compari-
son matrices are formed the AHP is employed to de-
termine the criterion weights utilizing the eigenvector
method shown in Eq.(2).
The pairwise comparison matrix was established
using a nine-point scale, Table 4. Then the weight for
each criterion was determined by using the eigenvec-
tor method (see Column 10 of Table 4). Then alterna-
tives were compared based on the different criteria
and the eight matrices (the order of the matrices is
33). The weight of each alternative was then deter-
mined using the eigenvector method (see Table 5).
Fig. 4 Alternatives for the processing plant site,
letters A, B and C



694 Mining Science and Technology Vol.20 No.5

Table 4 Pairwise comparison matrix for processing Table 6 Calculated weights for each criterion for
plant site criteria each processing plant site
 C1 C2 C3 C4 C5 C6 C7 C8 Weight Weight A B C
C1 1 3 1/3 5 1 1/3 3 1/5 0.136 C1 0.136 0.111 0.111 0.778
C2 1/3 1 1 1 1 1 3 1/5 0.083 C2 0.083 0.111 0.111 0.778
C3 3 1 1 3 1 1 3 1 0.137 C3 0.137 0.079 0.225 0.696
C4 1/5 1 1/3 1 1/5 1/5 1 1/5 0.041 C4 0.041 0.111 0.111 0.778
C5 1 1 1 5 1 1 7 1 0.176 C5 0.176 0.143 0.143 0.714
C6 3 1 1 5 1 1 5 1 0.176 C6 0.176 0.079 0.225 0.696
C7 1/3 1/3 1/3 1 1/7 1/5 1 1/3 0.036 C7 0.036 0.143 0.143 0.714
C8 5 5 1 5 1 1 3 1 0.215 Overall priority 0.1078 0.1535 0.7387

Table 5 Comparisons of the alternatives with Table 7 Calculation of the consistency ratios
reference to C1~C8 Weight max I.I R.I.I I.R
A B C Weight Objective 1 8.9823 0.1403 1.41 0.0995
A 1 1 1/7 0.111 C1 0.136 3 0 0.58 0
C1 B 1 1 1/7 0.111 C2 0.083 3 0 0.58 0
C 7 7 1 0.778 C3 0.137 3.0968 0.0484 0.58 0.0834
A 1 1 1/7 0.111 C4 0.041 3 0 0.58 0
C2 B 1 1 1/7 0.111 C5 0.176 3 0 0.58 0
C 7 7 1 0.778 C6 0.176 3.0968 0.0484 0.58 0.0834
A 1 1/3 1/7 0.079 C7 0.036 3 0 0.58 0
C3 B 3 1 1/5 0.225 C8 0.215 3 0 0.58 0
C 7 5 1 0.696
A 1 1 1/7 0.111 According to this result the calculated inconsis-
C4 B 1 1 1/7 0.111 tency ratio is below 10% and the prepared selecting
C 7 7 1 0.778 matrices may be considered consistent.
A 1 1 1/5 0.143
C5 B 1 1 1/5 0.143
7 Conclusions
C 5 5 1 0.714
A 1 1/3 1/7 0.079
Selection of a mineral processing site is one of the
C6 B 3 1 1/5 0.225
most important factors affecting the whole process.
C 7 5 1 0.696
This problem involves taking into account many
A 1 1 1/5 0.143
C7 B 1 1 1/5 0.143
factors such as economical, geological, technical,
C 5 5 1 0.714
environmental and tectonic parameters. Therefore, it is
A 1 1 1/7 0.111 necessary to use a decision making method that
C8 B 1 1 1/7 0.111 considers multiple criteria when solving this problem.
C 7 7 1 0.778 In this paper the AHP, which is the most powerful and
flexible tool for use in solving multiple criteria
Once the component weights are calculated they problems, was applied. The problem was to select an
are synthesized to obtain the rank scores of each al- appropriate site for the mineral processing plant at the
ternative. The weights are synthesized from the high- Sangan iron ore mine (phase 1). For this purpose three
est level down by multiplying the weights by their alternatives, A, B and C, were evaluated with regard to
corresponding parent component from the level above eight criteria. After using the AHP method location C
and then adding them for each component within a was identified as the best site for locating the mineral
level according to the component it affects. The re- processing plant. In addition, a calculated inconsis-
sults for the processing plant site are tabulated in Ta- tency ratio of 0.0783 shows that the judgment matrices
ble 6 where it can be seen that alternative C has the are consistent and that the selection is acceptable.
highest weight. This means that the most suitable
processing plant site for the Sangan iron ore mine Acknowledgements
project is alternative C.
The authors would like to acknowledge the Sangan
6 Consistency ratios iron ore mine for kind cooperation in this research.

The judgments used in the process of deciding on References


the most suitable site were validated from the consis-
tency ratios. The results for are presented in Table 7.
[1] Venkata R R. Decision Making in the Manufacturing
Environment: Using Graph Theory and Fuzzy Multiple
SAFARI Mohsen et al Mineral processing plant location using the analytic hierarchy process 695

Attribute Decision Making Methods. London: Springer- [7] Saaty T L. The Analytic Hierarchy Process: Planning,
Verlag, 2007. Priority Setting, Resource Allocation. New York:
[2] Hwang C L, Yoon K. Multiple Attribute Decision Making. McGraw-Hill, 1980.
Berlin: Springer-Verlag, 1981. [8] Lee A H I, Chen W C, Chang C J. A fuzzy AHP and
[3] Saaty T L. Fundamentals of Decision Making and BSC approach for evaluating performance of IT depart-
Priority-Theory with the Analytic Hierarchy Process. ment in the manufacturing industry in Taiwan. Expert
Pittsburg: RWS Publications, 2000. Systems with Applications, 2008, 34(1): 96-107.
[4] Bottero M, Peila D. The use of the analytic hierarchy [9] Boroushaki S, Malczewski J. Implementing an extension
process for the comparison between micro-tunneling and of the analytical hierarchy process using ordered
trench excavation. Tunneling and Underground Space weighted averaging operators with fuzzy quantifiers in
Technology, 2005, 20(6): 501-513. ArcGIS. Computers & Geosciences, 2008, 34(4): 399-
[5] Samanta B, Sarkar B, Murherjee S K. Selection of 410.
opencast mining equipment by a multi-criteria decision- [10] Lurka A. Location of high seismic activity zones and
making process. Transactions of the Institute of Mining seismic hazard assessment in Zabrze Bielszowice coal
and Metallurgy, 2002, 111: A136-A142. mine using passive tomography. Journal of China Uni-
[6] Karadogan A, Bascetin A, Kahriman A, Gorgun S. A versity of Mining & Technology, 2008, 18(2): 177-181.
new approach in selection of underground mining [11] Lurka A, Swanson. Improvements in seismic event loca-
method. In: Proceedings of the International Conference: tions in a deep western U.S. coal mine using tomo-
Modern Management of Mine Producing, Geology and graphic velocity models and an evolutionary search al-
Environment Protection. Varna: SGEM, 2001. gorithm. Mining Science and Technology, 2009, 19(5):
599-603.




Vous aimerez peut-être aussi