Determination of optimum cutoff grades

of multiple metal deposits by using the

Golden Section search method
by M. Ataei and M. Osanloo*

risk of financial failure. Maximization of NPV

also means that the capital invested is being
Synopsis used most efficiently. In addition, cash taken
today is more certain than cash promised next
In recent years several mine researchers have studied the year.
optimization of cutoff grade with the purpose of maximizing Net The basic algorithm to determine the cutoff
Present Value (NPV). The best and most popular one is the Lane
grades which maximize the NPV of an
algorithm. Based upon the Lane algorithm it is possible to calculate
operation subject to mining, milling, and
the optimum cutoff grade for mine, mill, and refinery capacities for
single metal deposits with variable grades. The Lane algorithm
refining capacities was proposed by Lane
cannot be used in multiple metal deposits. The reason is that, while (Lane 1964; 1988). His theory takes into
in single metal deposits six points are possible candidates for the account the costs and capacities associated
optimum cutoff grade, in multiple metal deposits an infinite number with these stages. Mine capacity is the
points are possible candidates for the optimum cutoff grades and maximum rate of mining the deposit, mill
objective function evaluation of these infinite points is impossible. capacity is the maximum rate of processing
In this study the Golden Section search method was used to ore, and refinery capacity is the maximum rate
calculate optimum cutoff grades of multiple metal deposits for the of production of final product. The determi-
same conditions as Lane assumed in single metal deposits. Based on nation of cutoff grade is based on the fact that
this method, grade-tonnage distribution uncertainty space of either one of those stages will alone limit the
problem is guessed. In the next step, by selecting test points in total capacity of operation or a pair of stages
uncertainty space and evaluating the objective function at these
may limit the entire operation. The theory also
points, a part of uncertainty space will be eliminated. The procedure
takes into account the grade distribution of the
is then repeated until the uncertainty space has been sufficiently
reduced to ensure that the optimum cutoff grades have been located
deposit and the opportunity cost (time costs)
to the required accuracy. of mining low grade ore while high grade ore
Key words: Optimum cutoff grades, Optimization, Net Present is still available in the deposit. The optimum
Value (NPV), Golden Section search method cutoff grades theory introduced by Lane
determines the cutoff grades year by year. This
procedure depends upon the graphical
approach for the determination of optimum
cutoff grades. However, if only one of the
capacities is limited, then analytical determi-
Introduction nation is also possible.
Deposits containing more than one metal
In open pit mines, cutoff grade is used to
are usually dealt with by converting all metals
discriminate ore from waste. It is one of the
to their equivalent in terms of one basic metal,
most crucial decisions that must be faced by
and aggregating several values (Liimatainen
mine engineers. If the mineral grade is equal to
1998; Zhang 1998). For example, lead and
or above cutoff grade, the material is classified
zinc often occur together. Assuming zinc to
as ore and if the grade of mineral is less than
have twice the value of lead, the lead content
the cutoff grade, the material is classified as
can be divided by two and added to the zinc
waste. Depending upon the mining method,
content in order to obtain total zinc content.
waste is either left in situ or sent to the waste
Then any analysis can be conducted exactly as
dumps, whereas ore is sent to the treatment
if mineralization consists of a single metal.
plant for further processing and eventual sale
(Taylor 1985).
There are many theories for the determi-
* Department of Mining, Metallurgy and Petroleum
nation of cutoff grades. But most of the
Engineering, Amirkabir, University of Technology,
research that has been done in the last three Tehran Polytechnic, Iran.
decades shows that determination of cutoff The South African Institute of Mining and
grades with the objective of maximizing net Metallurgy, 2003. SA ISSN 0038223X/3.00 +
present value (NPV) is the most acceptable 0.00. Paper received Aug. 2002; revised paper
method. Maximizating NPV helps to reduce the received Jun. 2003.

The Journal of The South African Institute of Mining and Metallurgy OCTOBER 2003 493
Determination of optimum cutoff grades of multiple metal deposits
The equivalent grades method has some flaws, but if
minerals have fairly stable values, this procedure is valid and
P = (s1 r1 )Qr1 + (s2 r2 )Qr 2 mQm cQc fT [1]
simplifies the problem. If one of the metals is subject to
Where m: mining cost ($/tonne of material mined), c: concen-
market limitation, this technique becomes invalid, because,
trating cost ($/tonne of material concentrated), r1: refinery
the production in excess of the contracts for that metal cannot
be sold and therefore ore cannot be valued on the basis of cost ($/unit of product 1), r2: refinery cost ( $/unit of
contract price. Therefore, it is the influence of capacities in product 2), f: fixed cost, s1: selling price ($/unit of product 1),
both plant and market which invalidate the combined value s2: selling price ($/unit of product 2), T: the length of the
criterion (Lane 1988). This method also has operational and production period being considered, Qm: quantity of material
economical flaws (Barid and Satchwell 2001). to be mined, Qc: quantity of ore sent to the concentrator, Qr1:
Other methods for discrimination of ore/waste in multiple the amount of product 1 actually produced over this
metal deposits are: Critical level method, Net Smelter Revenue production period, Qr2: the amount of product 2 actually
(NSR) method, Single grade cutoff approach, Dollar value produced over this production period.
cutoff approach (Annels 1991;Barid and Satchwell 2001). If d is discount rate, the difference v between the present
None of these methods is an optimized technique, because values of the remaining reserves at times t=0 and t=T is
the distribution of grade of mined material, mining operation (Hustrulid and Kuchta 1995):
capacity constraints, and the effect of time on money value
v = P VdT [2]
are not considered. These methods usually lead to sub-
optimal exploitation of the resource. In reality, there are Where V is the present values at time t =0. Substituting
potentially operational, economic and logic serious issues in Equation [1] into Equation [2] yields:
the application of these methods. Therefore determination of
optimum cutoff grades of multiple metal cutoff grade decision v = (s1 r1 )Qr1 + (s2 r2 )Qr 2 [3]
making is formidable. In this paper, selecting cutoff grades
with the purpose of maximizing NPV subject to the mQm cQc ( f + Vd )T
constraints of mining, concentrating, and refining capacities
The quantities refined Qr1 and Qr2 are related to that sent
of two metals will be discussed.
by the mine for concentration Qc by:

Objective function Qr1 = g1 y1Qc [4]

For an operating mine, there are typically three stages of Qr 2 = g2 y2Qc [5]
production: (i) the mining stage, where units of various
grade are extracted up to some capacity, (ii) the treatment Where: g 1 is the average grade of metal 1 sent for concen-

stage, where ore is milled and concentrated, again up to some tration and g 2 is the average grade of metal 2 sent for
capacity constraint, and (iii) the refining stage, where the concentration.
concentrate is smelted and/or refined to a final product which Substituting Equations [4] and [5] into Equation [3]
is shipped and sold. The latest stage is also subject to yields:
capacity constraints. For simplicity, assume a two metal
deposit. In this deposit, ore is sent to a concentrator and the [
= (s1 r1 )g1 y1 + (s2 r2 )g2 y2 c ] [6]
concentrator will produce two concentrates. Each concentrate Qc = m Qm ( f + Vd )T
for smelting and finally refining is sent to a refinery plant.
Each stage has its own associated costs and a limiting One would now like to schedule the mining in such a way
capacity. The operation as a whole will incur continuing fixed that the decline in remaining present value takes place as
costs. (See Figure 1). rapidly as possible. This is because later profits get
By considering costs and revenues in this operation, the discounted more than those captured earlier. In examining
profit is determined by using following equation: Equation [6], this means that v should be maximized.

Cost Symbol Capacity Symbol

m M Mine

c Concentrator
C

r1 R1 Metal 1 Concentrate 1
Refinery 1

r2 Metal 2 Concentrate 2
R2 Refinery 2

Figure 1The flowchart of the mining operation in a two metal deposit

494 OCTOBER 2003 The Journal of The South African Institute of Mining and Metallurgy
Determination of optimum cutoff grades of multiple metal deposits
Equation [6] is the fundamental formula and all the
f + Vd
cutoff grade optimum can be developed from it. The time s1 r1
taken T is related to the constraint capacity. Four cases arise vr1 = R1 [13]
depending upon which of the four capacities are actually g y + (s r ) g y c
limiting factors. 1 1 2 2 2 2
If the mining rate is the limiting factor then the time T is Qc m Qm
given by: (s1 r1 ) g1 y1 +
Q
T= m [7] vr 2 = f + Vd [14]
s2 r2 R g2 y2 c
M
If the concentrator rate is the limiting factor then the time 2
T is controlled by the concentrator: Qc m Qm
Qc Now, for any pair of cutoff grades, it is possible to
T= [8] calculate the corresponding Vm, Vc, Vr1, and Vr2. The
C controlling capacity is always the one corresponding to the
If the refinery output of metal 1 is the limiting factor then least of these four equations. Therefore:
the time T is controlled by the refinery of metal 1:
Q gyQ
[
max ve = max min(vm , vc , vr1 , vr 2 ) ] [15]

T = r1 = 1 1 c [9] In Equations [11] to [14], V is an unknown value

R1 R1 because it depends upon the cutoff grades. Since the
If the refinery output of metal 2 is the limiting factor then unknown V appears in these Equations an iterative process
the time T is controlled by the refinery of metal 2: must be used. Equations [11] to [14] are known two
dimensional and instead of a curve in one metal deposit , Vm,
Qr 2 g2 y2Qc
T= = [10] Vc, Vr1, Vr2 and Ve are surface and may be represented in a
R2 R2 series of contours . Figure 2 shows Ve for a two metal
Substituting Equations [7], [8], [9] and [10] into deposit.
If only one capacity is dominant, the limiting economic
Equation [6] yields:
maximization may be accomplished analytically. To find the
[
vm = (s1 r1 ) g1 y1 + (s2 r2 )g2 y2 c ] [11]
grades which maximize the NPV under different constraints,
one first takes the derivative of Equations [11] to [14] with
f + Vd
Qc m + Qm respect to g1 and g2. In the next step, setting derivatives of
M Equations [11] to [14] equal zero. It will obtain four line
equations; the general form of these lines is:
(s1 r1 ) g1 y1 + (s2 r2 ) g1 g2
vc = + =1 [16]
g y c + f + Vd [12] p q
2 2 C Where the p and q values of Equation [16] for different case
Qc mQm are given in Table I. Equation [16] is useful when a single

0.20

0.15
molybdenum (%)

0.10

0.05

0.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0
Copper (%)

Figure 2Ve for a two metal deposit

The Journal of The South African Institute of Mining and Metallurgy OCTOBER 2003 495
Determination of optimum cutoff grades of multiple metal deposits

Table I
guessed. In the next step, by selecting test points in the
uncertainty space and evaluating and comparing the objective
p and q values of Equation [16] for different cases function at these test points, a part of the uncertainty space
Limiting capacity p q will be eliminated. This reducing procedure is repeated until
the uncertainty interval in each direction is less than a small
Mine c c specified positive value , where is the desirable accuracy
(s1r1) y1 (s2r2) y2
for determining optimum cutoff grades (Rardin 1998).
Concentrator f + Vd f + Vd The ratio of the remaining length, after the elimination
c+ c+
C C
(s1r1) y1 (s2r2) y2 process, to the initial length in each dimension is called the
reduction ratio. The dichotomous search method, the
Refinery 1 c c
Fibonacci search method, and the Golden Section search
(s1r1 f + Vd) y1 (s2r2) y2
R1 method are examples of elimination methods. Among these
Refinery 2 c c methods, the reduction ratio of the Golden Section search
(s1r1) y1
(s2r2 f + Vd) y2 method is optimum and equals 0.618 (this number is called
R2 the golden number). In this method, the ratio of eliminated
length to initial length will be equal to 0.382. In addition,
using the Golden Section rule means that every stage of the
uncertainty range reduction (except the first one), the
process is limiting; however, when the maximum occurs at a objective function need only be evaluated at one new point
balancing point, where more than one capacity restricts (Chong and Zak 1996; Rao 1996; Bazarra, Hanif and Shetty
throughput, no satisfactory analytical technique has been 1993).
developed. The problem geometrically is one of four Figure 3 shows the Golden Section search method for a
intersecting surfaces forming hills. The peaks are compara- one dimensional function. In the first step, assume (L, U) to
tively easy to locate but the ridges and valleys where they be the initial interval of uncertainty and note that the initial
intersect are more difficult (Lane 1988). Infinite points are interval includes the optimum point. Then select two test
possible candidates for the optimum cutoff grades. For this points, g1 and g2 (Figure 3.a). The locations of these points
reason, the maximum is best located by a search process. The are:
Golden Section search technique to calculate the optimum
cutoff grades of multiple metal deposits has been found quite
g1 = L + (U L) x 0.382 [17]
effective.
g2 = L + (U L) x 0.618 [18]
Calculation of optimum cutoff grades In the next step, the objective function will be evaluated
One of the fastest methods to calculate the optimum point of in the g1 and g2 points. Depending on the objective function
unimodal functions is the elimination method. In the first value of these points, the length of the new interval of
step of this method the uncertainty space of the problem is uncertainty is successively reduced in each iteration

g1 = L + (U L) x 0.382
g2 = L + (U L) x 0.618
(3.a)

L g1 g2 U
f (g1) < f (g2)
(3.b)

L g1 g2 U

f (g1) > f (g2)

(3.c)
Figure 3Golden Section search method for one dimensional function

496 OCTOBER 2003 The Journal of The South African Institute of Mining and Metallurgy
Determination of optimum cutoff grades of multiple metal deposits
(Figures 3.b and 3.c). The process is then repeated by placing Example
a new observation. Repeating this in the new range will find Consider a hypothetical condition that a final pit limit has
the optimum point with desirable accuracy. been superimposed on a mineral inventory. The pit outline
This procedure can be extended for multivariable contains 15 million tonnes of material. The gradetonnage
problems. Applying the Golden Section search method for distribution and average grade of ore for each metal are
unimodal two-dimensional functions, in the first step the shown in Tables II, III and IV. The associated costs, prices,
initial interval of uncertainty for each variable must be capacities, quantities and recoveries are demonstrated in
determined. Assume (L1, U1) to be the initial interval of Table V.
uncertainty of variable 1 and (L2, U2) to be the initial interval
of uncertainty of variable 2. Then select four test points (A,
(L1, U2) (U1, U2)
B, C, D) in the uncertainty space (Figure 4).
The locations of these points are:
a1 = L1 + (U1 L1 ) x 0.382 [19]
b2
a2 = L1 + (U1 L1 ) x 0.618 [20] C D
b1 = L2 + (U2 L2 ) x 0.382 [21]
b1
b2 = L2 + (U2 L2 ) x 0.618 [22] A B
In the next step, calculate the amount of the objective
function for each four test points. By comparing the objective
function values of these points, the optimum point in this (L1, L2) a1 a2 (U1, L2)
iteration and a new space of uncertainty is determined:
If point A is optimum then U1 = a2 and U2 = b2 and a Figure 4Golden Section search method for one-dimensional function
part of the uncertainty space is eliminated.
The remaining space is shown in Figure 5.a.
If point B is optimum then L1 = a1 and U2 = b2 and a Table II
part of the uncertainty space is eliminated. Grade-tonnage distribution of copper and
The remaining space is shown in Figure 5.b. molybdenum
If point C is optimum then U1 = a2 and L2 = b1 and a
part of the uncertainty space is eliminated. Molybdenum (%)
The remaining space is shown in Figure 5.c. Copper (%)
00.05 0.050.1 0.10.15 0.150.2 >0.2
If point D is optimum then L1 = a1 and L2 = b1 and a
00.1 1400000 900000 285000 315000 510000
part of the uncertainty space is eliminated. 0.10.2 400000 300000 250000 135000 60000
The remaining space is shown in Figure 5.d. 0.20.3 800000 530000 300000 210000 30000
0.30.4 1500000 570000 375000 135000 60000
In the remaining space for finding the optimum point, 0.40.5 410000 255000 75000 60000 60000
only one test point is left. In the next step, three new test 0.50.6 510000 300000 210000 105000 110000
points must be selected. This operation is repeated until the 0.60.7 375000 270000 210000 90000 90000
> 0.7 645000 690000 570000 500000 400000
optimum point is found with desirable accuracy (Kim 1997).

(L1, U2) (U1, U2) (L1, U2) (U1, U2)

b1 b1
C D C D
b2 b2
A B A B

(L1, L2) a1 a2 (U1, L2) (L1, L2) a1 a2 (U1, L2)

(5.a) (5.b)

(L1, U2) (U1, U2) (L1, U2) (U1, U2)

b1 b1
C D C D
b2 b2
A B A B

(L1, L2) a1 a2 (U1, L2) (L1, L2) a1 a2 (U1, L2)

(5.c) (5.d)

Figure 5Elimination of a part of the uncertainty space by the Golden Section search method

The Journal of The South African Institute of Mining and Metallurgy OCTOBER 2003 497
Determination of optimum cutoff grades of multiple metal deposits

Table III b1 = 0 + (0.2 0) x 0.382 = 0.0764

Average grade of copper for different copper and b2 = 0 + (0.2 0) x 0.618 = 0.1236
molybdenum intervals
Thus the four test points in the first iteration are:
Copper (%) Molybdenum (%) (0.2674, 0.0764), (0.2674, 0.1236),
0 0.05 0.050.1 0.10.15 0.150.2 >0.2
(0.4326, 0.0764), (0.4326, 0.1236)
00.1 0.02 0.03 0.02 0.03 0.05
In the next step, calculate the amount of ore, the amount
0.10.2 0.12 0.17 0.16 0.19 0.14
0.20.3 0.25 0.27 0.25 0.22 0.26 of waste, the average grade of ore for each metal, the amount
0.30.4 0.33 0.32 0.35 0.34 0.37 of total mined material, the amount of mined material must
0.40.5 0.44 0.47 0.45 0.48 0.46
0.50.6 0.53 0.55 0.57 0.54 0.55
be sent to the concentrator, the amount of metals product of
0.60.7 0.67 0.63 0.65 0.64 0.66 refinery 1 and 2 and calculate the amount of Vm, Vc, Vr1, Vr2
> 0.7 0.98 1.04 1.02 1.09 1.01 and Ve (objective function) for each four test points. Table VI
shows the result of these calculations.
Among the four test points, point (0.2674, 0.0764) is the
optimum point. Therefore the boundaries of the new search
Table IV
space are:
Average grade of molybdenum for different copper
and molybdenum intervals L1 = 0 U1 = 0.4326 L2 = 0 U2 = 0.1236
By repeating this process, optimum cutoff grades can be
Copper (%) Molybdenum (%)
found with desirable accuracy. In this example it is assumed
00.05 0.050.1 0.10.15 0.150.2 >0.2
that a cutoff grade with accuracy of 0.001% is desired. For
0-0.1 0.004 0.052 0.104 0.152 0.216 finding cutoff grades with 0.001 per cent accuracy, the
0.1-0.2 0.034 0.064 0.120 0.190 0.228
0.2-0.3 0.022 0.056 0.132 0.182 0.254
operation is repeated. Table VII shows the result of repeat
0.3-0.4 0.062 0.084 0.116 0.188 0.276 operations for the first year of mine life. Optimum cutoff
0.4-0.5 0.012 0.070 0.108 0.182 0.238 grades are obtained after 15 iterations (4 + 14 x 3 = 46 test
0.5-0.6 0.024 0.078 0.140 0.164 0.304
0.6-0.7 0.028 0.058 0.124 0.170 0.240 points). According to these calculations, for the first year of
> 0.7 0.018 0.076 0.126 0.172 0.256 mine life optimum the cutoff grades of copper will be
0.3441% and the optimum cutoff grade of molybdenum will
be 0.0254%.
After doing calculations for the first year of mine life, the
Table V grade tonnage curve of the deposit must be adjusted. To do
Economic parameters this, ore tonnes in the first year of mine life from the grade
distribution intervals above optimum cutoff grades and waste
Parameter Unit Quantity
tonnes in the first year of mine life from the grade distri-
Mine capacity Tons per year 2500000 bution intervals below optimum cutoff grades should be
Mill capacity Tons per year 1000000 subtracted. These calculations will be repeated until the end
Refining capacity (copper) Tons per year 6000
Refining capacity (molybdenum) Tons per year 1000
of mine life. The output obtained in these calculations gives
Mining cost Dollars per tonne 1 the cutoff grade policy and the production schedule as shown
Milling cost Dollars per tonne 3.5 in Table III.
Refining cost (copper) Dollars per tonne 88.5
Refining cost (molybdenum) Dollars per tonne 254
Fixed costs Dollars per year 790000 Conclusion
Price (copper) Dollars per tonne 1700
Price (molybdenum) Dollars per tonne 6700 One of the important parameters of open-pit mine design is
Recovery (copper) % 82 determination of cutoff grade. The cutoff grade is used to find
Recovery (molybdenum) % 82 the destination of material to be mined. In multiple metal
Discount rate % 20
deposits none of the common methods such as critical level,
equivalent grade and net smelter revenue methods is an
optimized technique, because in these methods the distri-
According to Figure 1, the objective function of this
bution of grade of mined material, mining operation capacity
problem (ve) is unimodal and the Golden Section search
and the effect of time on money value are not considered. The
method can be used.
Based up on gradetonnage distribution (Table II), the
initial uncertainty interval for copper cutoff grade is 0,0.7
Table VI
and initial uncertainty interval for molybdenum cutoff grade
is 0,0.2 therefore: Objective function value (Ve) in four test points at
L1 = 0 U1 = 0.7 L2 = 0 U2 = 0.2 first iteration

Considering Equations [19] to [22], the possible space Test point Objective function value

for optimum point yield: (0.2674, 0.0764) 13 400 582

a1 = 0 + (0.7 0) x 0.382 = 0.2674

(0.2674, 0.1236) 7 263 290
(0.4326, 0.0764) 10 043 473
a2 = 0 + (0.7 0) x 0.618 = 0.4326 (0.4326, 0.1236) 5 176 957

498 OCTOBER 2003 The Journal of The South African Institute of Mining and Metallurgy
Determination of optimum cutoff grades of multiple metal deposits
Table VII methodology outlined in this paper is Golden Section search
Result of repeat operations for first year of mine life method. This method provides a fast procedure to determine
the optimum cutoff grades for multiple metal deposits. For
Iteration Cutoff grades Objective function (ve)
Copper (%) Molybdenum (%)
this purpose hypothetical data of one Cu/Mo ore deposit was
used to find the optimum cutoff grades and maximize the
1 0.2674 0.0764 13400582 present value. The total deposit is assumed to be 15 million
0.2674 0.1236 7263290
0.4326 0.0764 10043473 tonnes. Based on the Golden Section method and grade-
0.4326 0.1236 5176957 tonnage distribution, the uncertainty space of problem was
2 0.1652 0.0472 15609347 found. By selecting test points in the uncertainty space and
0.1652 0.0764 14557106
0.2674 0.0472 16886550 calculating the amount of ore, amount of waste, average
0.2674 0.0764 13400582
grade of ore for each metal, amount of total mined material,
3 0.2674 0.0292 16336005
0.2674 0.0472 16886550 amount of mined material that must be sent to the concen-
0.3305 0.0292 18010784 trator, amount of metals product of refinery 1 and 2, amount
0.3305 0.0472 15404960
of Vm, Vc, Vr1, Vr2 and Ve (objective function) for each four
4 0.3305 0.0180 16949965
0.3305 0.0292 18010784 test points were determined. Based on the result of the
0.3695 0.0180 17467282 objective function at these points, a part of the uncertainty
0.3695 0.0292 16555937
5 0.3064 0.0292 17299971 space can be eliminated. These operations were continued
0.3064 0.0361 17904442 until the optimum cutoff grades (0.001%) were found with
0.3305 0.0292 18010784
0.3305 0.0361 17167138 high desirable accuracy .
6 0.3305 0.0249 17637253
0.3305 0.0292 18010784 Reference
0.3454 0.0249 18026841
0.3454 0.0292 17557996 ANNELS, A.E. Mineral Deposit Evaluationa partial approach, Chapman &
7 0.3454 0.0223 17838465 Hall, London, 1991. pp. 114117.
0.3454 0.0249 18026841
BARID, B.K. and SATCHWELL, P.C. Application of economic parameters and
0.3546 0.0223 17816747
0.3546 0.0249 17679804 cutoffs during and after pit optimization, Mining Engineering, February
8 0.3397 0.0249 17905342 2001, pp. 3340.
0.3397 0.0265 18041904 BAZARRA, M.S., HANIF, D.S., and SHETTY, C.M. No linear programming theory
0.3454 0.0249 18026841 and algorithms, John Wiley & Sons, Inc., New York. 1993.
0.3454 0.0265 17875922
9 0.3362 0.0265 17946174 CHONG, E.K.P. and ZAK, S.H. An introduction to optimization, A Wiley-
0.3362 0.0276 18030117 Interscience Publication, John Wiley & Sons, Inc., New York, 1996.
0.3397 0.0265 18041904 p. 409.
0.3397 0.0276 17978534
HUSTRULID, W. and KUCHTA, M. Open-pit mine planning and design, vol. 1,
10 0.3397 0.0259 17990464
0.3397 0.0265 18041904 A.A.Balkema, Rotterdam, 1995. pp. 512544.
0.3419 0.0259 18049111 KIM J. Iterated grid search algorithm on unimodal criteria, Ph. D. Dissertation
0.3419 0.0265 18015664 in statistics, Blacksburg, Virginia, 1997. p. 115.
11 0.3419 0.0255 18017498
0.3419 0.0259 18049111 LANE, K.F. Choosing the optimum cut-off grade, Quarterly of the Colorado
0.3432 0.0255 18053530 School of Mines Quarterly, 1964. vol. 59 (4), pp. 811829.
0.3432 0.0259 18036136 LANE, K.F. The economic definition of orecut off grades in theory and practice,
12 0.3432 0.0253 18034062 Mining Journal Books Limited, London, 1988. p. 145.
0.3432 0.0255 18053530
0.3441 0.0253 18056247 LIIMATAINEN, J. Valuation model and equivalence factors for base metal ores,
0.3441 0.0255 18047846 Proceeding of mine planning and equipment selection (ed.) Singhal J.,
13 0.3441 0.0252 18044244 A.A.Balkema, Rotterdam, 1998. pp. 317322.
0.3441 0.0253 18056247
RAO, S.S. Engineering optimization (Theory and Practice), Third edition, A
0.3446 0.0252 1804554
0.3446 0.0253 18052363 Wiley-Interscience Publication , John Wiley & Sons, Inc., New York, 1996.
14 0.3437 0.0253 18047794 p. 903.
0.3437 0.0254 18055211 Rardin R. L. 1998. Optimization in operations research, PrenticeHall
0.3441 0.0253 18056247
International, Inc., p. 919.
0.3441 0.0254 18061286
15 0.3441 0.0254 18061286 TAYLOR, H.K. Cutoff gradessome further reflections, Trans. Inst. Min. Metall
0.3441 0.0255 18058403 (Sect. A: Min. industry), 1985. pp. A204A216
0.3442 0.0254 18057018 ZHANG, S. Multimetal recoverable reserve estimation and its impact on the cove
0.3442 0.0255 18050509
ultimate pit design, Mining Engineering, July, 1998. pp. 7377.

Table VIII
Optimum cutoff grades for different years of mine life
Year cutoff grade Qm Qc Qr1 Qr2 Profit NPV
Copper (%) Molybdenum (%) (tonne) (tonne) (tonne) (tonne) ($) ($)

1 0.3441 0.0254 2500000 1000000 5979 953 8990988 31058889

2 0.3382 0.0257 2462121 1000000 5945 948 8937434 28279583
3 0.3256 0.0218 2282898 1000000 5776 919 8657884 24998065
4 0.2683 0.0218 1986094 1000000 5324 900 8104657 21339794
5 0.2450 0.0172 1814563 1000000 5132 863 7729465 17503096
6 0.1242 0.0162 1523546 1000000 4577 838 6966054 13274251
7 0.0917 0.0104 1378169 1000000 4358 800 6512926 8963047
8 0.0714 0.0034 1052609 857348 3520 644 5091277 4242731

The Journal of The South African Institute of Mining and Metallurgy OCTOBER 2003 499
 Sasol director appointed chairperson to the
National Science and Technology Forum*
The NSTF Executive Committee and its stakeholders are The NSTF welcomes Marriott as the chairperson of the
proud to announce the appointment of John Marriott as the NSTF and looks forward to his expertise in issues of science,
new chairperson of the NSTF from 1 June 2003. Marriott engineering and technology to ensure continued growth of
succeeds Dr S.J. Lennon, who successfully served as the discipline in South Africa. Marriott stated that he was
Chairperson of the NSTF for the past three years. honoured by the appointment and looked forward to being
Marriott, is currently a director of Sasol Technology and able to contribute to the activities of the NSTF.
is also the general manager of Sasol Ltd. A chemical The NSTF wishes both Lennon and Marriott success in
engineer by training, he has spent several years at the their new challenges in uplifting the economic growth
highest level in the corporate world and has simultaneously through science, engineering and technology.
maintained an outstanding reputation in the technical
world. He has forged close associations with higher
education institutions in South Africa, where his
management skills and technical expertise helped provide * Issued by: Office of the Chief Executive Officer, National
marked insights into alliances between education and Science and Technology Forum (NSTF)
industry to ensure the provision of technical and scientific For more information contact:
skills. Mrs Wilna Eksteen on 082 442 4983.

Thuthuka Project Managers commissions its first project

with Outokumpu*
The recent commissioning of the cleaner flotation plant at Outokumpu as the main process equipment supplier, sub-
Botswanas BCL Mine (Bamangwato Concessions Limited) contracting to Thuthuka.
near Selebi Phikwe marked several firsts for Thuthuka As the main turnkey contractor on the project, Thuthuka
project managers. Project Managers was responsible for the civil, structural
Last year, Thuthuka was granted the tender for the steel, piping, electrical and sections of the instrumentation
upgrade of the cleaner flotation plant at the concentrator contracts. As the project manager, Bezuidenhout carried
plant. BCL is a copper/nickel-mining complex and one of overall responsibility for the projects execution. Other
Botswanas largest private employers. responsibilities included engineering and procurement,
Gerhard Bezuidenhout was the Thuthuka project document control, progress control, scheduling, cost control,
manager in charge of the upgrade to BLCs cleaner flotation quality assurance and control, and reporting and liaising
plant. The full scope of the project involved the supply and with BLC project personnel. It should be noted that the
installation of the cleaner flotation plant as part of a larger assistance of the BLC project personnel was of a high quality
upgrade of the concentrator plant as a whole. The overall and contributed to the success of the project, adds
aim of the upgrade was to reduce spillage and increase Bezuidenhout.
economic mineral recovery within the concentrator plant. Bezuidenhout concludes, The BCL flotation plant
Bezuidenhout says, This project was a very exciting one represents the first association between Thuthuka and
for Thuthuka for several reasons. Firstly, this is the first Outokumpu, and were certain that the success achieved will
metallurgical processing plant of this nature that Thuthuka result in future joint ventures.
has managed as a turnkey project and secondly, this is
Thuthukas first venture with Outokumpu and were very
pleased to have established a good working relationship
with one of South Africas leading suppliers of flotation
* Issued by: Alison Job, V Squared Marketing,
equipment.
Tel: (011) 678 2227,
Outokumpu had previously supplied flotation cells to
E-mail: [email protected]
Tati Nickel, a Lin Ore and Botswanan company that owns
BCL, so it was the natural extension of an existing On behalf of: Bill Pullen,
relationship for them to partner us on this project. Thuthuka Project Managers,
Thuthuka Project Managers won the tender to project Tel: (011) 315 7376,
manage the upgrade, while BCL Limited selected E-mail: [email protected]

500 OCTOBER 2003 The Journal of The South African Institute of Mining and Metallurgy