Keynote Address — Strategic Planning and Scheduling
Strategic versus Tactical Approaches in Mining 2011, Perth, Australia
7
Figure 3 Production rate and head grade history at CSA Mine, (head grade in red squares – right
axis, production rate in blue diamonds – left axis)
Examination of the production history of nearly every mine shows a strong
relationship between mining
rate and head grade. Larger scale mining is less selective, but the grade problem is as much due to human
nature as it is to technology; if people are set unrealistic goals then “waste plus ore equals more ore”. The
author once reviewed a mine where the stope designs included discrete zones of hard, abrasive waste rock.
Upon
suggesting that a smaller, more selective stope design would be advantageous he was told, “We
looked at that but we couldn’t get the scheduled tonnes”. The mill throughput was later reduced and the
stopes redesigned, for a substantial improvement in mine NPV.
In general industry practice, the tonnage capacity of the process plant sets the rate. If the plant has been
constructed with surplus capacity or expanded to that point then great pressure is put on the mine to fill
the mill, often with scant regard for the effect on the quality of the material delivered.
5
A simplified optimisation model
If historical production data from a mine is analysed whereby:
grade is expressed as a percentage of the weighted mean
historical grade, and
annual production is expressed as a percentage of the average annual historical production.
The linear best-fit line will have as its slope a dimensionless constant k such that:
k = -ΔG(%) / Δt(%)
(1)
where:
G
=
realised head grade (in % metal).
t
=
the mining rate in tonnes per annum.
This is the value of k = 0.9 that defines the slope of the line in Figure 2. Values of k for selected Australian
non-gold mines are given in Table 1 and for gold mines in Table 2. Note that the R
2
value (or the Pearson
Coefficient of Determination) is an indicator of how well the line fits the data. The closer the R
2
value is to 1
the better the fit. The non-gold results are based on historical production data compiled by Mudd (2009).
The derived k values for non-gold mines cluster around a mode of 0.3 and 70% of them lie between 0.2 and
0.5. These k values should be used with caution because they incorporate the
influences of changes in
cutoff grade, planned dilution, unplanned dilution and in
some cases mining technology, as well as the
effects of changes in orebody characteristics over time. Nevertheless, they reflect valid choices of
operating
points for these orebodies and it is better to use these values than to assume that the mining rate has no
effect on head grade.
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1990
1991
1992
1993
1994
1995
1996
1997
1999
2000
2001
2002
2003
2004
2005
2006
2007
Dostları ilə paylaş: