Strategy Optimisation For Underground Mines
Strategy Optimisation For Underground Mines
Strategy Optimisation For Underground Mines
MininG
Strategy
optimisation for
underground mines
By ChrIS alFOrD, alFOrD MInInG SySTeMS, anD BrIan hall, aMC
COnSUlTanTS
J
ValMax Ideal improvement proposal:
eff Whittle pioneered these Change production rate &
Hill
developments, but today, many cutoff grade to maximise
software suppliers have similar value
tools. For underground mine design
(using stoping methods), no equivalent
tool has been available, and in that Increasing
sense, it is difficult to claim that designs Value
are optimal. Underground mining is
different, as mining is selective, and
while mining method and geotechnical
ground conditions may impose
constraints on the stope sequencing
options, there is greater flexibility
to optimise stope selection. AMIRA
Increasing
International coordinated the first Production
large-scale research project to identify Rate Increasing Cutoff Grade
an integrated and optimal approach to
mining method, cut-off grade, and 3 automated scenario generation and by-stope scheduling optimisation
production and development rates, computing execution controllers down to a time increment of one day
among others. 3 Hill of Value analysis: to visualise the can be optimised, with constraints on
The first three-year AMIRA research overall response of value to changes development, production and treatment,
project (2007–2010) was P884 ‘Planning in various decisions and external and with the objective to maximise net
and Rapid Integrated Mine Optimisation’ scenarios in order to facilitate present value.
(PRIMO). This was followed by two informed decision-making, rather The tool chain developed is being
additional AMIRA projects, where the than simply identifying a single applied to sponsor supplied case
research team was reduced to two optimum strategy. studies, and an extension to the current
providers, Alford Mining Systems (AMS) The goal in each stage is to produce research project is anticipated to
and AMC Consultants Ltd. The two designs in a rapid, automated and widen the capability of the framework
projects were P1037 ‘Optimisation of comparable format, with calibration for different case studies. A third
Stope Design and Stope Layouts’ and factors where the automated designs commercial release of the stope
P1043 ‘Rapid Automated Underground do not exactly match what might be optimisation engine with all the latest
Mine Optimisation Framework’. These produced with a fully engineered design. research developments is planned for
projects were complementary and A research user interface has been 12–18 months, and the remaining
built on the methodology identified developed so that the research deliverable research is expected to be of sufficient
in PRIMO, but with the goal of an can be accessed by the sponsors. This generality for a product release of the
integrated tool chain built on a common would be replaced by a commercial user whole tool chain to be considered in
software architecture. Sponsors agreed interface after the research deliverable the next two to three years. In the
that the acronym SIRUS (Latin for an goes through a product development meantime, the sponsors will be able to
‘underground granary’) be applied to the phase where additional functionality use the research deliverable with the
new project. can be added, and extensive testing and research interface.
The projects have a somewhat validation completed.
unique structure, as the goal of the One of the key objectives for the The projects are sponsored by key mining
research projects is to deliver the P1043 project was to explore new software companies Cae Datamine, Maptek,
research as a commercially supported mathematical optimisation techniques GijimaaST/MinerP and Deswik and mining
product. Too many research projects for sequence optimisation. Methods companies newmont, BhPB, MMG, Barrick,
produce a piece of research software like mixed integer programming (MIP) vale, Xstrata/Glencore and Boliden.
on a CD, and the value is lost as the and genetic algorithms (GA) have
industry software providers are unable been available for decades and were Further information can be obtained from
to deploy the results. tested in the PRIMO project, but newer Chris du Plessis, Program Manager, aMIra
The P1037/P1043 projects have hybrid optimisation techniques have International.
developed optimisation engines for been applied in the SIRUS project.
each stage of the strategy optimisation Preliminary results indicate that stope-
tool chain:
3 stope shapes – producing the optimal
set of stope shapes, given a cut-off,
stope geometry and geotechnical
constraints
3 stope layout – aggregating shapes
into larger mining blocks and panels,
removing outliers, and, in some
cases, choosing different stoping
methods depending on ore width
3 development layouts – generating
level layouts that ‘make the mining
method work’, based on engineering
design rules
3 access – adding ore passes, declines,
shafts, et cetera, for access and ore
transport to ‘make the mine work’
3 stope, development and access
sequence optimisation for the life of
the mine