Maths professor set to shake up open cut mine design

A MELBOURNE maths professor who is already established as a world leader in developing tools for the mining industry has an idea that could revolutionise the design of open cut mines. So far it has shown net present value increases of between 5% and 10%.
Maths professor set to shake up open cut mine design Maths professor set to shake up open cut mine design Maths professor set to shake up open cut mine design Maths professor set to shake up open cut mine design Maths professor set to shake up open cut mine design

A Melbourne professor has come up with an algorithm that could boost the NPV of open pit mines.

David Upton

University of Melbourne Professor Hyam Rubinstein will visit Perth at the end of August to seek $1.86 million from sponsors to develop and commercialise the idea, based on an algorithm he developed with his PhD student Juan Yarmuch.

Development of the mine design tool is being led by AMIRA International, which has a global reputation for successfully bringing together academia and industry on collaborative research projects.

Trials of the tool on existing mines have already shown it increased the NPV of open cut mines by between 5% and 10%, depending on the size, duration and complexity of the mine.

Rubinstein told Australia's Mining Monthly the algorithm was all about optimising the design of pushbacks - a volume that can be mined in a period between two and four years, within the ultimate pit shell.

"The existing software produces an end-state design, which might optimise theoretical NPV, but the designs are not practical," he said.

"They produce an idea, but an experienced mine planner has to look at what's produced and tinker with it over two to three weeks to produce a proper design.

"Practical pushbacks need to be connected, have the right bench width for access of mining equipment and have the right volume for the mill production capacity.

"What we have is an algorithm that will bypass a lot of that extra work and in just two to three hours produce a better design that is very, very close to the final design.

"It's really quite a major step forward. All of the improvement in NPV goes straight to the bottom line. And the more complicated and longer a mine is running, the greater the value created."

Rubinstein has already had success applying maths to solve problems for the mining industry.

In 2008, he received backing under an AMIRA-managed project to commercialise mathematical tools that optimised underground mine design.

Known as DOT and PUNO, the software was developed by MineOptima and licensed to Maptek and Deswik. MineOptima was bought last year by RPM Global.

The algorithm for open cut design has been patented by the University of Melbourne and licensed to a special-purpose company, Thinking Mine Design, to develop and commercialise the software.

Rubinstein is project director and Juan Yarmuch, a mining engineer who has worked on large open cut mines in Chile, is responsible for technical development and management of software development resources.

The development team also includes David Niall, an experienced project manager for software and technology products.

The AMIRA-run project is seeking $1.86 million from sponsors for Phase 1 over a period of about 18 months.

Phase 1 will focus on scaling the software to work efficiently on very large mines and the production of a commercial version of the software for early release to sponsors.

Rubinstein said developing the project under AMIRA helped open doors in the mining industry and could really benefit the project as well as the mining industry.

"If we can work with a range of mining companies and see different designs, try out our tools and sharpen them up in cooperation with the leading mining companies, the development will be much faster and we will end up with a much better product," he said.

Rubinstein said the algorithm was based on a branch of maths known as linear programming, which was developed in the US at the end of WW II.

"It's an extremely powerful way of optimising the NPV of an open cut mine, but what no-one else has been able to do is find a way of modelling all the difficult constraints of a pushback in a way that you can come up with an optimised solution, satisfying the mining requirements," he said.

Rubenstein said the University of Melbourne's decision to patent the algorithm was unusual, but reflected the commercial potential of the idea.

"This is the way we are doing it because there are really interesting ideas there and we really want to produce the best possible solution for the mining industry," he said.