Search

Home > Research Areas > Amira P9
Research Areas

Amira P9

AMIRA P9 is the world’s largest university-based mineral processing research programme. In its 50-year history it has re-shaped the practice of designing and optimising mineral processing plants, using mathematical modelling and computer simulation. The research team includes some of the world’s leading researchers in this area.

P9 began in 1962, when the Australian Minerals Industry Research Association (AMIRA) commenced a collaborative research project between the University of Queensland and thirteen Australian mining companies. The project, commonly called ‘The Mineral Processing Project’, was the ninth AMIRA project, and hence termed P9. Over the years, the project has been extended 16 times, in three to four year cycles. The current extension is P9P (2012-2015).

The programme is the ultimate example of a collaborative venture. The P9P project involves five universities. It is led by the JKMRC (University of Queensland), and includes contributions from the University of Cape Town, Haceteppe University (Turkey), the Federal University of Rio de Janeiro and Chalmers University, Sweden.

The project is currently sponsored by 20 companies on five continents, representing all parts of the global mineral industry community. It also achieves an extraordinary degree of leverage of resources through additional government funding in Australia and South Africa, and through the participation of the large number of companies supporting the work. P9 has won a number of awards for its collaborative research achievements and associated software outcomes (eg JKSimMet, JKSimFloat).

UCT joined the project in 1996, when it was recognised that the CMR flotation research group had a major contribution to make in this field. Since then, significant and successful bi-national collaboration has been formed between the CMR and the JKMRC, which has added enormous synergy and value to the overall project in the areas of flotation, comminution and classification. The collaboration led to the establishment of the CRV in 1998, and has been well supported by South African Government (THRIP) funding.

A key element of the P9 philosophy has been the reliance on the energy and talent of postgraduate students. Their theses have been the intellectual basis of the research and many of the students have gone on to enjoy successful careers in the industry, to the further benefit of the sponsors.

Over the years, the P9 project has developed mathematical models of various units in the minerals processing chain, which can be used to create strategies for designing and optimising processes through simulation. The models have evolved from being typically full scale and empirical in nature – and hence able to predict the outcomes of changes that can be done on an existing plant – to become more mechanistic, including the effect of both the machine and the feed characteristics. Tools have been developed to characterise the hardness of feed ores and gas dispersion sensors to measure the hydrodynamic properties of flotation pulps. Recently, the use of the Discrete Element Method (DEM) and Computational Fluid Dynamics (CFD) has allowed much improved computational modelling of milling, classification and flotation.

The major limitation of the current models is that they cannot predict the liberation of the minerals in the ore in the comminution process, or the performance of the downstream separation units on the basis of the mineral composition of the particles in the feed streams. As a result, the models are comminution-, classification- or flotation-specific, and cannot be used to simulate an entire mineral processing chain.

The vision of the P9P project is to develop an integrated multi-component simulator structure of the entire comminution, classification and flotation process chain, and multi-component models of the unit operations. The project will also deliver new measurement and characterisation (and other) tools, which will considerably enhance the ability to predict and improve plant performance.