The operation of copper flotation circuit is a challenging task for the site metallurgists and plant operators. To optimise the performance requires meeting both grade and recovery objectives. These plants have numerous pieces of equipment that are monitored by hundreds of sensors. The challenge is to review these sensors and adjust the numerous control inputs to maintain peak operational efficiency for extended periods. There is also a need to respond to equipment failure. These challenges lead to in operational inefficiencies and can result in a reduction in revenue.
This paper details an overview of an operational assistance tool implemented at Newcrest's Telfer plant’s copper flotation circuit. This tool is based on a digital twin of the plant. A digital twin is a model of the plant that is simulated in parallel to the normal plant operation. It is fed with the plant data in real time to provide a prediction of plant operation. The model outputs directly generate the metrics that are used to assess the plant performance, and these are compared with the metrics calculated from the actual plant’s operation. Deviations between the model estimate and the plant’s actual performance are used to gather insight into plant health and alert the operator to poor performance. The operators gather more information regarding the root cause by reviewing the model’s inputs that have deviated outside their normal operating range. An evaluation of the plant performance, following the commissioning of the digital twin, showed an overall reduction of acid soluble copper in the tails.
This paper discusses the implementation of a digital twin of the plant to monitor plant performance during live operation. It will include details on data collection and cleansing, data reduction techniques, model development, and real-time implementation. The results of the case study of an implementation of a digital twin at Newcrest's Telfer copper flotation plant will be used throughout the paper.
Oliver, S J and Tooher, R, 2018. Optimising flotation plant operation by using a digital twin in the control room, in Proceedings 14th AusIMM Mill Operators' Conference 2018, pp 93–100 (The Australasian Institute of Mining and Metallurgy: Melbourne).