Operational and Control Decision Making for Aluminium Smelters

Abstract

Aluminium smelters are continually facing challenges which include not only the drive to achieve high productivity and high effi ciency but to also comply with strict environmental regulations and reduce energy consumption. Focusing solely on improving automatic control is not enough. The influence of human factors on process outcomes has been recognised as a very crucial factor. However, the literature survey shows the understanding of the mechanism of human decisions in aluminium smelting operation and control and the interaction between human and systems (including operational systems, control systems and management systems) is insufficient. This research therefore studied human reasoning, decision making and human behaviour with respect to their influence on operation and process control in the complex aluminum smelting industrial environment. A decision making model, which synthesises the naturalistic decision making process (RPD), an advanced control model, as well as a human and system interactive approach, was developed to provide guidance to staff for making operational and control decisions. A feedback model was also developed for decision making improvement. These concepts and models have been incorporated into a new system named Advanced Supervisory Control and Management System (ASCMS), which provided a means for verification of the models and hypothesis testing. Two experiments which were conducted in a laboratory setting provided evidence that visual guidelines and a human and system interactive decision making approach (decision making model) improved the operators' performance on detecting abnormalities while controlling the process. Two field studies carried out in operating smelters involving bath height control and the implementation of a new anode setting procedure showed the effectiveness of the visual and procedural guidelines in the supervisory system on improving the operational staff's decision making and hence the process performance. The field investigation of perception of workload increased the understanding of operators' thinking and behaviour towards their routine tasks. In all field studies, it was found that provision of feedback to staff (at every level) was crucial. Furthermore, modification of management systems, such as work design, was required to allow operators' performance improvement. These findings support the feedback model. Apart from the major findings above, from this research, it is also understood that attention and perception play significant roles in human thinking, reasoning and information processing. They directly affect the results and quality of every decision made. More research on improving and developing strategies which take into account of attention and perception, is strongly recommended for an information rich process such as aluminium smelting process. In conclusion, the understanding of human reasoning, decision making and human behaviour in the context of aluminium smelting process operation and control has been increased through this research. The learning and knowledge about human decision making and behaviour gained have provided feedback to improve ASCMS design and implementation. The findings in this research also provide feedback and guidance to smelters for future decision making in operation and control improvements which contributes to tackling the challenges present in the industry.