Abstract:
Over the last three decades, process control in aluminum smelters has improved significantly as it is the biggest leverage for cost and energy reduction in production and product quality improvement, as well as for meeting environmental compliance. The implementation of computerized automatic control systems two decades ago was a step change in improvement in the state of the art of process control. However, the complex and dynamic nature of the process requires human monitoring, diagnosis, and intervention from time to time. This study investigates the use of the supervisory screen of the control system of a smelter, as well as the effectiveness of visual guidelines to help the operators to identify process abnormalities. The results show that visual guidelines such as voltage patterns which are used as a reference improve the performance of the operators. Detection time and falsealarm rates were reduced in addition to increasing detection sensitivities. It is proposed that a higher level of human and system interaction would improve the overall performance process control.
