Development of a dynamic maintenance algorithm with multiple scenarios: a case study for surface mining

Ölmez Turan, Merv
Surface mining operations such as ore extraction and overburden stripping activities highly depend on machine performance. These machines’ operational plan aims to handle required amount of material within a specific period with the lowest maintenance cost and the highest availability. In order to achieve these objectives, the machines should be adapted to the production schedule properly. On this basis, maintenance policies play crucial roles in the sustainability of operations. A maintenance policy is basically a combination of work packages that cover the answers of what, when, and how to maintain a machinery system. It should be determined specifically not only the machine itself but also operational dependencies between machines in the fleet. Although the literature on mining machine maintenance modelling commonly concentrates on corrective and preventive actions, opportunistic maintenance that examines whether an opportunity exists for the preventive maintenance of a running component in case of failure of another dependent component is not discussed as required. On this basis, the current study intends to develop an integrated simulation model that considers mathematical interaction of corrective, preventive, and opportunistic maintenance to stochastic uptime and downtime behaviors of subsystems in Arena® Software. The implementation of the simulation algorithm helps to understand the effect of maintenance policy content on total maintenance cost or annual production amount. The proposed algorithm is implemented for two different scenarios. In the first example, an operation with three shovels where corrective, preventive, and opportunistic maintenance activities are applied under the policy is simulated for a period of one year. The simulation model is also performed for five different inspection intervals. According to results, the maximum attainable annual production of three shovels was obtained as 7,266,714 m3 . The increase of inspection intervals were detected to have no significant effect on the fleet availability. However, increasing time between inspections caused a shift from preventive maintenance to corrective maintenance that may cause a remarkable jump in the machine deterioration rate. In this sense, sensitivity of corrective maintenance statistics to the inspection intervals for each subsystem were evaluated. In the second case, six different maintenance policies with different combinations of corrective, preventive, and opportunistic maintenance were applied for a dragline system. In addition, effect of inspection intervals on the total maintenance cost was also evaluated for the policies that include preventive maintenance. The result shows that the total maintenance cost is minimized to 913,480 $ by applying just corrective and opportunistic maintenance. This means that opportunistic maintenance, which is applied during corrective maintenance hours, is good enough to prevent approaching failures; and preventive maintenance in inspections is redundant for the system. Moreover, the corrective maintenance statistics explains that the machinery house and movement subsystems are the most sensitive to inspection intervals where this is least for the rigging and hoisting subsystems.
Citation Formats
M. Ölmez Turan, “Development of a dynamic maintenance algorithm with multiple scenarios: a case study for surface mining,” Thesis (M.S.) -- Graduate School of Natural and Applied Sciences. Mining Engineering., Middle East Technical University, 2019.