When pivotal pieces of equipment on mine sites –like crushers or furnaces – need to be shut down for maintenance, the operation is disrupted, often at massive cost. Each day of inactivity can represent hundreds of thousands of dollars of lost production, so maintenance shutdowns need to be carried out as efficiently and quickly as possible.
Shutdowns are usually longer than the duration of the single longest maintenance activity, as the range of activities undertaken are often dependent on other operations, or need to be performed in a particular sequence. The maintenance workforce usually requires diverse skills, from boilermakers to electricians and mechanics. There are not enough labour hours available to do all tasks at once, so optimal scheduling of activities using the resources available is crucial. It may even be cost effective to hire additional maintenance staff for a shutdown, even considering their additional transport and accommodation costs, to shorten its duration.
Every shutdown is different, and any problems encounter enduring early stages of maintenance have the potential to change the nature and timing of subsequent activities.Managing the ever-changing connected activities of a maintenance shutdown to achieve the greatest efficiency isa complex and high-stakes problem.
Each shutdown has a set of inter-related activities, a set of actual or potential resources (both equipment and labour skills, and their associated costs. Mathematical algorithms can be designed to model the general sequencing and supply chain optimisation problem, but need to be customised to the individual maintenance shutdown circumstances. Optimising a maintenance schedule to maximise the total number of activities in the shortest space of time is only part of the problem. As the maintenance schedule progresses and problems arise, equipment breakages or shortages become apparent, or the time required to complete a maintenance task changes, the remainder of the schedule is affected and needs to be re-optimised to meet the new circumstances.Tracking the progress and success of each task becomes critical to allow schedule optimisation on a daily basis.
The Optimisation and Control Research Group has created optimisation models for various shutdown maintenance scheduling problems. These models consider the critical activities, precedence and time constraints and resource limitations, resulting in a number of different optimised scenarios for planners and managers to consider. Once the preferred scenario is selected and the external resources finalised, a detailed schedule and timeline of activities can be created to minimise the total shutdown time. Once the shutdown begins, additional corrective maintenance activities can be added on a daily basis, with the models used to find the best schedule to fit new activities in, or prioritise tasks and resources to minimise the amount of overtime required.
Maintenance shutdown time can be minimised by using mathematical modelling to test scenarios to find the best set of maintenance resources and conditions, and then optimise scheduling activities once the ‘boundary conditions’ for the shutdown are chosen. Continually optimising a live maintenance schedule as circumstances change requires not only a robust model of the system, but also accurate tracking and recording of operations as they progress in real time. Commercial partners are addressing this limitation, moving towards the scenario where real-time optimisation and scheduling of a‘live’ and evolving maintenance operation is possible. This combination of activity tracking and live optimisation has the potential to improve every type of supply chain and logistics challenge.
