A world leader in the production of commodities globally. Primary business focus lies on extraction of minerals, but with significant operations in refining, especially of iron ore.
Regular maintenance shut windows are scheduled to ensure that a major processing plant is maintained optimally for it to process at its maximum capacity. During these shut windows, there are thousands of maintenance tasks that must be carried out that all have:
Each shut window varies in duration and requires a large workforce to conduct the maintenance. A major concern during each shut is keeping an acceptable level of safety risk by managing the amount of workers on site.
The challenge was to create an effective shut strategy that aimed to reduce worker related safety risks, maintain the plant effectively, while maximising plant throughput.
To achieve the desired result of increasing site safety, it was important to lower the number of workers on site per shut. Off the shelf software was not available and the size and complexity made this a particularly difficult problem to solve. Working closely with the customer, Polymathian designed and developed bespoke optimisation software that integrated with their existing system.
This software acts as a decision support tool and utilises the latest in mathematical programming to provably select the best allocation of maintenance tasks to shuts given the challenge. The tool is able to be used operationally and strategically; operationally to allocate maintenance tasks to shuts, and strategically to evaluate alternate shut strategies. Not only is the tool able to produce the optimal strategy, it helps to validate and cleanse many thousands of rows of maintenance related data.
In addition to the software, Polymathian provided consulting services to run a number of scenarios and to make recommendations.
Polymathian were able to recommend a strategy, that removed unnecessary shut windows and resulted in:
As a result, the tool successfully produced a strategy that saw a 40% reduction of workers during peak maintenance periods. This was a phenomenal result as it greatly reduced the site’s safety risks while improving overall plant throughput. This was all achieved with minimal disruption to previous planning processes and rapidly delivered in fourteen weeks.