Case Study 2 - Process Modelling to support Capital Decisions

The Business

An export mining business, featuring multiple mines feeding a central railway system which transports the ore to port. At the port, material is stockpiled prior to being exported by ship.

The Problem

The company was seeking to expand production.  They needed to understand the existing capacity of their logistics system, and the additional capacity that could be released through incremental improvements and / or capital expenditure.

The logistics system, which comprises mine stockpiles, rail system, port stockpiles and ship loading, is a highly-coupled system – with the performance of each part impacting every other part.  It was therefore important to model the system as a whole, rather than focussing on the individual components.

The Approach

The approach used was a combination of Driver Tree analysis to determine the steady-state performance of the system and Discrete-Event Simulation to understand the dynamic behaviour.

Details

Driver Tree and Capability Analysis were used up-front to:

  • Engage the stakeholder group, and improve their understanding of system performance.
  • Display current performance data in a way that facilitated the testing of assumptions and quick what-if analysis.
  • Generate and prioritise ideas for improvement.
  • Provide a quick estimate on the capacity limit of the current system, and the likely impact of incremental improvements.

This was followed up by detailed dynamic simulation work, which focussed on providing a more robust estimate of the impact of expansion scenarios that would impact the dynamic behaviour of the system.

The Results

The entire process was very well received by the client, and stakeholders at all levels of the organisation.  Rather than just receiving a set of simulation results at the end of the project, the client and stakeholders were taken on a journey which significantly altered their understanding of the way their system works. The results showed that intuition is not always a good guide in highly-coupled systems; with the some simulation results diverging significantly from expectations.

Dynamic models tend to be a bit of a “black box”, which sometimes makes it difficult for the client to accept unexpected results.  The Driver Trees, with their highly visible logic, made it easier to communicate with the stakeholder groups, check model logic and verify assumptions.   

Note: This project implemented in association with Optika Solutions Pty Ltd.