Mine Smelter Process Improvement Evaluation Case Study

A study of the impacts on ore transfer via gantry cranes and AGVs during process improvement implementation at a mine smelter using Flexsim Simulation Software


A large mining company wanted to evaluate the impacts of competing process improvement plans on the productivity of their smelting process during implementation. The plant being studied initially had five reduction kilns and four smelting furnaces.

The reduction kilns were used to heat the ore to approximately 1200 degrees Fahrenheit to reduce the moisture content and ore oxide. The smelting furnaces melted the ore into slag and a metal matte that was further refined at a later process not included in the simulation model. Transfer of the ore from the reduction kilns to the smelting furnaces was done through a combination of gantry cranes and AGVs. Each of the five AGVs travelled along its own unique set of rails that ran between one of the reduction kilns and the corresponding hoist well. There were five gantry cranes that all traveled along the same set of rails that were available to transport the containers from the hoist wells to the feed bins. The ore transfer process is outlined below:

  1. Ore is loaded into a brick lined container on an AGV at a reduction kiln
  2. The AGV transports the full container to a hoist well
  3. A gantry crane places an empty container onto the AGV and then removes the full container
  4. The AGV takes the empty container back to the reduction kiln to be refilled
  5. The gantry crane hoists the full container up to the feed bin level of the plant
  6. The gantry crane transports the full container to a feed bin that is ready and empties the container into the feed bin
  7. The gantry crane transports the now empty container back to a hoist well
  8. The gantry crane hoists the container down to a ready position and waits for the AGV to return with a full container

Issues to Solve

The ultimate goal of the process improvement plans was to increase the total production capacity of the smelter. With a firm understanding of the implementation costs and capacity increase for each process improvement plan, the main concern for the mine turned to the impacts on smelter productivity during implementation.

The main aspects for the process improvement plans were upgrading and/or the addition of smelting furnaces and the upgrading and/or the addition of reduction kilns. The implementation sequence for each process improvement plan had a built in flexibility to allow many of the different tasks to be worked on at the same time or in any order.
The main goal of the simulation was to quantify the impacts of the different implementation plans so that the mine could determine the best plan.

Simulation Approach

There were two possible approaches to quantify the impact of any given process improvement plan. The first approach was to simulate the entire plan and the second was to simulate each step individually. The second approach was used to allow the user the easiest method for changing the sequence of a given plan and to reduce the number of simulation models that would need to be built.
To give the user the ability quickly and easily simulate any of the different process improvement plan steps the model needed to allow the user to quickly change the following model parameters without editing the model directly:

  • The quantity and parameters for:
  • Reduction kilns
  • Smelting furnaces
  • AGVs
  • Gantry cranes
  • Cross Feeding
  • What reduction kilns could supply material to each smelting furnace and the priority associated with each relationship
  • What gantry cranes could service each reduction kiln and the priority associated with each relationship
  • What gantry cranes could service each smelting furnace and the priority associated with each relationship.

To insure that the model would not need to be modified by the user an Excel spreadsheet was used as the interface for changing these model parameters. This method was also used because it is faster and generally a more recognizable interface for changing large amounts of data when compared to custom GUIs.


The mine smelter operates as a continuous process that changes very little and very slowly over time when compared to other processes. As a result, the original algorithms and methodologies used by the smelter to determine production and dispatching were very rigid. Flexible algorithms and methodologies were required to allow users to change the number of equipment and cross feeding capability, as well as random equipment break downs. This proved to be the hardest concepts to implement in the simulation model.
Calculating the production settings for the reduction kilns and smelting furnaces is a complicated task that is determined by the ability of the reduction kilns to supply the smelting furnaces. The original rigid calculations were hard coded with the number of equipment and cross feeding pattern. A solution was developed based on the Simplex linear programming algorithm. This new method allowed for any combination of reduction kiln and smelting furnace quantities. The cross feeding pattern was specified in a table that was automatically updated when equipment would breakdown.

The original method for assigning gantry cranes to transfer full ore containers to the feed bin was to limit each crane to a designated reduction kiln. With few exceptions, each reduction kiln was generally limited to supplying a single furnace at any given time. A new algorithm was developed that kept track of the available equipment, full containers, and feed bins ready for ore. The algorithm then compared these lists, cross feeding pattern and weights to find the best possible assignment. The cross feeding pattern was continuously updated when gantry cranes would breakdown. This was down to ensure that gantry cranes were not assigned a task that would require them to travel past another gantry crane that was broken down.


The final simulation model proved to be very successful in helping to evaluate and quantify the impacts of implementing the various process improvement plans. Using an Excel spreadsheet to specify the model parameters allowed multiple individuals to work on the designs for various plan steps without requiring them to be familiar with the simulation model. This increased collaboration between different groups and decrease the overall duration of the project.
The model has proven so successful that the mine plans on using the model to evaluate future projects and operating procedure changes that may impact smelter production.


Model and case study completed by Brandon Peterson. Brandon is a simulation engineer with Flexsim Software Products and can be reached at [email protected] or 801-224-6914.