simon

Mar 22, 2023

Tailings planning with uncertainty - a Risk-Based Approach

This video showcases Muk3D's capability to run dynamic simulations for tailings planning taking into account uncertainty in both data and the system modelled. Muk3D Simulation Add-On connects Muk3D with Goldsim simulation software to run monte-carlo simulations for your tailings deposition and sequencing.

From deterministic to risk-based planning

For technical and time reasons among others, the traditional tailings planning approach has been deterministic, assuming fix input values while often ignoring uncertainty in both knowledge and randomness of the system.

With the capability of Muk3D to rapidly run tailings depositions, these limitations can now be bypassed. With the Muk3D Simulation Add-On, Muk3D can use the power of Goldsim in representing uncertainty using distributions to dynamically simulate tailings deposition and sequencing.

This approach brings a more sensible approach to tailings planning and allows for the quantification of risks.

Interface between Muk3D and Goldsim

The Muk3D Simulation Add-On creates an interface between Goldsim and Muk3D. Decide which variables you want to import/export and it creates the necessary DLL and server for the software to talk to each other. You can choose to import/export among the following variable types.

  • Float

  • 1-D array

  • Lookup Table

Define your deposition characteristics in Muk3D, develop your simulation model and define distributions in Goldsim and you're ready to go!

Muk3D deposition run inside of a Goldsim simulation

Example of a stochastic tailings planning model

The example showcased in this video simulates the following system.

System Modelled

Find out more

This paper discusses how Muk3D and Goldsim have been integrated allowing for 3D tailings deposition to be incorporated into a water balance model. The difference in results between the 2D models for a TSF used in GoldSim and the same models run using a 3D deposition model are explored, and advantages and limitations of this approach are discussed.