The sizing of a starter dam for a Tailings Storage Facility (TSF) involves balancing many competing demands._x000D_
It should cost as little as possible, be high enough to store tailings until the first raise can be constructed, it must be able to contain water safely, and it must be designed in such a way that the long-term integrity of the higher raises are not compromised, among other things._x000D_
In addition to these demands there is the complication of the behaviour of the process plant, and the tailings during commissioning._x000D_
Variations in slurry density, excess water in the TSF, initial settled density of the tailings, reliability issues, and beach slopes above and below water can all impact the way the tailings beaches are developed and ultimately impact the ability of the starter dyke to contain the required volume of tailings and water.The conventional approach to the design of a TSF starter dyke is to begin looking at the elevation required for a set of deterministic design parameters, find ways to reduce the size or volume of it as much as possible to reduce cost, and then possibly look at the sensitivity of some of the input parameters._x000D_
This paper discusses a stochastic modelling approach for starter dykes that allows a wide range of design inputs to be simulated quickly, with an envelope of possible results produced._x000D_
The sensitivity to different parameters can also be examined to understand which design inputs have the biggest impact on the required dyke sizing._x000D_
By integrating the 3D TSF modelling tool Muk3D with GoldSim, it will be demonstrated how a wide range of scenarios can be evaluated and how that information can be used to make better decisions about the selection of design options.CITATION:Cooper, C A and Suter, G A, 2018. Starter dyke optimisation with Muk3D and GoldSim, in Proceedings Mine Waste and Tailings Stewardship Conference 2018, pp 466-475 (The Australasian Institute of Mining and Metallurgy: Melbourne).