Quantitative Methods for Copper Smelter Re-Engineering Projects

The quantitative techniques to evaluate the enhancement of a particular unit operation should consider the corresponding changes to the adjacent unit operations. For example, the expansion of off gas handling capacity should consider the corresponding changes in converter cycle times, matte grade, and other operational parameters, that can provide the optimal system-wide benefit. In general, the upgrading of smelter equipment is risky due to the size of capital investment, as well as any unforeseen impacts on rest of the operations. However, this risk can be mitigated by applying the principles of Front End Loading (FEL) in the formulation of multiphase re-engineering projects. The initial phases of engineering are relatively inexpensive, and determine whether or not to pursue the subsequent phases that are comparatively more expensive. Each phase of engineering depends on increasingly detailed descriptions regarding the critical phenomena that can effect design decisions; the increasing levels of detail may result from numerical simulations, as well as laboratory and plant trials. The quantitative framework should be sufficiently flexible to accommodate these increasing levels of detail, regardless of the particular phenomena, so that it may evolve throughout the project. The current article describes the combined usage of mass balancing, thermochemistry and multiphysics, in the development of discrete event simulation (DES) and finite differences (FD). The incorporation of finite differences into a DES framework depends on adaptive time-stepping, to determine threshold-crossing times, e.g. when the matte temperature surpasses a critical level that necessitates the addition of cold charge. The hybridization of DES and FD is appropriate for multiphase smelter development projects, as it supports semi-continuous material flows, operational policies and scheduling, as well as unexpected events and trends, in increasing levels of detail. This type of framework will support the installation and tuning of modern control systems, including state-of-the art sensors and machine learning algorithms.