98 over the tonnage extracted. An example of a mine with four layers with distinct characteristics is shown in Figure 2. It depicts a representative realization of P2O5 variability in each layer, along with light shaded bands representing the range of fluctuation over 100 simulations. By default, one layer is excavated from a mine each day until it is exhausted, at which point the next layer is excavated. A user-defined layer excavation schedule which specifies which layer should be excavated from each mine on each day can be provided for the simulation. Overburden and interburden layers are ignored for simplicity. Piles store inventory up to a capacity limit, which is typically high enough to be irrelevant. Piles optionally can start with an initial inventory. Fixed mine and pile parameters are presented in Table 1. Figure 2. Variability of P2O5 composition in the feedstock from four layers in one mine. Each layer has a different average composition. The curves a representative generated feedstock composition with respect to cumulative extracted tonnage, and the light shaded bands represent the range of values attained over 100 different realizations. Table 1 – Fixed mine and pile parameters. Parameter Value Units Mine production capacity 5000 tonnes/day Mine operating cost 5 $/tonne Number of piles 4 Pile capacity 100000 tonnes Pile initial inventory 0 tonnes Process pathways define possible processing pathways through which material can flow, rather than specific processes themselves. For example, process pathways for phosphate processing could include washing; washing and flotation; washing, grinding, and flotation; calcination; directly selling material with no mineral processing; or sending material to the waste pile. Each process pathway has a defined processing cost per ton, a tonnage reduction factor, and a set of process transformation factors (one for each tracked chemical compound, e.g., one for P2O5 and one for MgO) that describe the way material is transformed through processing. To represent stochasticity, these process factors are generated in a similar way to feedstock variability—they are drawn from a probability distribution defined by a mean and variance, but they vary over time, rather than in space. Process parameters are displayed in Table 2.
RkJQdWJsaXNoZXIy MTM0Mzk2