From Mountains to Metal (or Chemical)
The vocabularly (subtlety of differences in - and range) of a language reflects the experiences and expectations of the users (or developers, as the developer is a super user).
Metsim "talks" everything from mountains to atoms - not just in general terms but with complete mathematical integrity across all the individual schools of science between.
To be able to talk about a mineralized mountain peak or a completely liberated mineral particle in a flotation concentrate in the same mathematical model you need to be able to talk (or describe) using different degrees of definition - ranging from initially very vague (like "soft digging", "brecciated", "weathered") to something more specific like "fromboidal pyrite".
You also need to be able to accept, understand, and deal with - in a mathematical way - the uncertainties and interpretations that lead to particular descriptions. The geologist's postulation on the genesis of an ore body is important, even if it is of the kind "on the one hand it could be that...., or on the other hand it could be that....". The open pit contractor will talk in terms of blastability, diggability. The underground mine engineer will talk about ore competence, support requirements. The term "Merenksy Reef" says a lot to some people.
The mineral processor, hydromet, pyromet, and environmental scientist should be listening to all of these descriptions, translating the original parameters (such as soft digging, blocky, arsenic) into other parameters that they will each need (such as compressive strength, liberation characteristics, likely presence of elements that will attack refractory, heap permeability and percolation rates, deportment elements between licensed effluents or gaseous stack emissions etc.).
It is also important to understand (and see in real numbers) the actual exploration drill holes results, the kriging interpretation (and the impact of different algorithms), the potential risk of assumptions about "missing" data (such as from low recovery of drill core) - and how these factors may be biassing an interpretation of both the metal content and process response (and hence the project value).
With time, based on the process plant response and the detailed analysis of the process plant products and other streams, the geologist's postulations will be confirmed or improved, the kiging algorithm will be refined, and the economic value of blocks will be revised.
Then the mine plan can be revised and alternate block sequences can be run to find the best NPV for the "go forward" position. Metsim's "Mine Module" provides the data structure and functionality (for use as intended or customisation using APL) for the translation of the material and ore types to component specifications by size fraction, and for incorporation of block models, which when combined with dynamic sequence modelling through all process types enables complete environmental and economic evaluation.
By using historical data (either real or simulated results) to "feedback" to improve the material type parameters and component translations, Metism provides the basis for a business knowledge system that increases the value recovery for the ongoing operation and provides a valuable starting point for evaluation of other similar ore bodies being considered for development.