MAO Essentials Information about Material Allocation Optimizer

MAO Essentials

Lexicon of terms and acronyms

• Material Allocation Optimizer (MAO): optimizes allocation of Inputs to Destinations using Linear Programming.
• Input: one of the following material resources:
• Mine production subdivided into rock types and “grade classes”.
• Stockpiles.
• External.
• Destination: one of the following material destinations:
• Processing method.
• Blended product (defined as a pseudo processing method).
• Stockpile.
• Grade Class: ore (raw material) mass with mineral elements content satisfying one of the following criteria:
• Grade (quality) of a selected element falls within certain limits; for example, Au grade is between 1.8 and 1.85 grams per ton.
• Weighted value of all products per ton.  more...
• Value of a formula defined function of elements falls within certain limits; for example Lime Saturation Factor (LSF) (=100*CAO/(2.8*SIO2+1.1*AL2O3+0.7*FE2O3)) is between 2.3 and 2.4
• Stockpile: place where ore (raw material) of a given type is stored and from which it can be later reclaimed for processing. In and out flows can be controlled with:
• Maximum storage capacity.
•  “Quality constraints”.
• Maximum reclaim rates and rehandling costs.
• External: external source of ore (raw material).
• Processing Method: actual economic model processing method like Mill, Leach, Kiln etc. or a pseudo-method defining a blended product.
• Blended Product: ore (raw material) blend satisfying predefined quality targets. For example, Iron Ore H: 67% <Fe <68% and 2% <Mg<3% and Iron Ore L: 64% <Fe<65%.
• Quality Target: a rate or ratio of element grades that must be maintained within specified limits; for example Cu>0.8% or Silica Modulus between 2.3 and 2.4.
• Global Target: a rate or ratio of inputs and destinations in mass units that must be maintained within specified limits.
• Input Parcel Model: parcel model built by MAO from geological model and economic settings containing original sub-cell grades and end economic data for all processing methods. Whenever possible, the same input parcel model is used in consecutive MAO runs and even shared between MAO case studies.
• Result Parcel Model: parcel model containing all results of MAO optimization, in particular:
• Parcel destination, processing plant or stockpile.
• The year when it will be mined.
• The sequence number in the Scheduler EOS.
• The revenue from the parcel and the costs of processing.
• The products recovered from the parcel.
• The model information can be added to the original sub-cell model and saved as Datamine file or exported to other application.

Linear Programming (LP) Model

MAO LP Model is setup in terms of the following basic variables:

• Input-Destination; for example 'OXID-MILL', 'SULF-LEACH', etc.
• Input-Destination-Element; for example 'OXID-MILL-AU', 'OXID-MILL-CU', 'SULF-LEACH-SIO2', etc.
• Input-Destination-Product-R; for example 'OXID-MILL-AU-R', 'OXID-MILL-CU-R', where 'R' stands for “recovered”.

The optimization objective and the constraints are linear functions of these variables. Most commonly used constraints like processing capacity limits and quality targets are formulated internally. If needed, you can define other constraints directly in terms of basic variables as “global constraints”.

How it works

• An Input Parcel Model is built or retrieved.
• The ore (raw material) from the mine is divided into grade classes according to user specifications.
• The Linear Programming problem is solved.
• A Result Parcel Model is built; all reports are based on this model.

	Related Topics
	Quick Start Capital Costs Targets Import Essentials Economic Model Essentials  Pit Optimization Essentials Pushback Essentials Scheduler Essentials MFO Essentials