Optimal Selection of Selective Mining Unit Size For Geostatistical Modeling

Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

Optimal Selection of Selective

Mining Unit Size for Geostatistical


Modeling

Oy Leuangthong, Steve Lyster,


Chad Neufeld, and Clayton V. Deutsch
Centre for Computational Geostatistics
School of Mining and Petroleum Engineering
Department of Civil & Environmental Engineering
University of Alberta
Outline

• Conventional Approach
• Proposed Method
• Example
• Sensitivities
• Practical Considerations
Conventional Approach
• Conventional definition:
– The smallest volume of material on which ore / waste classification is
determined
• Build models at SMU scale:
– Estimate the grades (Block kriging)
– Estimate the distribution (IK, MG, ...)
– Simulate the grades (SGS, SIS, ...)
• Predict production based on the model
• Use blastholes to calculate final tonnages and grade
Applying the SMU

• Estimation (Kriging,…)
– Calculates estimate and variance at SMU scale
– Data-configuration and search-dependent accounting for change of support
• Estimating a distribution (IK, MG, ...)
– Predict a point-scale distribution
– Scale up the distribution with analytical change of support model
– Provides distribution of possible values at SMU scale
• Simulation (SGS, SIS,…)
– Construct multiple realizations
– High resolution at data scale
– Block average to SMU size (accounts for change of support)
Applying the SMU
• Geological features:
– The smaller the SMU, the better the boundaries are defined
– Structures smaller than the SMU are lost
– Sub-blocks and grid refinement help for boundaries
– Our goal is to make block model resources/reserves match practice
Considerations

• In practice it is impossible to freely select an SMU


– The smallest size may not be the best choice
• SMU depends on:
– Selectivity of the equipment
– Grade control practice
– Available data (blastholes, grade control drilling, ...)
– Process used to calculate dig limits
– Dilution - Internal (inside blocks), External (geological boundaries),
and Operational
Proposed Method
Definition of SMU:
– The block model size that would correctly predict the tonnes of ore, tonnes
of waste, and diluted head grade that the mill will receive

1. Create high resolution (HR) model:


– Area equal to monthly / quarterly production
– Grid size equal to 1/3 to 1/10 blasthole spacing
– Populate using simulation or kriging
– Account for trends and geological controls
2. Use HR model to simulate blasthole /exploration
drilling
– Account for sampling errors
– Grade control drilling (if planned)
Proposed Method

3. Simulate grade control to get dig limits


– Apply dig limits to HR model to get reference production
– Total waste (Tw), total ore (To), and average ore grade (zo)
4. Create multiple realizations of the area from drillhole data
5. Choose a range of SMU sizes and block average
– Account for drillhole spacing
– Say from 5m x 5m to 50m x 50m
6. Calculate tonnes and ore grade for all SMU sizes
– Average multiple realizations at the same SMU
7. Plot results and choose SMU
– Match results to reference values
Choosing an SMU Size

• For cutoff grade below


the mean (Zc < m)
• The true values comes
from HR model
• Look for SMU that is
closest to the truth
Choosing an SMU Size

• For cutoff grade above the


mean (Zc > m)
An Example
Mean = 6.30
Median = 5.47

1m x 1m x 5m Blocks 5% Cut-off
Example - Conventional Method
Blasthole Data (10m spacing) Kriged Map (2m x 2m)

HR Model Dig Limits

Blasthole Dig Limits

Waste areas are included


Example - Conventional Method
• Reference values
– To = 2.426 million tonnes (above 5%)
– Tw = 2.434 million tonnes
– Zo = 7.35%
Example - Simulation

Exploration Drillholes 5m x 5m Realization


(50m Spacing)
Example - Block Averaging
Probability of Ore Grade of Ore Probability of Ore Grade of Ore

5x5 25x25

10x10 30x30

15x15 40x40

20x20 50x50
Example - Choosing an SMU
Tonnes of Ore Tonnes of Waste

To = T - Tw

2.43 MT 2.43 MT

Grade of Ore

7.35%
Example - Choosing an SMU
• 20m x 20m SMU:
– Tonnes are correct
– Grade is low, 7.22%
– 1.77% error in total metal tonnage (Grade * Tonnes)
• 6m x 6m SMU:
– Grade is correct
– Ore tonnage is low, 2.37 million tonnes
– 2.47% error in total metal tonnage
• 20m x 20m SMU has the least error
• Operationally grade may be more important
– 7.22% for 20m x 20m vs. 7.35% for 5m x 5m
Example-Sensitivities
Optimal SMU sizes based on different reference
ore and waste polygons
Diglim program output
8 different ore/waste polygons
Hand digitized polygons
Example-Sensitivities
Optimal SMU sizes for different cutoff grades

• Based on tonnage:
– SMU size is max at cutoff 5.0%
• Based on grade:
– SMU size is max at cutoff 5.0%

• Mean grade for all points is 5.211%


• Median grade is 4.911%

SMU size based on tonnage

SMU size based on grade


Practical Considerations
• Need to calibrate results to actual production
– Many site-specific things to consider

• Refine process using previous mined areas


– Model area and use true mill tonnes and grade

• Implement automatic dig limits


• Need to check sensitivity to:
– How blastholes are sampled
– Geological boundaries
– Effects of dilution
– Multiple metals and contaminants

You might also like