Department of Mining Engineering
Geotechnical considerations to reduce adverse impacts
of mine waste rock dumps on the environment
Rennie B. Kaunda
Department of Mining Engineering
What gets measured gets managed
Peter Drucker(?)
(slideshare.net)
2
Department of Mining Engineering
Outline
• Problem statement and motivation
• Principal component analysis (PCA)
• Results
• Summary and conclusions
3
Department of Mining Engineering
Acknowledgements
• Tailings and Mine Waste ‘18 Conference Organizers and Session Chair
• Colorado School of Mines
4
Department of Mining Engineering
Problem statement: How should the slope stability analysis be assessed??
5
Department of Mining Engineering
Strength envelopes
t – s coordinates
(normal and shear stresses acting on the failure
plane)
s
t
s3 s1s3scis3 s1
t = c + s tanf
c
f
Highly nonlinear especially here
6
Department of Mining Engineering
Factors affecting shear strength and stability analysis
• Dump Height
• Density
• Moisture
• Weathering
• Transport
• Construction method
• Ore type
• Foundation/topography
• Mineralogy
• Clay content
• Mechanical properties
• Geochemistry
• Particle shape
• Porosity
• Petrology
• Other 7
Department of Mining Engineering
Empirical correlations to establish shear strength
(Dorador et al., 2017, Material characterization, In: Hawley, M., & Cunning, J. (eds.), Guidelines for Mine Waste Dump and Stockpile Design: 13-27. The
Netherlands: CSIRO PUBLISHING/Balkema)
8
Department of Mining Engineering
Large-scale triaxial testing
(Linero et al., 2008) 9
Department of Mining Engineering
Gradation analysis
(Dorador et al., 2017, Material characterization, In: Hawley, M., & Cunning, J. (eds.), Guidelines for Mine Waste Dump and Stockpile Design: 13-27. The
Netherlands: CSIRO PUBLISHING/Balkema)10
Department of Mining Engineering
Data Dimension Reduction: Principal Component Analysis (PCA)
• A statistical technique used to reduce the dimensionality of a data set
• Useful technique to discover patterns in high-dimensional data
• Basic idea is to capture as much of the variation in the data as possible, using covariances and eigen vectors
11
Department of Mining Engineering
• Calculate the covariance between all possible axes (i.e. dimensions)
• Example for three axes expressed in covariance matrix C:
),cov(),cov(),cov(
),cov(),cov(),cov(
),cov(),cov(),cov(
zzyzxz
zyyyxy
zxyxxx
C
)1(
))((
),cov( 121
n
yyxx
AA
n
i
ii
Data Dimension Reduction: Principal Component Analysis (PCA)
12
Department of Mining Engineering
Data Dimension Reduction: Principal Component Analysis (PCA)
• Calculate the eigen vector and eigen value of the covariance matrix of adjusted original data (transformed using the mean of the data)
• The eigen values with the largest absolute values capture most of the variability in the dataset (principal components)
Ce = λe
13
Department of Mining Engineering
0
5
10
15
20
25
PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
Var
ian
ce (
%)
Data Dimension Reduction: Principal Component Analysis (PCA)
14
Department of Mining Engineering
Global Database:120 test data
*D10, D30, D60, and D90 = 10%, 30%, 60% and 90% percent passing sieve sizes respectively; cu = coefficients of uniformity; cc = curvature; GM = gradationmodulus; FM = fineness modulus; R = ISRM hardness rating; “min UCS” = minimum uniaxial compression strength of group; “max UCS” = maximumuniaxial compression strength of group; “dry unit wt” = dry unit weight at corresponding confining stress.
GM=∑(Cumulative percentage passing sizes through specified sieves)/100
15
{ 1 ½ in. (37.5 mm), ¾ in. (19 mm), 3/8 in. (9.5 mm), #4(4.75 mm), #8(2.36 mm), #16(1.18 mm), #30(600 µm),
#50(300 µm), #100(150 µm), and #200 (75 µm)}
Parameter*D10
(mm)
D30
(mm)
D60
(mm)
D90
(mm)Cc Cu GM FM
Hardness
(R.)
Min UCS
(Mpa)
Max UCS
(Mpa)
Dry unit
wt
(kN/m3)
Confining
Stress
(Mpa)
Shear
Stress
(Mpa)
Min 0.02 0.92 3.2 10 0.42 1.36 0.2 3.94 1 1 5 9.32 0.002 0.0065
Max 33.9 42.4 80.1 100 22.3 220 5 8.8 6 250 400 38.9 4.21 3.92
Mean 5.49 9.25 20.4 42.7 2.27 33.7 2.72 6.35 4 73.4 166.6 20.6 0.67 0.63
STDEV 10.05 11.56 15.92 24.1 3.6 46.73 1.26 1.27 1 41.84 92.17 5.44 0.76 0.66
(N = 120)
Department of Mining Engineering
*Parameter A=D10, B=D30, C=D60, D=D90, E=coefficient of uniformity, F=coefficient of curvature, G=gradation modulus, H= finenessmodulus, I= ISRM hardness rating, J= minimum uniaxial compression strength, K= maximum uniaxial compression, L= dry unit weight,M=confining stress, N=shear stress.
Data Dimension Reduction: Principal Component Analysis (PCA)
16
A B C D E F G H I J K L M N VarianceCum.
Proportion
PC(1) 0.305 0.344 0.359 0.264 -0.2 -0.13 -0.366 0.362 0.284 0.287 0.284 -0.141 0.037 0.06 6.09 43.50%
PC(2) 0.28 0.205 -0.063 -0.282 0.086 0.017 -0.019 0.042 -0.055 -0.076 0.014 -0.216 -0.6 -0.608 2.45 61.00%
PC(3) 0.207 0.237 0.207 0.274 0.344 0.172 -0.238 0.212 -0.428 -0.334 -0.365 0.321 0.042 0.018 2.07 75.70%
PC(4) -0.041 -0.02 0.116 0.063 0.474 0.592 0.091 -0.123 0.18 0.372 0.35 0.276 -0.096 -0.053 1.4 85.80%
PC(5) 0.014 -0.031 -0.091 0.104 -0.035 0.613 -0.028 0.115 0.115 -0.193 -0.187 -0.69 0.127 0.107 0.76 91.10%
PC(6) 0.46 0.284 -0.162 -0.588 0.277 -0.05 0.061 -0.052 -0.058 0.017 0.124 -0.074 0.371 0.296 0.47 94.50%
PC(7) -0.295 -0.115 0.199 0.081 0.601 -0.388 0.014 0.111 -0.246 0.177 0.076 -0.48 -0.012 0.033 0.34 96.90%
PC(8) 0.151 0.236 0.476 0.104 -0.255 0.065 0.435 -0.461 -0.337 0.23 -0.131 -0.16 -0.023 0.05 0.18 98.20%
PC(9) -0.348 -0.082 0.417 -0.603 -0.188 0.201 -0.169 0.331 -0.127 0.209 -0.212 0.082 0.069 -0.061 0.12 99.10%
PC(10) -0.065 0.177 0.234 -0.076 0.258 -0.158 0.019 -0.231 0.677 -0.069 -0.536 0.027 -0.065 0.045 0.05 99.40%
PC(11) -0.119 0.082 0 0.022 0.006 0.005 -0.476 -0.476 -0.049 0.01 0.058 -0.074 0.498 -0.516 0.04 99.70%
PC(12) -0.103 -0.003 0.077 -0.135 -0.04 0.05 -0.533 -0.411 -0.107 -0.157 0.167 -0.059 -0.45 0.488 0.03 99.90%
PC(13) 0.27 -0.232 -0.343 0.08 0.005 -0.006 -0.258 -0.049 -0.118 0.656 -0.472 -0.012 -0.096 0.067 0.01 100.00%
PC(14) -0.489 0.73 -0.397 0.053 -0.043 0.039 0.039 0.066 -0.088 0.191 -0.061 0.013 -0.063 0.076 0.01 100.00%
*Principal
Component
Parameter
Department of Mining Engineering
Summary and Conclusions
• Many previous studies have recognized the correlation between rock
particle size distribution and shear strength in waste rock dumps
• This study has identified and ranked key geotechnical parameters in
terms of their influence on the shear strength of waste rock, based on
a compiled global database (120) and principal component analysis
• These include GM, FM, D60% passing size; for a given confining
stress, rock particle hardness, UCS, and density.
• However, when possible site specific tests and data should be
acquired
17