Master’s degree in Mineral and Environmental Engineering

Mateusz Janiszewski

Geotechnical risk assessment in the Pyhäsalmi mine with a

focus on seismic risk

Master’s Thesis

Espoo

Supervisor: Professor Mikael Rinne

 Aalto University, P.O. BOX 11000, 00076
AALTO
www.aalto.fi
Abstract of Master's thesis

Author Mateusz Janiszewski

Title of thesis Geotechnical risk assessment in the Pyhäsalmi mine with a focus on
seismic risk
Department Civil and Environmental Engineering
Professorship Rock Engineering Code of professorship RAK-32
Thesis supervisor Prof. Mikael Rinne
Thesis advisor Wen Zhijie
Date 05.08.2014 Number of pages 79 +37 Language English

Abstract

Due to a constant depletion of shallower deposits underground mining is pushing

production to larger depths resulting in more difficult ground conditions and an increase
of rock stresses. Mining at larger depth effects in an increase in seismic activity, therefore
seismic risk is becoming more important and requires special attention. This necessitates
development of frameworks and guidelines to tackle those risks more efficiently.
The scope of this study was to apply and test the Geotechnical Risk Assessment
guideline for underground mines developed at Aalto University as part of the project:
Innovative Technologies and Concepts for the Intelligent Deep Mine of the Future
(I2mine), work package 3 - “Rock mechanics and ground control”, under the 7th
framework of the European Union. The goal was to use seismic monitoring data from the
Pyhäsalmi mine in order to quantify the seismic risk, using proposed guideline as a
framework for evaluation. Furthermore, the research was aimed to give answers on: what
is the applicability of the guideline, what are the main issues encountered during its usage
and how to improve it.
Geotechnical Hazard Potential was evaluated using a proposed methodology. Next,
seismic risk in the Pyhäsalmi mine was assessed. Seismic events from the database were
clustered into logical groups in order to speed up the analysis. Then, the maximum
predicted size of a seismic event was found for each cluster group. Seismic risk was
assessed based on the largest possible seismic event attributable with a particular damage
that is expected. The final risk was evaluated using two methodologies: one that is
currently used in the mining industry and second under the development.
The level of seismic risk in the Pyhäsalmi mine was found to be low. The biggest risk
was found in areas located at the northern ore-waste contact zone. Installation of
additional ground support was recommended in order to lower the risk. The Geotechnical
Risk Assessment guideline was found to be suitable framework for risk assessment
purposes, which supports in selection of appropriate assessment approach and aids in
selection of tools that are used throughout the evaluation. Its applicability was found to
be high, however minor changes were recommended in order to use its full functionality.

Keywords Mining, risk assessment, seismic hazard, Pyhäsalmi mine

ii
Foreword

Foremost, I would like to express my sincere gratitude to my supervisor Prof. Mikael

Rinne for his continuous support and guidance that helped me throughout this study.
Besides my supervisor, I would like to thank my advisor Wen Zhijie for his enthusiasm,
knowledge and insightful observations in all the time of research. Completion of this
thesis would not have been possible without their commitment and effort.

My sincere thanks also goes to Pekka Bergström from the Pyhäsalmi mine who supported
me with collection of all the required data and gave answers to many questions during
research and writing of this thesis.

I thank my colleagues from the Rock Engineering department at Aalto University for their
assistance and all the stimulating discussions we had for the last months. In particular,
I am grateful to Otto Hedström for solving many organizational glitches that would take
much more time without his assistance.

Finally, I would like to thank my parents for supporting me spiritually during my whole
life.

Espoo 05.08.2014

Mateusz Janiszewski

iii
Table of contents

Abstract ............................................................................................................................. ii

Foreword ..........................................................................................................................iii

Table of contents ............................................................................................................... 1

List of figures .................................................................................................................... 4

List of tables ...................................................................................................................... 6

Nomenclature .................................................................................................................... 8

1 Introduction ............................................................................................................... 9

1.1 Motivation .......................................................................................................... 9

1.2 Objectives ........................................................................................................... 9

1.3 Structure ........................................................................................................... 10

2 General description of the Pyhäsalmi mine ............................................................ 11

2.1 Geology ............................................................................................................ 12

2.2 Mining method ................................................................................................. 13

2.3 Rock mechanics and ground support................................................................ 14

2.4 Seismic monitoring system .............................................................................. 16

3 Theoretical background ........................................................................................... 19

3.1 Geotechnical risk assessment guideline ........................................................... 19

3.1.1 Geotechnical hazard potential ................................................................... 19

3.1.2 Geotechnical risk assessment .................................................................... 21

3.2 Seismicity in mines .......................................................................................... 23

3.2.1 Seismic event size ..................................................................................... 23

3.3 Seismic hazard .................................................................................................. 25

3.3.1 Quantitative Seismic Hazard and Risk Assessment Framework .............. 27

3.3.2 Probabilistic Approach for Seismic Risk Assessment .............................. 32

1
 3.4 Data clustering .................................................................................................. 34

3.4.1 Quality Threshold clustering algorithm .................................................... 35

3.4.2 Single linkage hierarchical clustering algorithm....................................... 36

4 Geotechnical risk assessment in Pyhäsalmi mine ................................................... 36

4.1 Geotechnical hazard potential evaluation ......................................................... 37

4.2 Risk assessment tool selection ......................................................................... 39

4.3 Hazard identification ........................................................................................ 42

5 Seismic Risk Assessment ........................................................................................ 43

5.1 Seismic data analysis ........................................................................................ 43

5.1.1 Events statistics ......................................................................................... 43

5.1.2 Seismic events clustering .......................................................................... 47

5.2 Selection of the investigation area .................................................................... 55

5.2.1 Division of selected mining levels ............................................................ 57

5.3 Seismic hazard assessment ............................................................................... 61

5.3.1 PASRA ...................................................................................................... 62

5.3.2 QSHRAF ................................................................................................... 67

5.4 Seismic risk evaluation and representation ...................................................... 70

5.5 Seismic hazard and risk mitigation .................................................................. 71

6 Discussion ............................................................................................................... 73

6.1 Seismic hazard and risk assessment ................................................................. 73

6.2 Geotechnical risk assessment guideline ........................................................... 74

7 Conclusions ............................................................................................................. 75

8 Recommendations for future work.......................................................................... 75

8.1 Seismic risk assessment and mitigation ........................................................... 76

8.2 Geotechnical risk assessment guideline improvement ..................................... 76

References ....................................................................................................................... 77

2
Appendices ...................................................................................................................... 80

A Ground support system standard ............................................................................. 80

B Seismic events clustering ........................................................................................ 81

C Seismic Risk Assessment – PASRA ....................................................................... 86

a. Level 1225 ............................................................................................................ 86

b. Level 1250 ............................................................................................................ 88

c. Level 1275 ............................................................................................................ 91

D Seismic Risk Assessment – QSHRAF .................................................................... 93

a. Level 1225 ............................................................................................................ 93

b. Level 1250 .......................................................................................................... 100

c. Level 1275 .......................................................................................................... 108

E Seismic risk mitigation.......................................................................................... 115

3
List of figures
Figure 2-1. Location of the Pyhäsalmi mine. Source: Google Maps .............................. 11

Figure 2-2. Pyhäsalmi mine underground layout and deep ore dimensions
(Numminen, 2012). .................................................................................................... 12

Figure 2-3. Stope size and mining sequence at the Pyhäsalmi mine.
(Gleeson, 2010). ......................................................................................................... 13

Figure 2-4. Schematic diagram (left) and picture of borehole geophone

installed in the mine. .................................................................................................. 16

Figure 2-5. Seismic system sensitivity map – vertical section, view looking
east, triangles illustrate installed geophones (Meyer, 2014). ..................................... 18

Figure 3-1. Gutenberg-Richter frequency-magnitude power law relation

(F-M diagram, Eq. 4) for large population of seismic events
(from Hudyma (2008): Figure 2.11).......................................................................... 27

Figure 3-2. Risk assessment process applied within MS-RAP

(from Mikula, et al., 2008: Figure 30). ....................................................................... 28

Figure 3-3. EVP vs PPV diagram showing zones with expected damage
(from Mikula, et al., 2008: Figure 38). ....................................................................... 31

Figure 3-4. Steps of clustering process (from Halkidi et al. (2001): Figure 1). .............. 34

Figure 4-1. Results of the geotechnical risk assessment approach selection. ................. 41

Figure 5-1. Number of seismic events subdivided by the magnitude. ............................ 44

Figure 5-2. Number of seismic events per year greater than or equal to
magnitude 0; *projected based on first two months. .................................................. 45

Figure 5-3. Daily histogram of seismic events for the period from
25-11-2002 to 27-02-2014.......................................................................................... 46

Figure 5-4. Diurnal chart showing the distribution of seismic events

by the hour of day; subdivided by the magnitude. ..................................................... 46

Figure 5-5. Flowchart presenting three stages of clustering process. ............................. 47

Figure 5-6. Plot of 646 QT clusters created in the second stage of

clustering using QTCLUST algorithm; plotted in the X-Y space. . ........................... 50

Figure 5-7. Plot of 646 QT clusters created in the second stage of

clustering using QTCLUST algorithm; plotted in the Y-Z space,
view looking east. ....................................................................................................... 51

Figure 5-8. Frequency-Magnitude relation of the cluster QT121. .................................. 52

4
Figure 5-9. Frequency-Magnitude relation of the cluster QT84. .................................... 52

Figure 5-10. A schematic summary of the clustering process ........................................ 55

Figure 5-11. An example of rockburst damage that occurred 24.12.2001

on mining level 1200, at the upper level of stope GBL21. ........................................ 56

Figure 5-12. All damages mapped on the mining levels in 2013, subdivided
into severity classes. ................................................................................................... 57

Figure 5-13. Mining level -1225 with assessment zones.

Colors represent different support type. ..................................................................... 59

Figure 5-14. Mining level -1250 with assessment zones.

Colors represent different support type. ..................................................................... 60

Figure 5-15. Mining level -1275 with assessment zones.

Colors represent different support type. ..................................................................... 61

Figure 5-16. Local magnitude and seismic energy relation plotted

from monitoring data from years 2011-2013. ............................................................ 62

Figure 5-17. Results of REC/AEC for mining level 1225. ............................................. 64

Figure 5-18. Results of REC/AEC for mining level 1250. ............................................. 65

Figure 5-19. Results of REC/AEC for mining level 1275. ............................................. 66

Figure 5-20. EVP vs. PPV diagram illustrating assessment zones with the largest
rockburst damage potential (Mikula, et al., 2008). .................................................... 69

Figure 5-21. The number of assessment zones on each mining level assigned with
Seismic Risk Ratings. ................................................................................................. 70

Figure A-1. Minimum standard for 5.5m wide meshed heading – for wider headings
additional support to be installed (Bergström, 2014). ................................................ 80

Figure A-2. Minimum standard for 5.5m wide heading without mesh – for wider
headings additional support to be installed (Bergström, 2014). ................................. 80

Figure B-1. Cluster groups created in the third stage of clustering using SLINK
clustering algorithm, plan view. ................................................................................. 84

Figure B-2. Cluster groups created in the third stage of clustering using SLINK
clustering algorithm, view looking east. .................................................................... 85

5
List of tables
Table 2-1. Primary stress field (Bergström, 2014). ......................................................... 14

Table 2-2. Primary rock properties (Bergström, 2014). .................................................. 14

Table 2-3. List of active geophones in the seismic monitoring system

(de Jongh, 2013). ........................................................................................................ 17

Table 3-1. Rockmass classification for risk, based on safety margin

(Mishra, 2012). ........................................................................................................... 20

Table 3-2. Rockmass classification for risk, based on modified stability number
(Mishra, 2012). ........................................................................................................... 20

Table 3-3. Geotechnical Hazard Potential classification. (Mishra, 2012). ..................... 21

Table 3-4. Richter magnitude of seismic events compared with

qualitative description of how those events are felt in the mine
(Hudyma & Potvin, 2004) .......................................................................................... 24

Table 3-5. Mechanisms of damaging rockbursts (Ortlepp, 1997). ................................. 25

Table 3-6. Comparison of SHS with relative seismic hazard and the largest event
magnitude (Hudyma & Potvin, 2010; Mikula, et al., 2008). ...................................... 27

Table 3-7. Ground support capacity scale – E2 factor in Eq.7 (Potvin, 2009). ............... 29

Table 3-8. Geological structure - E4 factor in Eq.6 (Potvin, 2009)................................. 30

Table 3-9. Rockburst damage potential scale showing expected rockmass and support
damage (redrawn from Mikula, et al., 2008). ............................................................. 31

Table 3-10. Seismic risk assessment matrix constructed from quantitative exposure
rating (vertically) and rockburst damage potential (horizontally)
(Mikula, et al., 2008). ................................................................................................. 32

Table 4-1. Results of the modified stability number for massive

sulfide and volcanites. ................................................................................................ 38

Table 4-2. Results of the safety margin for massive sulfide and volcanites. .................. 38

Table 4-3. Results of rockmass classification for massive sulfide;

risk in operation stage................................................................................................. 38

Table 4-4. Results of rockmass classification for volcanites; risk in operation stage. .... 38

Table 4-5. Bow Tie Analysis for large seismic event. .................................................... 42

Table 5-1. Basic statistics of seismic event parameters. ................................................. 44

6
Table 5-2. Sensitivity analysis of QTCLUST algorithm parameters. ............................. 49

Table 5-3. Five most populous cluster groups created in the third stage of clustering
using SLINK algorithm. ............................................................................................. 54

Table 5-4. Absorbed Energy Capacity scale for ground support

(*including fibrecrete when mesh is not used and shotcrete when mesh is used). .... 63

Table 5-5. Ground support capacity (E2) scale for ground support. ............................... 67

Table B-1. Five most populous clusters created using the QTCLUST algorithm with
parameters used for evaluation and comparison in the third stage of clustering
procedure. ................................................................................................................... 81

Table B-2. Cluster groups created in the third stage of clustering using SLINK
clustering algorithm. ................................................................................................... 82

Table C-1. Results of the seismic risk assessment using PASRA

for mining level 1225. ................................................................................................ 86

Table C-2. Results of the seismic risk assessment using PASRA

for mining level 1250. ................................................................................................ 88

Table C-3. Results of seismic risk assessment using PASRA

for mining level 1275. ............................................................................................... 91

Table D-1. Results of seismic risk assessment using QSHRAF

for mining level 1225. ................................................................................................ 93

Table D-2. Results of seismic risk assessment using QSHRAF

for mining level 1250. .............................................................................................. 100

Table D-3. Results of seismic risk assessment using QSHRAF

for mining level 1275. .............................................................................................. 108

7
Nomenclature
AEC Absorbed Energy Capacity

ASTH Apparent Stress Time History

EVP Excavation Vulnerability Potential

F-M diagram The Gutenberg-Richter Frequency Magnitude Diagram

GHP Geotechnical Hazard Potential

GRA Geotechnical Risk Assessment

I2mine Innovative Technologies and Concepts for the Intelligent

Deep Mine of the Future

MAG, ML Local magnitude

PASRA Probabilistic Approach for Seismic Risk Assessment

PPV Peak Particle Velocity

QSHRAF Quantitative Seismic Hazard and Risk Assessment Approach

developed by the Australian Centre for Geomechanics

QTCLUST Quality Threshold Clustering algorithm

REC Released Energy Capacity

RDP Rockburst Damage Potential

SHS Seismic Hazard Scale

SLINK Single Linkage Hierarchical Clustering algorithm

SRR Seismic Risk Rating

8
1 Introduction

These days, shallower deposits are being depleted and underground mines are pushing to
larger depths. This results in much more difficult rock stress conditions that pose more
geotechnical risks on mining operations. Larger stresses are also attributable with an
increase in seismic activity, therefore seismic risk becomes more significant and
necessitates special attention. Development of frameworks and guidelines for
underground mines is required to assess and manage those risks. This study is part of the
I2mine project (Innovative Technologies and Concepts for the Intelligent Deep Mine of
the Future) work package 3 - “Rock mechanics and ground control”, under the 7th
framework of the European Union, which aims for new approaches in rock mechanics
and ground control to prevent geotechnical related accidents and financial losses in
underground mines with increasing mining depths in the future.

1.1 Motivation

The main reason of this study is to investigate practical application of the Geotechnical
Risk Assessment guideline for underground mines proposed by Ritesh Kumar Mishra in
his Master thesis at Aalto University. A preliminary study to test this guideline was made
at the Garpenberg mine in the fall of 2013 (Froehlich, 2014). The study revealed that the
analysis of seismic events appears to be a valuable tool to analyze geotechnical risks. The
study at the Pyhäsalmi mine aims to improve this method by applying more extensive
database.

1.2 Objectives

The main objective is to apply the Geotechnical Risk Assessment guideline using data
from the Pyhäsalmi mine. The study focuses on the seismic hazard by evaluation of both
the probability and the consequence of an undesirable event. The goal is to establish a
risk ranking for different areas under investigation by analyzing the seismic events
database and other factors influencing the risk. The research question posed in this thesis
is:

 What is the level of seismic risk in the Pyhäsalmi mine and which areas in the
mine require special attention in terms of risk related to seismic events?

9
The second objective is to suggest on how to improve the Geotechnical Risk Assessment
guideline using the experience of testing it with real data from the mine. The research
questions associated with the second objective are:

 What is the applicability of the guideline and what are the main issues encountered
during its usage?
 Which features of the guideline can be changed to improve it?

1.3 Structure

The thesis is divided into seven main parts. Chapter 2 gives an overview of the Pyhäsalmi
mine and describes: geological and rock mechanical conditions, basic information on
mining method and ground support used in the mine, and seismic monitoring system.
Next, chapter 3 provides a theoretical background for this study and briefly defines three
aspects: the Geotechnical Risk Assessment guideline and its main aspects that are used
in this thesis, mine seismicity and basic definitions related to it and data clustering
procedure with a focus on clustering of seismic events. Then, chapter 4 describes the
process of geotechnical risk assessment performed in the Pyhäsalmi mine according to
the guideline, as well as evaluation of geotechnical hazard potential. Selection of risk
assessment approach and hazard identification is presented. Next, chapter 5 focuses on
seismic risk assessment using two methods: one that is used in the mining industry and
second new approach under the development. Chapter 6 gives a discussion on seismic
risk assessment in the Pyhäsalmi mine and strengths and weaknesses of the Geotechnical
Risk Assessment guideline encountered during this study. Finally, chapters 7 and 8
present final conclusions and recommendations for future work.

10
2 General description of the Pyhäsalmi mine

The Pyhäsalmi mine (63º39’31’’N - 26º02’48’’E) is

located in central Finland about 450km north from
Helsinki (Figure 2-1). It is an underground copper,
zinc and pyrite mine owned by First Quantum
Minerals. It is the oldest Finnish metal mine in
operation and one of the oldest in Europe.
Pyhäsalmi is the deepest mine in Europe, with the
depth reaching 1445m.

The orebody with reserves of 30Mt was discovered

in 1958, and production of ore started in 1962 with
an open pit mining and continued until 1976 when it
was switched to underground mining. In late 1990’s
the deposit was close to depletion and deep drilling
program lead to a discovery of the orebody
extension below the 1050m level and doubling of
ore reserves. Production in the “new mine” started
in 2001 with a projected mine life of about 14 years.
Figure 2-1. Location of the
The access to the mine provides a 1450m deep, 5m- Pyhäsalmi mine. Source:
diameter shaft installed during the expansion and an Google Maps
extended ramp (Figure 2-2).

As at December 31, 2012, the proven reserves amounted to 8,5Mt of ore, and measured
resources estimated to further 8Mt. In the year 2013 the ore milled amounted 1,3Mt,
resulting in production of 14 800t of copper, 21 600t of zinc and 825 800t of pyrite. (First
Quantum Minerals, 2013).

The mine has outstanding safety records, as well as high efficiency. Currently it employs
210 people directly and 50 through subcontractors (Trzaska, et al., 2010; First Quantum
Minerals, 2013). At present production commences on the levels between -1050 and -
1350. At the current metal prices, the end of mining is expected in 2019, with lower
tonnage already in 2018 (Pyhäsalmi Mine Oy, 2013).

11
 -1050

-1450

Figure 2-2. Pyhäsalmi mine underground layout and deep ore dimensions (Numminen,
2012).

2.1 Geology

The bedrock in the area of the Pyhäsalmi mine is part of the Svecofennian domain
between the Central Finland Granitoid Complex in the southwest and the Archean
Basement Complex in the east. Structurally it belongs to a crossing zone of the Raahe-
Ladoga Zone trending NW and the Oulujärvi Shear Zone trending SE. In terms of
lithology it belongs to the Savo Schist Belt trending NW (Puustjärvi, 2006).

The Pyhäsalmi orebody is a typical Zn-Cu volcanogenic massive sulfide deposit situated
in the hinge of a large syncline, surrounded by volcanites and an alteration halo. The ore
extents from the surface (outcropping with an area of 650m length and 80m width) down
to 1410m depth (Figure 2-2). The deposit consists of massive ore containing 70% of
medium- to coarse-grained sulfides. Originally it had no fractures or schistocity. The
composition varies both horizontally and vertically. In some places the ore is finely
banded with common thin porphyritic bands. Sporadically inclusion of week and altered

12
rock lenses can be found in massive sulfides. The contact between the ore and waste rock
is sharp. Around the massive ore there is a pyrite dissemination characterized by
brecciated structure (Puustjärvi, 2006). The main minerals are: pyrite, chalcopyrite and
sphalerite, and the average grades are: 1.2% Cu, 2.5% Zn and 43% S (Numminen, 2012).
The host rocks are mostly felsic pyroclastic rocks and porphyries rich in quartz that were
hydrothermally altered, metamorphosed, deformed and recrystallized. Other rock types
are mafic volcanic rocks, mainly coarse-grained tuff breccias and lavas, and volcanic and
felsic dykes (Puustjärvi, 2006).

The schistocity in host rocks is clear and leads to development of cracks that are most
dominant in the schistocity direction. Although there are no distinctive faults in the active
mining area that could be activated by mining, pegmatite veins that are present near the
ore contact zone have been identified as fault planes and might be important (Oye & Roth,
2005).

2.2 Mining method

The mining method currently utilized in the Pyhäsalmi mine is non-entry, bulk open
stoping. The mining sequence is carried from the middle of the orebody (at the bottom)
and continues upwards. Typically around 25 stopes are mined each year, with up to 5 in
production at the same time. The dimensions of primary and secondary stopes, as well as
the mining sequence is given on Figure 2-3.

Figure 2-3. Stope size and mining sequence at the Pyhäsalmi mine. Mining sequence: 1)
Mine out stopes 4/1 and 6/1; 2) Consolidated backfill of stopes 4/1 and 6/1; 3) Mine out
stopes 4/2 and 6/2; 4) Consolidated backfill of stopes 4/2 and 6/2; 5) Mine out stope 5/1;
6) Rock filling 5/1; 7) Repeat sequence for next stope (Gleeson, 2010).

13
The size of stopes varies from 50 000t to 200 000t (in 2013 the average size of around
55 000t). Primary stopes are backfilled with hydraulic backfill and consolidated using
material consisting of coarse tailings sand, slaked lime and slag which is pumped
underground as a slurry. Secondary stopes are backfilled either with hydraulic backfill or
with rock fill from a nearby quarry (Gleeson, 2010). The mine operates five days per
week - on weekdays with morning and evening shift. Stope blasting is performed with
emulsion explosives. For safety reasons (mainly because of sulfuric fumes after the blast)
blasting time is at 10pm, when all the crew leaves the mine and no one can access the
mine until the next morning.

The mine uses automated LHDs to haul the ore; the operator fill the bucket and unit is
released for automated hauling. The system is flexible so that barriers can be repositioned
to seal off the automated area and LHDs can move from one stope to another. Pyhäsalmi
uses also automated drilling equipment with current capability for one-hole automation.
The ore is dumped to ore passes, fed to jaw crusher and hoisted to the surface where it is
send to the mill for floatation (Gleeson, 2010).

2.3 Rock mechanics and ground support

The Pyhäsalmi mine is characterized by high horizontal stress field, what has an influence
on the mine planning. The average major principal stress measured at the -1125 level was
65Mpa dipping at 5º towards 310ºE. In order to avoid large stress failures stopes are
designed to be parallel to this direction. Results of measured and estimated in-situ stresses
can be found in table below.

Table 2-1. Primary stress field (Bergström, 2014).

Level σV [MPa] σH [MPa] σh [MPa]
-1135 (measured) 33 65 41
-1400 (estimated) 41 75 45

The quality of rock is good. Values of uniaxial compressive strength and elastic modulus
for intact rock are shown in Table 2-2.

Table 2-2. Primary rock properties (Bergström, 2014).

Rock type UCS [MPa] E [GPa] ρ [kg/m3]
Ore (massive sulfide) 90-120 90-140 4400
Waste rock (volcanites) 200-240 60-80 2700

Primarily, the rockmass is competent and stiff; rock is considered as prone to rockbursts.
With advancing mining one fracture set (azimuth 47º and dip 70 º) in the massive pyrite

14
has become more apparent (Hakala, et al., 2013). The jointing in the waste rock is related
to schistocity which is most prominent close to the ore. One of the major problems in the
mine is detachment and slip of the contact zone wall rock. It was observed first in 2009
from damage in shotcrete. Observed block movement and slip along surfaces in yielded
rockmass (along the schistocity surface in the ore contact area) has led to further
fracturing of the orebody. Due to this, the stress field has changed and relaxation took
place. The horizontal stress in remaining pillars inside the ore has been released, and the
high stress is concentrated along the contact zone and around the orebody. After stress
redistribution increased seismicity was observed outside the orebody.

The monitoring of rockmass conditions and stability consist of following elements:

 Microseismic monitoring system,
 Extensometers for local movement investigation and SMART cable bolts for
determining actual support loads in critical areas (stope brows, active stopes
vicinity, ore contact zone and ore passes),
 Fixed survey points,
 HI cells for measurements of 3D stress changes in time,
 Visual damage mapping in production levels,
 Stress modelling for stope scheduling (Bergström, 2014).

The Hollow Inclusion (HI) cell is an instrument to measure the absolute triaxial stress in
rocks by overcoring and to monitor stress changes over time. It consists of an arrangement
of strain gauges mounted in the wall of a hollow tube with known Elastic modulus. The
cell is installed in a borehole and grouted.
There are few different types of rock mechanical reporting currently performed at the
mine site. First of all, is daily seismic report for production personnel summarizing the
seismic activity for last 24h and indicating if any area needs inspection by the shift
supervisor. Second type is a weekly report summarizing seismic activity, extensometers
reading and damage mapping results. Third type is a monthly report of seismic activity.

The ground support system in the mine consists of four elements:

 Rock bolts - fully grouted 2.2m length and 20mm diameter Kiruna bolts with
wedge and 120mm washers; the minimum bolt density is 5 bolts per 1m of
development,
 Mesh – welded mesh 125mm x 125mm, with 5.5mm wires,
 Shotcrete – 50-75mm layer thickness and 30kg/m3 of steel fibers,

15
  Cable bolts – 5-9m cables, 15.2mm diameter (Bergström, 2014).

Depending on the local conditions and support requirements the set-up may vary from
only bolting and shotcreting to a system with all four above-mentioned elements. For
details please see Figure A-1 and Figure A-2 in Appendix A.

2.4 Seismic monitoring system

The seismic monitoring system was installed in 2001 by the Integrated Seismic System
International Ltd. (ISS) and since then is focused on recording microseismic events. The
latest system upgrade was completed in September 2013 by the Institute of Mine
Seismology (IMS) who currently manages the system. Maintenance of the system and
support service is also done by IMS. After upgrading, the system consists of:

 seismic server located at the surface for data processing and storage,
 7 IMS station seismometers and IMS GS seismometers,
 24 uniaxial (1G) and 10 triaxial (3G) geophones (see Table 2-3) with natural
frequencies of 4.5Hz and 14Hz respectively (de Jongh, 2013).

The 4.5Hz geophones have operational frequency bandwidth between 3Hz and 2000Hz
and are installed within 2 degrees of their pre-set orientation with respect to the vertical.
The 14Hz geophones have operational frequency bandwidth between 8Hz and 2000Hz
and can be installed at any angle. Geophones are fixed in 10m long, vertical boreholes
with a 76mm diameter that were drilled in the roof of excavations (see Figure 2-4).

Figure 2-4. Schematic diagram (left) and picture of borehole geophone installed in the
mine.

16
Table 2-3. List of active geophones in the seismic monitoring system (de Jongh, 2013).
Geophones
Station
X [m] Y [m] Z [m] Type
2276 8307 -99 3G
2159 8413 -107 1G
IMS1 @ surface
2279 8307 -57 1G
2313 8353 -108 1G
2260 8431 -397 3G
IMS2 @ -400 Level
2301 8383 -396 3G
2383 8300 -1040 3G
2247 8110 -1070 1G
IMS3 @ -1075 Level 2232 8285 -1067 1G
2192 8407 -1062 1G
2446 8424 -1021 1G
2290 8273 -1133 3G
IMS4 @ -1150 Level 2423 8346 -1111 1G
2231 8476 -1130 3G
IMS5 @ -1200 Level 2409 8357 -1178 1G
2414 8439 -1250 1G
2205 8500 -1239 1G
2198 8310 -1264 1G
IMS6 @ -1250 Level
2297 8219 -1242 1G
2110 8300 -1228 1G
2393 8236 -1156 1G
2209 8147 -1290 3G
2216 8136 -1313 1G
IMS7 @ -1325 Level
2099 8335 -1314 3G
2305 8337 -1318 1G
2403 8297 -1326 3G
2282 8183 -1340 1G
IMS 8 @ -1350 Level 2269 8408 -1340 1G
2185 8422 -1337 1G
2441 8405 -1353 1G
2291 8203 -1402 3G
2260 8279 -1390 1G
IMS9 @ -1410 Level
2414 8392 -1398 1G
2436 8243 -1399 1G

Six geophones are places above -400 level for backfill raise cave monitoring. Other
seismic sensors are placed in production areas on levels -1075 to -1425. After the upgrade
in 2013 more geophones were installed on the hanging wall side of the ore body resulting
in better coverage. The total area monitored is equal to 524,500 m2 and the average inter
sensor spacing is equal to 124.2m.

Microseismic events are recorded 24h per day and are stored in the database. Special Info
center was established providing real time access from any computer (including
underground) in order to infer the time, location and magnitude of recorded events
(Bergström, 2013).

17
Sensitivity of the system (as for January 2014) is described by:

 The minimal recordable magnitude varies from -2.1 to 1.3 (depending on the
location; the best is achieved close to production areas, in the vicinity of installed
geophones).
 3D location error varies from 10m (close to the center of the mine) to 50m (see
Figure 2-5).

-1050 -1050

-1450 -1450

Figure 2-5. Seismic system sensitivity map – vertical section, view looking east, triangles
illustrate installed geophones (Meyer, 2014).

18
3 Theoretical background

This chapter presents theoretical background for the thesis. It is divided into two parts.
The first part describes the geotechnical risk assessment guideline for underground mines,
proposed by Mishra (2012). The second part defines basic concepts related to seismicity
in underground mines.

3.1 Geotechnical risk assessment guideline

The geotechnical risk assessment guideline is part of the I2Mine project and aims to be
used with geotechnical data from underground mines in order to identify and investigate
geotechnical hazards that might pose risks. The goal is to evaluate the likelihood of
geotechnical hazard to occur, and to assess its possible consequences. The geotechnical
risk can be defined as:

𝐺𝑒𝑜𝑡𝑒𝑐ℎ𝑛𝑖𝑐𝑎𝑙 𝑟𝑖𝑠𝑘 = 𝐿𝑖𝑘𝑒𝑙𝑖ℎ𝑜𝑜𝑑 𝑜𝑓 𝑔𝑒𝑜𝑡𝑒𝑐ℎ𝑛𝑖𝑐𝑎𝑙 ℎ𝑎𝑧𝑎𝑟𝑑 𝑜𝑐𝑐𝑢𝑟𝑒𝑛𝑐𝑒 ∙

𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 𝑜𝑓 𝑔𝑒𝑜𝑡𝑒𝑐ℎ𝑛𝑖𝑐𝑎𝑙 ℎ𝑎𝑧𝑎𝑟𝑑

The guideline consists of two main parts: geotechnical hazard potential and geotechnical
risk assessment. Next subsections present basic ideas of the guideline. For more details
please consult Mishra (2012).

3.1.1 Geotechnical hazard potential

According to the guideline, the first step is identification of the Geotechnical Hazard
Potential (GHP) by using proposed risk level classification method. The GHP aims to
help in estimation of possible geotechnical hazards, and is meant to be used as a
preliminary step for risk ranking and justification of the required Geotechnical Risk
Assessment. GHP is based on following sub classification:

a) mining method
b) rockmass – based on stability number Nr (from Matthews stability graph
method):

N𝑟 = Q’ ∙ A ∙ B ∙ C (1)

Where:

Q’ – the Barton’s Q number with Jw (joint water parameter) taken as 1,

19
 A – the influence of stress on the excavation and the rockmass – ignored at this stage;
taken into account in SM parameter:

𝑟𝑜𝑐𝑘 𝑚𝑎𝑠𝑠 𝑠𝑡𝑟𝑒𝑛𝑔𝑡ℎ (2)

Safety Margin (SM) = −1
𝑚𝑎𝑗𝑜𝑟 𝑝𝑟𝑖𝑛𝑐𝑖𝑝𝑎𝑙 𝑠𝑡𝑟𝑒𝑠𝑠
B – the influence of discontinuity and joint orientation to the excavation surface.
C – the impact of the orientation of the excavation plane itself and accounts for the
influence of gravity.

C = 8 − 7 ∙ Cos α (3)

α is the angle of excavation plane inclination

After calculation of abovementioned parameters the results are used to assign the
rockmass competency using Table 3-1 and Table 3-2. The classification code starting
with 3 represents mining in operations stage.

Table 3-1. Rockmass classification for risk, based on safety margin (Mishra, 2012).
Range of SM Classification code Rock Competency
-1 to -0.8 3I Very low
-0.8 to 0 3II Low
0 to 0.5 3III Fair
0.5 to 2 3IV High
>2 3V Very high

Table 3-2. Rockmass classification for risk, based on modified stability number (Mishra,
2012).
Range of Nr Classification code Rock competency
0.0001 to 0.6 3I Very low
0.6 to 7 3II Low
7 to 30 3III Fair
30 to 250 3IV High
> 250 3V Very high

The last step is to infer the GHP from Table 3-3 and to rank the areas under investigation,
such as different sections of the mine. This can assist in a better allocation of resources
to high risk areas in order to reduce it. Based on the results, formal geotechnical risk
assessment is justified if the GHP is 3 or higher.

20
Table 3-3. Geotechnical Hazard Potential classification. Codes: O - open stoping, 1 –
mine in pre-feasibility stage, 2 – mine in bankable feasibility stage, 3 – mine in planning
and operations stage, I to V – hazard from very high to very low (Mishra, 2012).

GHP Description Code

Very Negligible chances of hazards arising from bulk rockmass property. O1V
Hazards can largely arise from random natural events, unforeseen O2V
Low (1)
discontinuity. O3V
Minor chances of hazards arising from bulk rockmass property. This O1IV
Low can be in terms of minor raveling and spalling.
Hazards arising from random natural events, unforeseen discontinuity O2IV
(2)
and human error. The extent of damage from such random event is
O3IV
noticeable but doesn’t hamper routine mining activity.
Fair chances of hazards arising from bulk rockmass property. This can
O1III
be routine if the rockmass is not supported/reinforced.
Fair Hazards arising from random natural events, unforeseen discontinuity O2III
(3) and human error. The extent of damage from this can be higher than
the routine visible failures. This can cause substantial damage to O3III
production.
High frequency of hazards arising from bulk rockmass property.
O1II
Accidents cause productivity loss recovered over weeks.
High An unsupported site may not be safe for onsite risk assessment itself. O2II
(4) Hazards arising from random natural events, unforeseen
discontinuities and human error. Such hazards cause major damage to O3II
production. May lead to closure of area.
Very high frequency of hazards arising from bulk rockmass property.
Accidents cause loss in productivity which may not be recovered over O1I
the year.
Very Site for risk assessment must not be visited without reinforcement and O2I
High (5) couple of days of observation.
Hazards arising from random natural events, unforeseen
discontinuities and human error. Such hazard may cause permanent O3I
loss of raw material in the form of trapped ore.

3.1.2 Geotechnical risk assessment

The geotechnical risk assessment (GRA) procedure follows five main steps:

Step 1 – Outline the scope of risk assessment

The scope of GRA has to be defined in order to know what resources are needed and who
will be performing the assessment. The requirements have to be specified accurately so
that the available resources will be used in the best possible way. Furthermore, risk
assessment parameters have to be selected based on the amount of available data: the
more data is available the easier is to quantify risks, hence in operational stage

21
quantitative parameters are preferred to qualitative ones. However, current practice in
many mines, including the Pyhäsalmi mine is to use qualitative parameters, even in
operational stage. The reason is their simplicity of transferring knowledge and its
comprehension by the workforce. In addition, outlining the scope of GRA involves
selection of appropriate approach from deterministic, probabilistic or possibilistic. Mishra
(2012) proposes a numerical tool for selection of abovementioned approaches (see
chapter 4.2).

Step 2 – Identify hazards within the scope

Selection of identification tool depends on the scale and hazard scope of the assessment.
For hazard specific GRA for large scale, Bow Tie Analysis (BTA) is considered as the
best tool to identify all hazards that can result in a top event. Other tools for hazard
identification are: Failure Mode & Effect Analysis (FMEA) for site specific GRA for
large and small scale, Fault Tree Analysis (FTA) and Event Tree Analysis (ETA) for
hazard specific GRA for small scale.

Step 3 – Evaluate the likelihood of hazard

The goal is to find the probability (likelihood) of an event to occur. Specific procedure
depends on the approach selected within the scope. For deterministic approach, precise
calculation is performed with measured parameters. This approach lies in line with
modelling of hazardous situation. For probabilistic approach, the probability of hazard is
evaluated by accounting for variability and uncertainty in data. If the approach is
possibilistic, a correlation is established between influencing factors and failure by using
indicative variables and utilizing empirical methods.

Step 4 – Assess the consequences

This step aims at an assessment of damages due to realization of hazards. It involves

evaluation of the exposure of people and assets to a hazardous condition, and
determination whether a hazard is tolerable or not. It is common to establish a common
benchmark for evaluation of consequences to provide an easy way of classifying
consequences. Ideally, severity is to be evaluated in terms of lost money, however it may
be difficult to quantify all damage types as a financial loss.

Step 5 – Rank the risk to formulate a strategy of risk reduction

22
Final risk is a result of multiplication of likelihood and consequences. It can be
represented in graphical format or using a matrix format. The first suits best when both
the probability and consequences are evaluated quantitatively, hence the risk is
quantitative. The latter is used when probability or consequences are not quantitative, and
is the easiest format for risk representation and ranking.

3.2 Seismicity in mines

Seismic events in the mining environment represent the response of rockmass to mining
activities that are inducing stress change and energy release. Seismic events are
predominantly generated in the process of rockmass failure. Rockmass deformation can
be associated with a series of small and large events over a period of time, where each
event is a single element of a continuing failure process. Because of inhomogeneity in the
rockmass, such as dykes, pillars or major discontinuities, some energy released during
the mining process can be stored and ultimately lead to large scale rockmass failure
(Hudyma, 2008).

A number of parameters have a direct influence on the occurrence of seismic events.

Blake and Hedley (2001) mentioned four main parameters: rock properties, depth, aerial
extent of mining and production rate. Hudyma (2004) lists some additional influencing
factors: pre-mining stress and induced mining stress, stage of extraction and percentage
of orebody extraction.

Seismic source is defined as an area within a rockmass where deformation or failure

(resulting in seismic event) is caused by a combination of stress, geological structure and
mining. Seismic source mechanism is referred as the type of deformation or failure that
results in seismic stress wave creation (release of seismic energy). It indicates the
maximum size of an event that can happen and is related to the timing of energy release
(stress driven event happens directly after the mine blast; structurally driven are not
related to mine blasting time) (Hudyma, 2008).

3.2.1 Seismic event size

Seismic event size or its intensity can be measured with a magnitude scale, with a most
common example of Richter magnitude scale (also shortened as Richter magnitude)
proposed by Richter (1935). It is a logarithmic scale of peak ground motion measured at
a distance of 100km from seismic source. Although this scale is widely used for

23
measurement of earthquakes, the local scale of the seismic monitoring system installed in
mines may differ from site to site. That is why bigger events, registered also by regional
or national seismological networks, are used to calibrate the local magnitude scale to an
approximate Richter magnitude scale. Table 3-4 presents a comparison of Richter
magnitude scale and a qualitative description of subjective feeling of people in the mine,
as a result of studies in Australian and Canadian mines.

Table 3-4. Richter magnitude of seismic events compared with qualitative description of
how those events are felt in the mine (Hudyma & Potvin, 2004)
Approximate
Richter
Qualitative description
Magnitude,
ML
 Small bumps felt nearby. Typically only heard relatively close to the
source of the event.
-3  Normally following development blasts.
 Event may be audible but vibration likely too small to be felt.
 Undetectable by seismic monitoring system.
 Felt locally as thumps or bangs at a short distance from the event.
 May be felt more remote from the source of the event (i.e. more than
-2
100 meters away).
 May be detectable by a microseismic monitoring system.
 Often felt by many workers throughout the mine (i.e. hundreds of meters
away).
-1
 Similar vibration to a distant underground secondary blast.
 Will be detected by a microseismic monitoring system.
 Vibration felt and heard throughout the mine.
 Bump commonly felt on surface (hundreds of meters away), but may not
0 be audible.
 Vibration felt on surface similar to those generated by a development
round.
 Typically felt and heard very clearly on surface.
1
 Vibrations felt on surface similar to a major production blast.

 Vibration felt on surface is greater than large production blasts.

2
 Vibration detectable with regional earthquake monitoring systems.

Mining-induced seismic events of low magnitude (ML from -4 to 0) are considered

as microseismic events (Young, et al., 1992). Typically, they do not cause any observable,
structural damage in mine openings, but are an indication of damage or inelastic
deformation in a brittle rockmass (Falmagne, 2001). Microseismic events are the main
portion of events recorded every day by seismic monitoring systems in mines (more than
95%), but they represent only a fraction of total seismic energy released due to mining

24
(Young, et al., 1992). Characteristic seismic source mechanisms of microseismic events
are listed by Hudyma (2008):

 Intact brittle rock fracture,

 Coalescence of rockmass fractures (rock joints),
 Large stresses in stope abutments,
 Shearing, crushing and volumetric fracturing of mine pillars,
 Shear or rupture of lithological contacts.

Rockbursts are seismic events with a magnitude higher than 0 that are characterized by a
sudden and violent release of stored energy that can result in dynamic failure of the
rockmass. They can cause losses in production, human life, or in the worst case,
a complete mine closure (Young, et al., 1992). In comparison to mining-induced
seismicity, large seismic events can be triggered by fairly insignificant mining-related
energy change that results in large failure of previously instable rockmass. Rockburst can
be caused either by an implosional failure observed near mine openings, or by shear-
related failure that outspread on greater distance from mine excavations (Hudyma, 2008).
Ortlepp (1997) suggests five mechanisms of rockburst divided by their source
mechanism, first motion recorded by seismic sensors and an approximate Richter
magnitude (see Table 3-5).

Table 3-5. Mechanisms of damaging rockbursts (Ortlepp, 1997).

Postulated source First motion from Richter
Seismic event
mechanism seismic source magnitude ML
Superficial spalling with
Usually undetected,
Strain-burst violent ejection of -0.2 to 0
could be implosive
fragments
Outward explosion of large
Buckling slabs pre-existing parallel Implosive 0 to 1.5
to surface of opening
Face crush/pillar Violent explosion from Mostly implosive,
1.0 to 2.5
burst stope face or pillar sides complex
Violent propagation of
Shear rupture shear fracture through Double-couple shear 2.0 to 3.5
intact rockmass
Violent renewed movement
Fault-slip on existing fault or dyke Double-couple shear 2.5 to 5.0
contact

3.3 Seismic hazard

Seismic hazard is defined as “an estimation of the mean probability (over space and time)
of the occurrence of a seismic event with a certain magnitude within a given time

25
interval.” (Gibowicz & Kijko, 1994). Kaiser et al. (2005) separates seismic hazard into
two types: rockmass degradation due to microseismic events, and dynamic loading due
to seismic wave propagation.

Seismic hazard estimation is the key objective in seismic monitoring, however there is no
unique and general measure to quantify it. It is commonly assessed as the largest possible
event that can occur; event size is proportional to the level of ground movement induced,
which creates the potential for rockmass damage. It is important to mention that it varies
in space and time and is influenced by the location of maximum ore extraction and related
stress concentration, as well as period of mine blasting. Seismic hazard can be evaluated
as long, medium, and short-term. In contrast to earthquake seismology, thorough
statistical seismic analysis is of limited importance for mine personnel. From this
perspective, an analysis that can relate to rockmass failure mechanisms, hence coupling
seismicity to mining activities, is more beneficial. (Hudyma, 2008).

One of the methods for quantification of seismic hazard is the so called Seismic Hazard
Scale (SHS) developed by Hudyma (2004). The SHS uses three mine seismicity
parameters: the rate of occurrence of events of a certain magnitude, the power law relation
for mine seismicity (as it is shown in Equation (4) and the maximum observed event
magnitude. The power law relation was developed by Gutenberg and Richter, and is
recognized as a relation between the magnitude of seismic event and the frequency of
occurrence:

𝑙𝑜𝑔10 (𝑁) = 𝑎 − 𝑏 ∙ 𝑀𝐿 (4)

Where:

N – number of events of at least magnitude ML,

ML – approximate Richter magnitude,
a – constant; measure of the level of seismicity,
b – slope of the power law relation; describes the relative number of small and large
events in a certain time interval (Hudyma & Potvin, 2004).
The diagram illustrating the Gutenberg-Richter power law relation is called
Gutenberg-Richter frequency-magnitude diagram (F-M diagram) and is presented on
Figure 3-1.

26
Figure 3-1. Gutenberg-Richter frequency-magnitude power law relation (F-M diagram,
Eq. 4) for large population of seismic events. This is an example of good dataset with
linear relationship that follows the power law (from Hudyma (2008): Figure 2.11)

The SHS is the x-axis intercept of the frequency magnitude relation, assuming b equal
to 1:
𝑆𝐻𝑆 ≈ 𝑙𝑜𝑔10 (𝑁) + 𝑀𝐿 (5)

Table 3-6 presents a comparison of SHS with relative seismic hazard and the largest event
that can occur. The SHS has been used to examine seismic hazard for events up to ML of
3, because events of larger magnitudes may not follow the relation (b value less than 1).

Table 3-6. Comparison of SHS with relative seismic hazard and the largest event
magnitude (Hudyma & Potvin, 2010; Mikula, et al., 2008).

Approximate magnitude of
SHS Relative seismic hazard
largest expected event, ML
-2 Nil -2
-1 Very low -1
0 Low 0
1 Moderate 1
2 High 2
3 Very high 3

3.3.1 Quantitative Seismic Hazard and Risk Assessment Framework

As an example of current practice in seismic risk analysis that uses SHS for seismic
hazard quantification, one can mention the Quantitative Seismic Hazard and Risk
Assessment Framework (QSHRAF) approach applied in Mine Seismicity Risk

27
Assessment Program (MS-RAP), developed by the Australian Centre of Geomechanics
(Mikula, et al., 2008). The risk assessment process is summarized on Figure 3-2.

Figure 3-2. Risk assessment process applied within MS-RAP (from Mikula, et al., 2008:
Figure 30).

First, the seismic data is grouped into clusters of events representing a single seismic
source. The risk assessment starts from quantification of seismic hazard by calculating
SHS for each cluster. Next, seismic hazard map is created using the mine geometry and
mine plans, where SHS is assigned on each point on map.

Next, the Peak Particle Velocity (PPV) is calculated at each location under an assessment
using the maximum local magnitude that is expected to occur. PPV is scaled for distance
using following relation:
𝑀𝐿
10 2 (6)
𝑃𝑃𝑉 = 1.4 ∙
𝑟

PPV – Peak particle velocity at the location under assessment [m/s]:

ML – local magnitude of the largest expected event,
r – distance from the cluster (seismic event location) to the location under assessment.

28
In the next step, excavation vulnerability potential for damages (EVP, see Equation (7) is
assessed on the base of four parameters stress conditions (E1), utilized ground support
(E2), excavation span (E3) and geological structure (E4). The EVP represents the
increasing likelihood and severity of rockburst damage. It is calculated using following
formula:

𝐸1 𝐸3
𝐸𝑉𝑃 = ∙ (7)
𝐸2 𝐸4

Where:
E1 – Stress condition factor as a ratio of static stress to rockmass strength in the vicinity
100∙𝜎1𝑀
of excavation, 𝐸1 = [-],
𝑈𝐶𝑆

σ1M – the mining induced maximum stress at the place under assessment [MPa],
UCS – the intact unconfined compressive strength of the rock [MPa],
E2 – Ground support capacity to withstand dynamic loading (see Table 3-7),
E3 – Excavation span, taken as a diameter of a circle drawn within mine opening [m],
E4 – Geological Structure (see Table 3-8) (Mikula, et al., 2008; Hudyma & Potvin, 2010).

Table 3-7. Ground support capacity scale – E2 factor in Eq.7 (Potvin, 2009).
Surface E2
Classification Reinforcement Example
support rating
Spot bolting with split sets or
Spot bolting
Low None 2 solid bar bolts, minimal surface
(spacing > 1.5 m)
support
Pattern bolting Pattern bolting with split sets or
Mesh or
Moderate (spacing 1–1.5 5 solid bar reinforcement, with
fibrecrete
m) mesh or 50mm fibrecrete
Pattern bolting with split sets or
Pattern bolting solid bar reinforcement, with
Mesh or with a second mesh or 50mm fibrecrete. Plus
Extra bolting 8
fibrecrete bolting (overall an additional pass of pattern
spacing < 1m) reinforcement, such as solid bar
bolts
Pattern bolting with split sets or
Pattern bolting
High static Mesh or solid bar reinforcement, with
and pattern cable 10
strength fibrecrete mesh or 50mm fibrecrete. Plus
bolts
pattern cable bolting
Pattern bolting with dynamic
Very high Dynamic
Pattern dynamic ground reinforcement such as
dynamic surface 25
support cone bolts, with dynamic
capacity support
resistant surface support system

29
Table 3-8. Geological structure - E4 factor in Eq.6 (Potvin, 2009).
E4 Description

Seismically active major structure

Major structural features such as faults, shears or discrete contacts intersect the
0.5 location and act as a potential failure surface promoting rockmass failure.
Example: The rock mass fails back or along a major fault, increasing the depth of
failure considerably more than would otherwise occur in the rockmass.
Unfavorable rock mass/no major structure.
The orientation of the rockmass discontinuity fabric may promote or enhance
rockmass failure. Generally, this factor is applied when there are local cases in
1 which the rock mass discontinuities promoted falls of ground much larger than
would be expected.
Example: A heavily jointed, blocky rock mass with kinematically unstable
rockmass blocks. The rockmass is prone to deeper than normal gravity driven
failure mechanisms.
Massive rock mass/no major structure.
1.5 The rockmass is essentially massive, or non-persistent rock mass discontinuities
may exist; including possible minor blast related fracturing. There are no major
structures such as faults or shears, which may promote or enhance rockmass failure.

The rockburst damage potential (RDP) is found by multiplying EVP by PPV at a specific
location (see Eq. 8).

𝑅𝐷𝑃 = 𝐸𝑉𝑃 ∙ 𝑃𝑃𝑉 (8)

Mikula, et. al (2008) gives a scale of the extent of rockburst damage from R1 (no damage)
to R5 (complete destruction of the support system) that was found empirically by
investigating and back analyzing 254 cases of rockburst damage by Heal, et. al in 2006.
The empirical chart relating the largest predicted ground movement (PPV) to the
excavation susceptibility to damage (EVP) is presented on Figure 3-3. The rockburst
damage potential increases with increasing EVP and PPV. The scale of projected damage
determines the necessary dynamic capacity of installed ground support. Proper ground
reinforcement that can withstand dynamic loading and large deformations is required in
order to reduce the rockburst hazard and protect workers, mine infrastructure and sustain
safe operation (Kaiser & Cai, 2012). By increasing the capacity of support, for example
by installation of yielding support and elimination of the weakest link, the rockburst
damage potential can be lowered (on Figure 3-3 point will move left, to lower rockburst
damage scale zone).

30
Figure 3-3. EVP vs PPV diagram showing zones with expected damage (from Mikula, et
al., 2008: Figure 38).

Below, Table 3-9 presents the scale of RDP with qualitative description of expected
rockmass and support damage.

Table 3-9. Rockburst damage potential scale showing expected rockmass and support
damage (redrawn from Mikula, et al., 2008).
Rockburst
Expected rockmass
RDP Damage Expected support damage
damage
Scale
0 to 25 R1 No damage / minor loose No damage
25 to 50 R1 No damage / minor loose No damage
Minor damage / less than 1t Support system is loaded, loose
50 to 130 R2
displaced mesh, plates deformed
130 to 170 R3 1 to 10t displaced Some broken bolts
170 to 230 R4 10 to 100t displaced Major damage to support system
Complete failure of support
230 to 280 R5 100+ t displaced
system

The final seismic risk is calculated by taking into account exposure of personnel.
Exposure is a function of the amount of time spent performing a number of tasks by mine
personnel, the level of protection and the number of people involved. Areas are assigned
with exposure ratings that were proposed by Owen (2004) after studies conducted at
underground mines.

31
The seismic risk ratings (SRR) are qualitative: VL = Very Low, L = Low, M = Moderate,
H = High, VH = Very High, E = Extreme, and depend on the type of mining activity
(quantitative exposure rating) and the level of rockburst damage potential. Graphical
representation of risk matrix is demonstrated in Table 3-10

Table 3-10. Seismic risk assessment matrix constructed from quantitative exposure rating
(vertically) and rockburst damage potential (horizontally). The seismic risk ratings are
qualitative: VL = Very Low, L = Low, M = Moderate, H = High, VH = Very High, E =
Extreme (Mikula, et al., 2008).
RDP=EVP 25- 65- 115- 170- 225-
<25 >280
x PPV 65 115 170 225 280
Excavation type/ Exposure
activity rating
Restricted access (no
entry) 100 VL VL L M M H H
Decline 1000 VL L M M H VH VH
Travelway - no active
mining 1000 VL L M M H VH VH
Travelway - mining on
the level 2000 VL L M M H VH VH
Production mucking
area 3000 VL L M M H VH VH
Busy level / travelway
drive / access 4000 VL L M M H VH VH
Development mining 7000 L M M H VH VH E
Production drilling 10000 M H H H VH E E
Production charge-up 10000 M H H H VH E E
Infrastructure areas /
workshops 14000 M H VH VH E E E

3.3.2 Probabilistic Approach for Seismic Risk Assessment

The Probabilistic approach for Seismic Risk Assessment (PASRA) has been developed
under the cooperation of Wen (2013). The first attempt to apply this approach was
performed by Froehlich (2014). It is based on calculation of two factors. The first is the
Released Energy Capacity (REC), as a relationship between the elevated energy due to
seismic event and accumulated energy (induced stress) in the area of investigation.
Second factor is the Absorbed Energy Capacity (AEC) based on the capacity of installed
ground support. Seismic hazard can be evaluated by dividing REC by AEC factor. By
applying it to sufficient data it is possible to estimate the likelihood of seismic hazard and
to assign it to different areas.

The PASRA methodology uses the same concept as QSHRAF to cluster the data into
logical groups representing a single seismic source in order to predict the maximum
possible seismic event within a group. The F-M diagram is used to find the maximum

32
size of an event within a population as x-axis intercept. This event size is used then to
calculate the maximum seismic energy of an event using following relationship:

𝑙𝑜𝑔 𝐸1 = 𝑎 ∙ 𝑀𝐿 + 𝑏 (9)

Where:
E1 – predicted largest expected seismic event energy,
ML - local magnitude,
a, b – constants.

The Released Energy Capacity (REC) formula links the largest predicted energy with the
accumulated stress energy in the area under assessment:

𝐸1 ∆𝐸 + 𝐸0 𝜎𝑢𝑛𝑖𝑎𝑥𝑖𝑎𝑙 ∆𝐸 𝐸1 ∙ 𝑒 −𝜑∙𝑟
𝑅𝐸𝐶 = = ∙ = (1 + ) ∙ 𝑘𝜎 = (1 + ) ∙ 𝑘𝜎 (10)
𝐸2 𝐸0 𝜎𝑚𝑎𝑥 𝐸0 𝐸0

Where:
ΔE – Evaluated energy condition due to seismic events based on the seismic wave
propagation at distance r from the seismic source in visco-elastic homogenous medium;
after Sambuelli (2009),
E0 – Accumulated energy in the investigated area representing the stress state in the area
under investigation,
E1 – Predicted largest expected seismic event energy calculated from the maximum
magnitude within a cluster group using Equation (9),
φ – Attenuation coefficient,
𝜎𝑚𝑎𝑥
kσ – Relation ,
𝑈𝐶𝑆

r – Distance to seismic cluster.

The Absorbed Energy Capacity (AEC) reflects the ground support capacity to withstand
and absorb the energy in the failure process following a significant seismic event.
Originally, the AEC equation aims at quantification of the support load capacity per
square meter in order to be comparable to REC. Here, due to the limitation of available
data, AEC is represented by a rating scale dependent on installed support. The higher the
capacity of the support to withstand dynamic damage, the larger the AEC. Seismic hazard
is evaluated by dividing the REC by AEC:

𝑅𝐸𝐶
𝑆𝑒𝑖𝑠𝑚𝑖𝑐 ℎ𝑎𝑧𝑎𝑟𝑑 = 100 ∙ (11)
𝐴𝐸𝐶

33
3.4 Data clustering
Clustering is an unsupervised technique to classify data into groups (clusters) based on
similarity. Data points contained within a valid cluster are more similar to each other,
compared to data points outside the cluster (Jain, et al., 1999). Clusters are defined as
connected regions of space containing large amount of points, opposed to regions with
less dense data. The main goals of clustering are: to reduce the amount of data in order to
allow for faster processing of information in representative groups of the entire data set,
to generate hypothesis about the data and to predict some features of the data within
specified clusters to extract useful knowledge. Clustering process can be divided into few
steps: feature selection, algorithm selection, results validation and interpretation (see
Figure 3-4).

Figure 3-4. Steps of clustering process (from Halkidi et al. (2001): Figure 1).

The first step is to select the features on which we perform the clustering (for example
spatial coordinates) depending on the type of data and the purpose of our analysis. The
second step is to select an algorithm that will result in good cluster organization.
Clustering algorithms are characterized by proximity measure and by clustering criterion.
Proximity measure defines how similar are two data points. (Halkidi, et al., 2001). One
of the most common proximity measures in data clustering is Euclidean distance that is
used to calculate the dissimilarity between two data points. In three-dimensional space
the Euclidean distance is defined as the length of the path between two points (x1, y1, z1
and x2, y2, z2) (Weisstein, 2012). It is calculated as follows:

34
 𝑑 = √(𝑥2 − 𝑥1 )2 + (𝑦2 − 𝑦1 )2 + (𝑧2 − 𝑧1 )2 (12)

Clustering criterion defines the type of function used for clustering and has an influence
on the type of achieved clusters. It has to be selected in such a manner that will fit well
the data set. The next step is to validate the results by utilizing suitable measures and
criteria in order to investigate if the algorithm was appropriate. Last step is to interpret
the results in order to draw conclusions about the data (Halkidi, et al., 2001).

3.4.1 Quality Threshold clustering algorithm

The Quality Threshold Clustering (QTCLUST) is a computationally efficient algorithm
originally developed for gene clustering that groups data into cluster of high quality
(Heyer, et al., 1999). Although it is a partitioning clustering method, it does not require
to specify the number of clusters beforehand, what is the biggest advantage compared to
other divisive methods (for example K-means clustering). The QTCLUST creates non-
overlapping clusters with some points remaining outside of clusters considered as
outliers. The distance from a point to a group of points is calculated as the maximum
length from any point from the group to the point (complete linkage using the Euclidean
distance – as in Equation (12). The QTCLUST algorithm has two parametrical
constraints: the minimum number of data points to be considered as a cluster (minimum
cluster size), and the maximum allowable radius of a cluster (measured from the centroid).
This ensures that the diameter does not exceed user-defined threshold and provides high
quality and compactness of created clusters (Bednarik & Kovacs, 2012). The procedure
of QTCLUST algorithm can be summarized with following steps:

 Create a potential cluster for random point by iteratively inserting the closest
points to the already existing group.
 Repeat step one until the size of cluster is larger than the threshold.
 Save the most populated potential cluster as the first true cluster.
 Remove the already clustered points from the data and repeat the procedure with
the reduced set of points.

The QTCLUST can be used in grouping of seismic events to break the data into compact
clusters. This can speed up (in terms of significant compression of data) and improve the
efficiency of analysis. (Kaiser, et al., 2005).

35
3.4.2 Single linkage hierarchical clustering algorithm
The Single linkage hierarchical clustering algorithm (SLINK), also referred as nearest
neighbor algorithm, is an agglomerative hierarchical algorithm that groups data by joining
similar clusters. At first, each data point is allocated to its own cluster. Then the two most
similar clusters are joined iteratively until there is only a single cluster. The distance
between two clusters is the minimum of all distances between cluster pairs. The SLINK
algorithm produces relatively few, elongated and chain-like clusters. The algorithm yields
a dendrogram that represent different levels of cluster grouping, that can be cut at different
heights to find the final clustering result (Jain, et al., 1999).

One of the advantages of SLINK algorithm is its simplicity and versatility. Hudyma
(2008) illustrated a successful application of SLINK algorithm in grouping of smaller
clusters of seismic data in order to distinguish single seismic sources from large data sets
of seismic events.

4 Geotechnical risk assessment in the Pyhäsalmi mine

The current practice of risk assessment regarding geotechnical hazards in the Pyhäsalmi
mine focuses mainly on two aspects. One is an evaluation of risks related to stoping. Mine
personnel keeps record of all the stopes in a special log summarizing data prior, during
and after mining. The data includes: area of the mine, development condition, mining
phase, past seismicity record, geological structures and important features, risk from
adjacent backfilling, and presence of orebody contact. The risk assessment approach is
based on a point-based ranking with a scale from small risk (rank I) to very high risk (rank
V). The stope log also contains information about expected stope performance and
procedures in case of higher risk. After mining each stope is evaluated and used for back
analysis (Pyhäsalmi Mine Oy, 2013). The second geotechnical assessment is a valuation
of risks and effects of ore subsidence and movements along the northern contact zone.
The mine utilizes monitoring instrumentation consisting of multi-point borehole
extensometers to monitor displacements around contact zone and smart cables to monitor
draw points brows near contact zones. Identified risks are: fall of ground in development
headings adjacent to contact zone, increased cost of ground support rehabilitation,
increased damage along the contact zone, large failures in stopes and increased
monitoring costs. (Bergström, 2012)

36
Due to recent increase of seismic activity, as an evidence deterioration of rockmass,
seismic risk is considered as high and large events can be expected. Monitoring of
rockmass condition principally relies on seismic system installed at the mine. Mine
personnel has a goal for further understanding of seismic data that requires performing
seismic analysis (Bergström, 2014). This analysis can help in estimating the seismic
hazard, especially to evaluate if larger events can have a negative influence, and how they
can be managed. This proofs a necessity for development and implementation of seismic
risk assessment approach. It also confirms that geotechnical risk assessment focusing on
seismic risk clearly fits into the Pyhäsalmi mine strategy to extend the knowledge of this
phenomenon.

Chapter 4.1 presents and attempt to evaluate the GHP using data from the Pyhäsalmi
mine. Chapter 4.2 describes the selection process of risk assessment approach. Chapter
4.3 presents hazard identification for seismic risk by the use of bow-tie analysis.

4.1 Geotechnical hazard potential evaluation

The assessment of the GHP has been performed according to the procedure described in
chapter 3.1.1. Evaluation was executed for the massive sulfide (ore) and volcanites (waste
rock). Following assumptions were taken into account:

 Open stoping is selected as mining method.

 Q’ value has been used in design modelling in the past (not up to date).
 Ore has one dominant fracture set with dip equal to 70°.
 Waste rock has foliation parallel to the ore contact, dip varies from 0º to 90º, here
20º is taken as a worst case scenario, because it is kinematically worst and gives
the lowest B factor (equal to 0.3) in calculation of stability number N r from
Equation (1).
 Excavation planes have an inclination of 0° and 90° for roof and wall respectively.
 Rockmass strength is taken from the inputs of the stress model; Hoek-Brown
failure criterion.
 Major principal stress is taken as an estimate for the -1400 level (see chapter 2.3).

First, the modified stability number was evaluated based on Equation (1) from chapter
3.1.1. The results can be found in Table 4-1.

37
Table 4-1. Results of the modified stability number for massive sulfide and volcanites.
ore waste rock
Parameter roof wall roof wall
Q' 100 100 35 35
B 0.8 0.3 0.3 0.3
C 1 8 1 8
Nr 80 240 10.5 84

Second, the safety margin was calculated using Equation (2) from chapter 3.1.1. The
results can be found in Table 4-2.

Table 4-2. Results of the safety margin for massive sulfide and volcanites.
Parameter Massive sulfide Volcanites
In-situ rock strength [MPa] 71 76
Major principal stress [MPa] 30 75
Safety Margin (SM) 1.4 0

As can be seen in Table 4-1 and Table 4-2, the lowest values of both the stability number
and safety margin are found for volcanites. The summarized results for massive sulfide
and for volcanites are presented in tables below. Based on those results, the rockmass
competency has been assigned according to Table 3-1 and Table 3-2 (see chapter 3.1.1).

Table 4-3. Results of rockmass classification for massive sulfide; risk in operation stage.
Parameter Value Rockmass competency Code
Nr 80 High 3IV
SM 1.4 High 3IV

Table 4-4. Results of rockmass classification for volcanites; risk in operation stage.
Parameter Value Rockmass competency Code
Nr 10.5 Fair 3III
SM 0 Low 3II

The results given in Table 4-4 and Table 4-3 indicate that the rock competency is high
(3IV category) and low (3II category) for massive sulfide and volcanites respectively.
Taking into account the mining method (open stoping) and rockmass competency (high
competency – code O3IV; low competency – code O3II) the classification presented in
Table 3-3 (see chapter 3.1.1) indicates that the Geotechnical Hazard Potential is low (2)
for massive sulfide and high (4) for volcanites. When the GHP is high, hazards related to
the rockmass properties are expected to arise from random natural events and to be of
high frequency. Such hazards may cause major damage leading to loss of production and
financial damage. This confirms the conditions and challenges that are currently

38
confronted in the mine. Given results indicate that the formal risk assessment is justified
for the volcanites, and not for the massive sulfide.

4.2 Risk assessment tool selection

The approach for geotechnical risk assessment is selected based on the methodology
proposed by Mishra (2012). In the selection process, a tool presented on Figure 4-1 is
used to find the most appropriate assessment approach. It is based on several GRA scope
categories, each of which has an assigned amount of points for three approaches:
deterministic, possibilistic and probabilistic. The amount of points for each category has
been proposed by Mishra and is not changed here. At the end, all the points are
summarized and the approach which scores the most is selected for further analysis.
Available alternatives of the GRA categories, as well as selected ones are described
below.

Type – the three possible types of GRA are: proactive, reactive and change
implementation. The proactive GRA is performed in advance of a planned operation and
is mostly done in the planning stage, when physical access to assess hazard is not possible.
The reactive GRA assesses a site with possible physical access, mostly after an event has
already occurred and when monitoring of the site is being carried out. Reactive GRA is
further subdivided into: routine reactive GRA concerning confirmation of risks expected
to happen and identification of new unexpected risks, and symptom based reactive GRA
that is commenced when a failure symptom has already occurred. The change
implementation GRA is performed when a change in existing mine design or operation
takes place and alternatives of different risk scenarios have to be assessed. In this study
GRA is aimed for quantification of expected seismic hazard, based on the analysis of data
from seismic monitoring system, so it can be considered as reactive – routine.

Area scale – the risk assessment can be performed on small (local) scale by investigation
of local geology, excavation geometry and reinforcement pattern, or on large scale, where
GRA focuses on rock mechanics of the entire area of a mine, applied mining method and
sequence and other large scale operations. In this study the risk will be assessed on large
scale, because large area of the mine (most hazard prone mining levels) will be evaluated
(see chapter 5.2).

Hazard scope – GRA can be: hazard specific when carried out to evaluate a single major
hazard in a mine, or site specific when assessing the entire area, where all potential hazard

39
are taken into account. Main focus of this study will be put on evaluation of risks related
with seismic hazard, therefore scope is considered as hazard specific.

Reporting requirements – GRA can be performed as part of internal reporting within a

company, or as part of external reporting to confirm that the operational risks are handled
according to legal standards. Although this study aims to investigate the applicability of
the proposed geotechnical guideline as a part of the I2Mine project, the results of GRA
will be used within the company (Pyhäsalmi mine), hence GRA can be considered as
internal.

Resources availability (data, equipment and personnel) – GRA can be performed with
resources availability considered as: high, when sufficient personnel, equipment (such as
monitoring system) and data is available to assess the likelihood of hazard, average when
not sufficient geotechnical data is available, but historical data of geotechnical hazards
exists, and finally low when data and personnel are limited and on-site instrumentation is
not available. In this study, despite the fact that only one person is carrying out the
investigation (externally) with a limited access and within limited time frame, there is a
sufficient amount of geotechnical monitoring data available in the mine (especially
largely abundant seismic monitoring data), hence the available resources are selected as
high.

Results of the selection, together with assigned points has been displayed on Figure 4-1.
The steps involved in using the tool are as follows:

Step 1 – Fill the checklist for the boxes on the right for each of the five GRA scope
categories. For example in the completed form below, the GRA is identified to be
“routine” (shaded in green). The amount of points for deterministic (De), probabilistic
(Pr) and possibilistic (Po) is transferred to the “GRA Type” box (following red arrows).

Step 2 – Repeat step 1 for the remaining four categories on the left.

Step 3 – Sum up the De, Pr and Po values for the boxes on the left and input them in the
“Total” box. The approach with the highest sum of points is the preferred risk assessment
approach.

The results of tool selection suggest the probabilistic approach (with a score of 9) to be
implemented in the geotechnical risk assessment procedure.

40
Figure 4-1. Results of the geotechnical risk assessment approach selection.

41
4.3 Hazard identification

The selection of hazard identification tool depends on the Geotechnical Risk Assessment
scale and hazard scope. If the risk assessment is to be performed on large area and with
hazard specific scope, the most appropriate hazard identification tool is the bow tie
analysis (BTA) (Mishra, 2012). BTA offers a simple approach to identify hazards and
potential threats that can result in an undesirable event with several adverse consequences.
BTA analysis consists of identification of hazards, threats and consequences related to
the top event. Hazards are considered as unsafe situations that can potentially result in the
top event by realization of specific mechanisms (threats). Table 4-5 present BTA for large
seismic event as a top event (major hazard) leading to damage of mine openings.

Table 4-5. Bow Tie Analysis for large seismic event.

Hazards Threats Top event Consequences
Inadequate ground Fatalities and injuries of
support Support load workforce
exceeds capacity Large Damage of infrastructure and
Inadequate mine design
seismic machinery
Active geological
Violent release of event Production interruption
structure (fault)
accumulated leading to
Accumulation of energy
energy damage of Increased cost
in rockmass
Large extent of ore mine
openings Ore loss
extraction Stress exceeds
High stress rockmass strength Reputation damage &
concentration community concern

To prevent or reduce consequences of the top event two types of control measures can be
utilized: pre-accident and post-accident. The first type of measures is used to prevent the
top event from happening. The latter is used to minimize the impact of an event after it
has already occurred. The goal of the GRA is to find the most appropriate control measure
that can result in effective risk reduction. For example if the ground support type or
capacity is not sufficient to withstand large movement of rock, changes in design have to
be made to prevent any damages to the tunnel profile. On the other hand, special
emergency chambers can be installed in case the damage happens to facilitate rescue of
mine personnel.

It is important to mention that hazards presented here should not be confused with seismic
hazard definition presented in chapter 3.3. The approach presented in this study evaluates
seismic hazard by taking into account multiple hazardous situations and threats (presented
in Table 4-5) as the influencing parameters.

42
5 Seismic Risk Assessment
This chapter presents an attempt to assess seismic risk in the Pyhäsalmi mine using the
two approaches described in chapters 3.3.1 and 3.3.2. The first subsection gives an
overview of the seismic data retrieved from the database in the mine. The data is first
analyzed statistically and quantified with methods used as current worldwide practice of
seismic risk assessment in underground mines. Data is also clustered in order to reduce
the amount of data, and allow for faster analysis. The second subparagraph describes the
selection of mining levels for further analysis. The third subsection presents the results of
seismic hazard assessment using Probabilistic Approach for Seismic Risk Assessment
(PASRA) method and the fourth subsection demonstrates calculation of seismic risk
using The Quantitative Seismic Hazard and Risk Assessment Framework (QSHRAF)
approach. The last subsection presents the final representation of seismic risk.

5.1 Seismic data analysis

Seismic events have been retrieved from the seismic database in the Pyhäsalmi mine. This
study examines events recorded in the period from 02-11-2002 to 27-02-2014. The total
number of events amounts to 206 157. Each record in the database is characterized by
several parameters: date, time, location in local coordinates system, location error,
seismic moment and seismic energy (for P- and S-wave), static and dynamic stress drop,
source size radius, and local magnitude.

The quality of data depends primarily on the process of waveform transformation and
used parameters. Investigation of the data quality is not part of this thesis and the data is
accepted as given. However, one abnormality has been detected in the location of seismic
events, so that 141 events were located above the ground surface. This might suggest
mistakes in data processing and using wrong parameters. To avoid any further biases
those events were deleted.

5.1.1 Events statistics

This subsection presents statistics of seismic events. Table 5-1 contains mean, median,
minimum and maximum of some seismic event parameters. As can be seen the minimum
local magnitude recorded by the seismic system amounted to -4.1 and the maximum to
2.2, with an average of -1.6.
43
The apparent stress is a measure of stress change at the seismic source. Depending on the
type of seismic source it can vary considerably; regions of high stress typically are a
source of high energy and high apparent stress. On contrary, low apparent stress events
take place in low-stress or highly fractured regions (Hudyma, 2008; Gibowicz & Kijko,
1994). The energy of seismic events varies from very low value of 6.82∙10-5 J to as high
as 1.99∙107 J, with a mean of 9.71∙102 J.

Table 5-1. Basic statistics of seismic event parameters.

Seismic Apparent Local

Location
X [m] Y [m] Z [m] Energy Stress magnitude
error [m]
[J] [bar] ML
Mean 8312 2280 -1245 5.0 9.71E+02 2.85E+00 -1.6
Median 8323 2297 -1268 5.0 5.29E+00 1.00E-01 -1.6
Min 8000 2000 -1499 0.0 6.82E-05 0.00E+00 -4.1
Max 8600 2599 0 253.0 1.99E+07 1.63E+04 2.2

As can be seen on Figure 5-1, the prevailing local magnitude range is from -2 to -1. Most
of the events have a magnitude between -3 and 0. Only 1% of events have a local
magnitude larger or equal than 0.

140000 132938

120000
Number of events

100000
80000
60000 43165
40000 28577
20000
201 1226 49 1
0

Magnitude range

Figure 5-1. Number of seismic events subdivided by the magnitude.

Seismic events with local magnitude higher than or equal to 0 have been plotted on Figure
5-2. It can be seen that the number of larger events increased since the beginning of
recording seismic activity, however in 2013 we can see a drop in large events. Number
for the year 2014 is projected based on first months, but the increase is visible.

44
 350

300

Number of events
250

200

150

100

50

0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014*
Year

Figure 5-2. Number of seismic events per year greater than or equal to magnitude 0;
*projected based on first two months.

Figure 5-3 shows daily histogram of seismic events. It is clearly visible that the number
of events per day increased. Moreover, in the past two years more extreme values start to
appear, with the largest number (2071 events per day) on 11-01-2014, and the second
largest frequency (1461 events per day) on 16-02-2014. One of the explanations of
increasing seismic activity is an intensification of movement of the northern ore contact
zone and an increase in subsidence rate. As it has been observed by Bergstrom (2012),
displacement increases were correlated with increased seismicity. Important fact to
mention is that the seismic system upgrade in September 2013 provided additional
geophones, and this results in higher sensitivity of the system, hence more events can be
recorded.

45
Figure 5-3. Daily histogram of seismic events for the period from 25-11-2002 to 27-02-
2014.

To investigate if there is a significant link between the time of day and occurrence of
seismic events, all events were grouped by the hour of day in which they took place.
Resulting diurnal chart is plotted on Figure 5-4.

As can be seen, considerable amount of seismic events occurred in the 22nd hour of day,
in comparison to other hours. It is an indication that there is a direct relation with blasting
time in the mine (at 22:00). This means that large portion of events can be considered as
stress driven.

18000
16000
14000
Number of events

12000
10000
8000
6000
4000
2000
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour of day
-4 ≤ MAG <-3 -3 ≤ MAG <-2 -2 ≤ MAG <-1 -1 ≤ MAG <0
0 ≤ MAG <1 1 ≤ MAG <2 2 ≤ MAG <3

Figure 5-4. Diurnal chart showing the distribution of seismic events by the hour of day;
subdivided by the magnitude.

46
5.1.2 Seismic events clustering
The main idea behind clustering of seismic events is to compress the amount of thousands
event into smaller groups to increase the efficiency of investigation. The assumption used
is that a single cluster (or cluster group) represents a single seismic source. Resulting
clusters (or cluster groups) can be used for the evaluation of maximum possible event that
can occur as a measure of seismic hazard.

The process of seismic events clustering is divided into three stages (see Figure 5-5).

Select & filter Cluster Group

clusters

• select spatial (XYZ) • QTCLUST algorithm to • SLINK algorithm to

and temporal range create clusters group the clusters
• filter out events with • Clusters validation • Selection of final
high location error and and selection cluster groups
outliers

Figure 5-5. Flowchart presenting three stages of clustering process.

In the first stage, the amount of seismic monitoring data is reduced using spatial and
temporal constraints in order to allow for trouble-free and faster analysis:

 Data is selected from the last 3 years of monitoring (from 01.01.2011 to

27.02.2014) – to limit the amount of data for processing, mainly because of high
computational demand of the clustering process. Analyzing more than 200000
events can be very slow, problematic and almost certainly unnecessary (Hudyma,
2014).
 Data is selected from inside the range for Y: from 2000 to 2600, for X: from 8000
to 8600, for Z: from -1500 to -900) – as events located close to the production
areas that have influence on the risk assessment process.

Next, data is filtered to select only good quality data:

 Reject events with high location error (above 95% percentile; ≥ 8m) to eliminate
the events that could bias the results.

47
  Reject outliers using density based approach (using isolation distance) – events
that do not have at least one neighbor within a sphere of a 20m radius around them
are rejected.

In the second stage, the Quality Threshold Clustering algorithm (QTCLUST) is used to
create compact clusters of seismic events. The main reason to use QTCLUST in this stage
is its computational efficiency compared to hierarchical clustering methods (like SLINK)
that are impossible to use with given amount of data.

Clustering routine was performed using an open source programming language and
environment for statistical computation – R, which is available as a free software. The
software has integrated collection of statistical analysis tools including the two clustering
algorithms applied in this study.

In order to find optimal values for two parameters of the QTCLUST algorithm a
sensitivity analysis has been performed. The goal is to maximize the number of seismic
event and the amount of seismic energy within the clusters, and at the same time keep a
reasonable number of created clusters. The cluster radius should be set up to be as low as
possible and at maximum equal to the sub-level spacing of production levels (25m in the
Pyhäsalmi mine). Small radius tends to ensure that seismic events within a cluster are
from a single seismic source, but can drastically increase computation time and the total
number of clusters. By decreasing the cluster radius and increasing the minimum number
of events (required to be classified as a cluster), the total number of events, as well as the
sum of the seismic energy being clustered is decreasing. A good clustering routine should
include at least 90% of total seismic energy and more than 75% of total number of seismic
events (Hudyma, 2008).

During the sensitivity analysis the cluster radius was differentiated between 20, 25 and
30m. The minimum number of events was varied from 10 to 40. Results are presented in
Table 5-2.

It can be seen that for a cluster radius of 20m results are rather poor in terms of the total
seismic energy within the clusters. For the minimum cluster size of 10 events the sum of
energy is only 82% of the total, what is below the minimum 90% required for good
clustering. Moreover, the number of QT clusters (901) is large and can be problematic to
handle in the next stage. The most efficient is cluster radius of 30m, but as it was stated

48
before, it exceeds the sublevel distance in the mine, and hence it is not appropriate to use
it. Ultimately, cluster radius of 25m is selected together with the minimum size of 15
events (highlighted in red) that gives good results in both the total number of events
clustered (96% of total) and seismic energy (90% of total).

Table 5-2. Sensitivity analysis of QTCLUST algorithm parameters.

Cluster
Min. Number of
radius Nr of events clustered Sum of energy [J]
events clusters
[m]
10 98450 95% 3.90E+07 82% 901
15 97340 94% 3.80E+07 80% 861
20 94908 91% 3.37E+07 71% 720
20 25 92304 89% 3.01E+07 64% 630
30 90872 87% 2.88E+07 61% 522
35 89379 86% 2.89E+07 61% 495
40 87429 84% 2.60E+07 55% 430
10 101826 98% 4.30E+07 91% 791
15 100071 96% 4.27E+07 90% 646
20 98289 95% 3.95E+07 84% 541
25 25 96926 93% 3.84E+07 81% 479
30 95757 92% 3.47E+07 73% 436
35 94577 91% 3.36E+07 71% 399
40 93030 90% 3.16E+07 67% 352
10 101541 98% 4.34E+07 92% 488
15 101572 98% 4.43E+07 94% 492
20 100405 97% 4.24E+07 90% 425
30 25 99471 96% 4.31E+07 91% 380
30 98707 95% 4.06E+07 86% 351
35 97675 94% 3.82E+07 81% 320
40 96819 93% 3.81E+07 81% 297

The results of QTCLUST are 646 QT clusters created from 100071 seismic events.
Figure 5-6 and Figure 5-7 illustrate all seismic events that were clustered, omitting the
outliers, with different coloring for different QT clusters. It can be seen that it is rather
difficult to analyze the results based on spatial distribution of clusters and further
grouping of cluster is inevitable.

49
 N

Figure 5-6. Plot of 646 QT clusters created in the second stage of clustering using
QTCLUST algorithm; plotted in the X-Y space. Colors illustrate different QT clusters.

50
 N

Y
Figure 5-7. Plot of 646 QT clusters created in the second stage of clustering using
QTCLUST algorithm; plotted in the Y-Z space, view looking east. Colors illustrate
different QT clusters.

The QT clusters were checked for validity using the F-M diagram drawn from the
population of seismic events within clusters. Graphs were plotted using range of local
magnitudes. The a and b parameters (from Equation (4) for each cluster were found using
linear regression (the least squares method). Ideally, the dataset would be expected to
follow linear relation for 2 or more orders of magnitude and have a slope of about -1
(Hudyma, 2008).

51
Figure 5-8 illustrates F-M diagram for cluster QT121. As can be seen, the linear relation
is preserved for magnitudes between -2 and -0.25, what is close to ideal. The slope of the
curve is -1, what represents the power law relation. Not all clusters showed such a good
fitting of F-M diagram and both the slope and the range of linear relation is not ideal.
Figure 5-9 illustrates F-M diagram of cluster QT84 which linear relation is only followed
for one order of magnitude and slope is -1.9. One of the reasons may be rather low
magnitude range of the dataset from Pyhäsalmi that results in faster flattening of the curve
and do not allow for larger magnitude range to be approximated with linear relation. Other
possibilities are: poor filtering of mine blasts, erroneous parameters of seismic events,
waveform corruption due to proximity of electrical noise, or even poor calibration of local
magnitude.

Figure 5-8. Frequency-Magnitude relation of the cluster QT121.

Figure 5-9. Frequency-Magnitude relation of the cluster QT84.

52
After validation of QT clusters, several parameters were calculated for each of them to
investigate their significance as a single seismic source, and to create a basis for further
analysis. The parameters were as follows:

o number of significant (MAG > 0) and large (MAG > 1) events,

o b value and a/b value (x-axis intercept) of the Frequency-magnitude
relation (see Equation (4),
o median of the S-wave to P-wave energy ratio,
o the sum of seismic energy,
o the sum of apparent stress.

The QT clusters created in the second stage of clustering were evaluated in order to limit
the number of clusters by rejecting marginal clusters that represent minor failure of
rockmass. In order to be categorized as marginal a cluster needs to meet at least two of
the following criteria:

 The number of events in a cluster is low (< 20-50).

 All events are small (MAG < -1).
 The total amount of Apparent Stress is low (< 30 bar).
 The total amount of seismic energy is low (< 1∙103 J).

After rejection of marginal clusters the resulting number of clusters was 366. Those
clusters were selected for the next stage. Five most populous QT clusters with calculated
parameters can be viewed in Table B-1 in Appendix B.

In the third stage of clustering, the Single linkage clustering algorithm (SLINK) was used
to link similar QT clusters into groups representing one seismic source. Euclidean
distance between cluster centroids was used as a distance metrics. Next, the similarity of
neighboring clusters was evaluated using following parameters:

 b-value of the Frequency-magnitude relation,

 median of the S-wave to P-wave energy ratio,
 top 5 maximum magnitudes recorded (values and dates),
 top 5 Apparent Stress peaks (values and dates),
 number of significant and large events,
 top 5 daily event histogram peaks (values and dates).

53
When the clusters showed similarities in those parameters (with a special emphasis on
first four) they were linked together into logical groups. It was possible to create 82
cluster groups, where:
 72 groups contain at least two QT cluster and maximum 12 QT cluster (group
S20).
 10 groups contain 10 single QT clusters with maximum number of events.
 2 groups contain 2 single QT clusters with the highest sum of seismic energy.

Table 5-3 presents five most populous clusters groups with centroids calculated as the
average of centroids from all the QT clusters within a group (for results of all 82 groups
please see Table B-2, Figure B-1and Figure B-2 in Appendix B). The maximum expected
magnitude is found by combining all seismic events from QT cluster contained within a
single group and producing an F-M diagram from resulting data set to find the x-axis
intercept. In the last step, each point within a single cluster group was assigned with
maximum predicted magnitude, with an assumption that seismic event of predicted size
can occur everywhere within a cluster group.

Table 5-3. Five most populous cluster groups created in the third stage of clustering using
SLINK algorithm.
Max. ∑ App.
Cluster Nr of ∑ Seismic
X [m] Y [m] Z [m] predicted Stress
group events Energy [J]
ML [bar]
S20 8337.8 2373.8 -1257.6 0.9 11097 3.60E+06 1.65E+04
S02 8268.7 2320.7 -1129.4 1.0 6564 1.16E+06 2.90E+03
S03 8366.2 2355.5 -1361.4 0.1 5581 2.30E+05 2.40E+03
S12 8379.3 2309.1 -1276.8 0.2 5579 1.89E+06 6.37E+03
S39 8352.9 2351.9 -1203.8 1.1 4464 1.16E+06 3.35E+03

Summary of the clustering process is presented in Figure 5-10. As can be seen, 82 Cluster
groups created from 297 QT cluster in the third stage contain 84% of total number of
seismic events (87 462 from 103 933 events). Furthermore, they represent 88% of the
total seismic energy (4.18∙107 J from 4.73∙107 J) and can be considered as a good
representation of the total population of seismic events.

54
Figure 5-10. A schematic summary of the clustering process
It has to be noted that due to the large amount of data, not all the results of the clustering
process are presented in this study in tabular format. Tables with all the results can be
presented on a request.

5.2 Selection of the investigation area

An analysis and interpretation of damages in production levels has been performed to find
the most unsafe levels, and to narrow down the area under investigation. It has been
assumed that high level of damage can be linked to high seismicity level, hence giving an
indication of high deterioration of the rockmass.

Damage data has been mapped by the mine personnel. Each damage record stored in the
database is described by several parameters:

 date of inspection,
 damage severity (subdivided into seven classes: very minor, minor, moderate,
major, serious, very serious, collapse),
 damage depth,
 support type used and condition its condition,
 location and surface extent,
 description,
 photograph (if the damage is more serious).

All the damages are also stored as a DTM surface file in Surpac software that allows to
localize each damage on the mining level map. The type of damage varies from superficial
cracks and fallouts of the shotcrete layer only in part of the excavation (very minor
damage), to more severe collapses of the all the tunnel profile, mainly under faulting
conditions (very serious to collapse type of severity). An example of more severe damage
can be found on Figure 5-11.

55
Figure 5-11. An example of rockburst damage that occurred 24.12.2001 on mining level
1200, at the upper level of stope GBL21.

During the analysis, all damages (according to the severity type) were summed for each
mining level. Results are plotted on Figure 5-12 together with the number of seismic
events with a local magnitude of 0 or higher.

As seen from Figure 5-12, the highest number of damages was mapped on the 1300 level.
However, many of them are only of very minor or minor severity. To select levels with
the highest level of deterioration, further investigation has been performed by ranking the
levels by number of damages in each class of moderate and higher severity. Next,
damages on levels ranked on the top in each category were summed up to find the ones
with highest number. Three mining levels which scored the maximum number are: 1225,
1250 and 1275. These are selected for further investigation of seismic risk.

56
 70 140

Number of events with a magnitude larger than 0

60 5 120
2
3 4
2
3
50 100
2 11
1
Number of damages

1 1
5 1
2 7
40 22 80
1 5 7
4 4 2
2 18 6
2 5 6
30 1 7 12 2 60
6 1
1 5 5 2
2
7 9 12 11
20 12 38 40
2 12 12 11
1
1
1 1 29
4 2 6 24 24
10 19 18 20
17 17
7 12 14 12 13
11 9
3 1
0 2 0
1150
1050

1075

1100

1125

1175

1200

1225

1250

1275

1300

1325

1350

1375

1400

1425
Mining level
Very minor Minor Moderate
Major Serious Very serious
Collapse MAG ≥ 0

Figure 5-12. All damages mapped on the mining levels in 2013, subdivided into severity
classes. Number of seismic events with a local magnitude 0 or higher.

It is possible that large amount of damages on these levels is related to higher ore
extraction. During the time period from September 2013 to January 2014 the largest
number of stopes were extracted on the level 1275. Furthermore, as it is shown on the
Figure 5-12, the number of seismic events with a magnitude larger than 0 (significant
events) correlates well with the damages observed on selected mining levels.

5.2.1 Division of selected mining levels

Selected mining levels were divided using a square grid consisting of 20m by 20m squares
indexed vertically with a capital letter from A to U and horizontally with a number from
1 to 19 (see Figure 5-13, Figure 5-14 and Figure 5-15). Each square in the grid
(assessment zone) was assigned with four characteristics:

 ground support type,

 maximum compression stress,
 intact rock strength (UCS)

57
  X, Y, Z coordinates of the excavation centroid point.

The support type was assigned based on the ground support map from the mine.
Additionally, each support type was indexed with a number (support index), according to
following rules:

 No bolts - 1
 Rock bolts - 2
 Cable bolts - 3
 Rock bolts and cable bolts – 4
 Mesh - 5
 Rock bolts and mesh - 25
 Rock bolts, cable bolts and mesh - 45

If there were mine openings with two different support types in one zone, then the name
of this zone also included the support index, for example on mining level 1275 zones
G16:2 and G16:25 for excavation supported with rock bolts and rock bolts with mesh,
respectively.

The maximum compression stress was assigned based on the rock mechanics numerical
model that was sectioned at the elevation of selected mining levels (Hakala, et al., 2013).
Intact rock strength was taken from the input parameters of the rock mechanics numerical
model as follows:

 Ore – 105 MPa

 Waste rock – foliation (on the ore-waste contact zone) – 110 MPa
 Waste rock– residual strength in zones were rockmass has yielded– 75 MPa
 Waste rock – 180 MPa

The resulting allocation of abovementioned parameters for assessment zones in all levels
can be found in Appendix C.

58
 Cable bolts designed,
not installed
Mesh designed, not
installed
No bolts
Rock & cable bolts
Rock bolts
Cable bolts
Mesh

Figure 5-13. Mining level -1225 with assessment zones. Colors represent different
support type.

Figures above and below show selected mining levels with assessment zones. Different
colors illustrate the ground support type, according to the legend. Dashed line in black
represents the ore-waste rock contact zone. Mining stopes are filled with gray color. The
alphabetical symbols represent mine infrastructure elements: KN - orepass, TN – backfill
raise, RTN, RTNP and IN – ventilation raises, VT – access ramp, TIMONKULU – shaft.

59
 Cable bolts

Mesh
No bolts
Rock & cb
Rock bolts
Cable bolts
Mesh

Figure 5-14. Mining level -1250 with assessment zones. Colors represent different
support type.

60
 Cable
Cablebolts designed,
bolts.
not installed
Mesh
Mesh designed, not
installed
No bolt
No
Rockbolts
& cable bolt
Rock & cable bolts
Rock bolt
Rock
Cablebolts
bolt
Cable
Mesh bolts
Mesh

Figure 5-15. Mining level -1275 with assessment zones. Colors represent different
support type.

5.3 Seismic hazard assessment

The assessment of seismic hazard has been performed using the two methodologies
described in chapters 3.3.1 and 3.3.2. The reason to select two approaches was that the
Probabilistic Approach for Seismic Risk Assessment (PASRA) methodology lacks a
possibility to compare the results with a rockburst damage scale developed from real
cases. It is difficult to assess the hazard as low or high, and only relative comparison
within a mine is possible. On the other hand, the Quantitative Seismic Hazard and Risk

61
Assessment Framework (QSHRAF) method gives the opportunity to quantify the hazard
using empirically developed scale.
As it was stated in chapter 2.3, due to fracturing of the orebody the stresses inside has
been redistributed and concentrated along the ore-waste contact zone and around the
orebody. This also resulted in an increase in seismicity. Taking this into account, hazard
assessment has been performed only on assessment zones located outside of the orebody
and on the contact zone.

5.3.1 PASRA
The seismic hazard was evaluated using the Probabilistic Approach for Seismic Risk
Assessment (PASRA) methodology described in chapter 3.3.2. First, the energy of
maximum possible event was calculated within each cluster group. To find the
relationship between local magnitude and seismic energy (as in Equation (9), a plot was
created from events database using magnitude and corresponding seismic energy. The
relationship was found using least squares regression. Results are plotted on Figure 5-16.
It is important to note that the data was selected from years 2011 to 2013. When seismic
events from 2014 were included, the linear relationship was not as good. This may be an
indication of poor system calibration after the upgrade at the end of 2013.

𝑙𝑜𝑔 𝐸 = 1.8 ∙ 𝑀𝐿 + 3.8

Figure 5-16. Local magnitude and seismic energy relation plotted from monitoring data
from years 2011-2013. Equation from linear regression (in red) used to calculate seismic
energy based on local magnitude.

Next, the Euclidean distance (Equation (12) was used to create a distance matrix with
length between all seismic events within cluster groups (seismic sources) and centroids

62
of mine opening within grids. The reason for calculating distance matrix for all points
within cluster groups (seismic sources) is a conservative assumption that the maximum
predicted event can occur at every point within a cluster group with equal probability.
The Released Energy Capacity (REC) was calculated for each cell in the distance matrix
according to Equation (10), with the energy of maximum predicted seismic event scaled
for distance. The attenuation coefficient was selected to be equal to 0.0001. Next, the
maximum value of REC was found for each centroid of mine opening.

The maximum intensities of REC were found for assessment zones: E14, I18 and L9 on
mining level 1225; I2, L7 and O13 on mining level 1250; K2, L3 and L2 on mining level
1275. From the results of REC it was observed that the distance from the centroid of
assessment zones to cluster groups did not have influence on REC result. For all
assessment zones on all levels the maximum predicted energy was related to only one
cluster group (S39 with the highest predicted local magnitude of 1.1). Even when the
distance from the source to assessment zone was as large as 313m (from zone E17 to
group S39) the highest REC was still related to this group. This is an indication that the
distance scaling factor in REC formula (as in Equation (10) is inappropriate.

The Absorbed Energy Capacity (AEC) was evaluated using qualitative scale of ground
support capacity present in Table 5-4.

Table 5-4. Absorbed Energy Capacity scale for ground support (*including fibrecrete
when mesh is not used and shotcrete when mesh is used).
Support type* AEC rating
No bolts 1
Rock bolts 2
Cable bolts 3
Rock bolts and cable bolts 4
Rock bolts and mesh 5
Rock bolts, cable bolts and mesh 6

Seismic hazard was calculated as the REC/AEC factor according to Equation(11). On the
mining level 1225 the top scores of REC/AEC were found for assessment zones: L9, L7,
L6 and R7 (see Figure 5-17). The average REC/AEC is 12.4. On the mining level 1250
the top scores of REC/AEC were found for assessment zones: P10, N8, N5 (see Figure
5-18). The average REC/AEC is 8.7. On the mining level 1275 the top scores of
REC/AEC were found for assessment zones: N7, R7:1, S7 and R8 (see Figure 5-19). The
average REC/AEC is 9.7. Detailed results of PASRA can be found in Table C-1, Table
C-2 and Table C-3 in Appendix C.

63
 REC/AEC

10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0

0.0
5.0
L9
L7
S7
L6
R7
N8
J1
K5
I18
N5
S5
T5:2
S4
E17
G16:2
O6
P6
Q6
R6
M3
H17
N6

Figure 5-17. Results of REC/AEC for mining level 1225.

M7
F17
E16
F15
L3
G17
M9
J18
K2
H18
D14
D13
C12
T5:3
T4
T3
REC/AEC - Level 1225

E14
L12
J17
J15
K12
M8
M5:4
F16
G16:25
J3
M12
L4:4
N10
D15
K4
M6
M5:45
M11
M10
K3
E15
K16
D11
L4:45
L13:4
K15
J14
L13:25
I15
K13
K14
64
 REC/AEC

10.0
15.0
20.0
25.0
30.0
35.0
40.0

0.0
5.0
P10
N8
N5
L7
O12
P11
N13
M4
M3
L2
M2
H17
J1:2
F17
C14
C9
M5:2
G17
E17
I1
D15
C13
M6
M8
F16
D14
C12:2
B11
B10

Figure 5-18. Results of REC/AEC for mining level 1250.

I2
M10
M9:2
K3:2
E15
D13
B9:2
B8
K5
L9
L5
J3:2
H1
C10
B7
L4
O13
M13
L13
L14
I3
REC/AEC - Level 1250

F13
L8
N12
L3
B9:3
M5:4
M7
C12:4
N9
K2
M9:4
C7
E14
C8
J1:45
G1
J3:4
M11
K3:25
J16
C11
D12:4
E1
E2
D2
D11
D12:4
M12
L12
K15
H2
F1
K12
J13
C3
D8
I17
C4
D8
C6
C5
65
 REC/AEC

10.0
15.0
20.0
25.0
30.0
35.0
40.0

0.0
5.0
N7
R7:1
S7
R8
L2:2
P6
O6
N6
Q6
R6
R7:2
M4
S6
M3
T6
N2
M2
K3
L5:2
H1
F1

Figure 5-19. Results of REC/AEC for mining level 1275.

I0
J1:2
O10
M7:4
J2:2
I1:2
H0
E0
N10
N9
L4:25
L3
M8
L2:45
REC/AEC - Level 1275

N11
M6
M7:45
M5
L4:45
M9:25
L9:4
M9:45
L5:45
K2
K4
L6
F0
K10
K5
M11
M12
M10
L11
K12
L9:45
L13
J2:25
I1:25
G0
L10
J1:45
K13
J13
66
5.3.2 QSHRAF
The Quantitative Seismic Hazard and Risk Assessment Framework (QSHRAF) approach
(described in chapter 3.3.1) has been used to evaluate seismic hazard. It is based on
evaluation of the maximum ground motion using Peak Particle Velocity (PPV) as in
Equation (6). Calculation of PPV was done for every centroid of assessment zones is
similar to calculation of REC in PASRA. Euclidean distance formula was used to create
a distance matrix with distances between all seismic events (grouped into cluster groups
- seismic sources) and assessment zones centroids. Then PPV was calculated using the
maximum predicted magnitude assigned to each seismic event, for each distance value,
to scale the PPV for length. Then, the largest expected PPV was found for each zone.
Results of PPV for each level can be found in Appendix D.

Next step was to calculate Excavation Vulnerability Potential (EVP) according to

Equation (7), stress condition factor (E1) was calculated as a ratio of static stress to
rockmass strength (UCS) in the vicinity of excavation. The values for stress were taken
from the rock mechanics numerical model and rock strength was assigned as described in
chapter 5.2.1. Ground support capacity (E2) has been assigned to assessment zones using
the support scale presented in Table 3-7, however slight changes has been made in order
to include the different support classes used in the mine. Meshing and shotcreting is used
in every excavation in the mine, therefore the rating for ‘No bolts’ support type has been
lowered from 5 (minimal surface support without rock bolts) to 4 (surface support without
rock bolts). On the other hand, if an additional mesh was used, the E2 rating was increased
to 6 (to account for increased capacity of additional surface support). Table 5-5 gives a
summary of support types and allotted E2 factor.

Table 5-5. Ground support capacity (E2) scale for ground support.
Support type E2
No bolts 4
Mesh 4
Rock bolts 5
Cable bolts 5
Rock bolts and cable bolts 10
Rock bolts and mesh 6

Excavation span (E3) was calculated as a diameter of a circle drawn within excavations
in assessment zones. Geological structure factor (E4) has been assigned with a value of
0.5 to assessment zones along the ore-waste contact zone as a potential failure surface
promoting rockmass failure. Other zones were assigned with E4 factor equal to 1.5.

67
The Rockburst Damage Potential (RDP) was calculated as a product of PPV and EVP (as
in Equation (8) and evaluated using the EVP vs PPV diagram (as in Figure 3-3) and
rockburst damage scale (as in Table 3-9). Results of the largest RDP for assessment zones
on three mining levels were plotted on Figure 5-20.

As can be seen on EVP vs PPV chart, four assessment zones on mining level 1225 are
within the R2 rockburst damage zone (minor damage/less than 1t of rock displaced): M9
(RDP = 121.3), N5 (RDP = 76.9), N8 (RDP = 57.9), K5 (RDP = 54.5). It is important to
note that N5 zone has surface support only (shotcrete and mesh) and no rock bolt, so the
E2 rating is 4. Other zones are within the R1 zone (no damage/minor loose), however two
zones: K2 (RDP = 49.8) and L9 (RDP = 46.5) are very close to the R2 zone and only
slight increase in PPV may result in damage. The average RDP on this level is 15.8.

On mining level 1250 are within the R3 rockburst damage zone (1 to 10t rock displaced):
N8 (RDP = 157.6), L7 (RDP = 147.8), N5 (RDP = 147.3). One assessment zone – M9:2
(RDP = 61.3) is in the R2 rockburst damage zone. Other zones are within the R1 zone (no
damage/minor loose), however two zones: M4 (RDP = 45.8) and L5 (RDP = 45.4) are
very close to the R2 zone and only minor intensification of PPV may result in damage.
The average RDP on level is 15.9.

Five assessment zones on mining level 1275 are within the R2 rockburst damage zone
(minor damage/less than 1t of rock displaced): M3 (RDP = 77.3), N7 (RDP = 75.2), K3
(RDP = 64.8), M2 (RDP = 59.5) and L3 (RDP = 53.6. Other zones are within the R1 zone
(no damage/minor loose), yet zone N10 (RDP = 47.5) is very close to the R2 zone and
only slight increase in PPV may result in higher hazard and damage. The average RDP
on this level is 16.5.

Detailed results of EVP, PPV and RDP calculation, as well as EVP vs PPV diagrams for
all assessment zones can be consulted in Appendix D.

68
 7

5
Peak Particle Velocity, PPV [m/s]

0
1 10 100 1000
Excavation Vulnerability Potential, EVP

Figure 5-20. EVP vs. PPV diagram illustrating assessment zones with the largest rockburst damage potential (Mikula, et al., 2008).

69
5.4 Seismic risk evaluation and representation
Final seismic risk is calculated using the Quantitative Seismic Hazard and Risk
Assessment Framework (QSHRAF) approach as a product of seismic hazard express in
RDP parameter and exposure to hazard of mine personnel that has been assigned using
the scale presented Table 3-10.

Table 3-10 illustrates as well the risk matrix with different Seismic Risk Ratings (SRR)
that are assigned to each assessment zone. Below, Figure 5-21 illustrates the number of
assessment zones with assigned SRR on each mining level. The SRR varies from ‘Very
Low’ (VL) to ‘Extreme’ (E). As can be seen, the majority of assessment zones have very
low seismic risk. On mining level 1225: 60 zones (87.0% of total) have very low SRR, 8
zones (11.6% of total) have low SRR and 1 zone (M9, 1.4% of total) has moderate SRR.
On mining level 1250: 70 zones (76.0% of total) have very low SRR, 18 zones (18.7% of
total) have low SRR and 3 zones (N8, L7, N5; 3.3% of total) have moderate SRR. On
mining level 1275: 54 zones (84.4% of total) have very low SRR, 9 zones (14.1% of total)
have low SRR and 1 zone (M3; 1.6% of total) has moderate SRR. Detailed results
showing SRR for all assessment zones can be consulted in Appendix D.

1225 1250 1275

80
Number of asessment zones per level

69
70
60
60 54
50
40

30
18
20
8 9
10 4
1 1 0 0 0 0 0 0 0 0 0
0
Very Low Low Moderate High Very high Extreme
VL L M H VH E
Seismic Risk Rating

Figure 5-21. The number of assessment zones on each mining level assigned with Seismic
Risk Ratings.

70
5.5 Seismic hazard and risk mitigation
The degree of damage due to seismic events (seismic hazard) can be lowered using
following strategies:

o Install additional support to decrease Excavation Vulnerability Potential

(EVP), hence reducing Rockburst Damage Potential (RDP) and risk
o Reduce span of excavation in newly designed areas

Risk mitigation can only be commenced when the acceptable level of risk is known. It
has to be specified by mine management beforehand, as a realistic target that will be used
to evaluate different mitigation scenarios. Here, mitigation measures are only considered
as initial recommendations, because the tolerable level of risk is not known. Furthermore,
measures are presented merely in existing mine openings, therefore only the first strategy
is evaluated.

Risk can be lowered through an increase of support capacity, therefore reducing the
damage potential. On mining level 1225 all assessment zones have maximum RDP within
the damage zone R2 (damage expected to be contained by support), therefore only
installation of surface support which is sufficiently to transfer load to individual
reinforcing elements is required. Following upgrades of the support system are
recommended for zones:

 M9 – installation of rock bolts, cable bolts and additional meshing, (reduction of

RDP from 121.3 to 60.6 and risk from moderate to low), further reduction is
possible only through installation of dynamic support system, such as cone bolts
with dynamic surface support.
 N5 - installation of rock bolts, cable bolts and additional meshing, (reduction of
RDP from 76.9 to 51.2 and risk from low to very low).
 N8, K5 and L9 – installation of rock bolts and mesh.
 K2 – additional mesh (reduction of RDP from 49.8 to 41.5 and risk from low to
very low).

On mining level 1250, RDP can be lowered in zones:

 N8 and N5 - installation of rock bolts, cable bolts and additional meshing

(reduction of risk from moderate to very low).

71
  L7 - installation of dynamic support system, such as cone bolts with dynamic
surface support (reduction of RDP from 147.8 to 29.6 and risk from moderate to
low).
 M4 and L5 – installation of additional mesh.

On mining level 1275, RDP can be lowered in zones:

 M3 - installation of cable bolts and additional meshing (reduction of RDP from

77.3 to 38.8 and risk from moderate to low).
 N7 - installation of rock bolts and mesh (reduction of RDP from 75.2 to 50.2 and
risk from low to very low).
 M2 - installation of additional mesh.

Results of RDP reduction can be seen on EVP vs PPV diagrams in Appendix E.

72
6 Discussion
This chapter is divided into two parts. First, it discusses the results and main aspects of
seismic risk assessment in the Pyhäsalmi mine and compares two assessment methods.
Second, it articulates strengths and weaknesses of the Geotechnical Risk Assessment
guideline as a result of experience gained in the process of risk assessment.

6.1 Seismic hazard and risk assessment

The process of seismic risk assessment has several limitations. First of all, data clustering
is very cumbersome and time-consuming procedure; automation through software would
be beneficial. Secondly, applied procedure has to be updated constantly in order to include
new events, so that risks can be monitored with advancing mining rate. Thirdly,
successful assessment depends highly on correctness of local magnitude scale and other
seismic parameters, especially seismic event location. Next, from the results of Released
Energy Capacity (REC, see chapter 5.3.1) it can be observed that the distance from the
centroid of assessment zones to cluster groups does not have influence on REC result.
For all assessment zones on all levels the maximum predicted energy is related to only
one cluster group (S39 with the highest predicted local magnitude of 1.1). Even when the
distance from the source to assessment zone is as large as 313m (from zone E17 to group
S39) the highest REC is still related to this group. This is an indication that the distance
scaling factor in REC formula (as in Equation (10) is inappropriate.

The results from two approaches used in the process of seismic hazard assessment,
namely the Probabilistic Approach for Seismic Risk Assessment (PASRA) results and
the Quantitative Seismic Hazard and Risk Assessment Framework (QSHRAF) are
compared here. The highest seismic hazard using PASRA method is found in following
assessment zones:

 on mining level 1225: L9,L7,L6,R7

 on mining level 1250: P10,N8, N5
 on mining level 1275: N7, R7:1, S7, R8.

Mining level 1225 has the highest average seismic hazard (average REC/AEC of 12.4).

On the other hand, the highest level of seismic hazard using QSHRAF method was found
on level 1275 (average RDP of 16.5). Most hazardous assessment zones are:

 on mining level 1225: M9, N5, N8, K5

73
  on mining level 1250: N8, L7, N5, M9:2
 on mining level 1275: M3, N7, M2, L3
Although there are some assessment zones that have high seismic hazard for both methods
(for example zones N8 and N5 on mining level 1250, and zone N7 on mining level 1275),
it can be seen that the results of two methods are different. The dissimilarity can be
explained by discrepancies in calculation of seismic hazard using those methods, for
example:

 PASRA is based on seismic energy and QSHRAF on local magnitude.

 Ground support factors are different for both methods.
 There is no geological structure factor in PASRA.

It has to be stated that QSHRAF method is more trustworthy, since it is implemented in

the Mine Seismicity Risk Assessment Program (MS-RAP) software that is currently used
in many mines all over the world.

6.2 Geotechnical risk assessment guideline

The geotechnical risk assessment guideline supports in successful selection of an
appropriate assessment approach. It gives a selection of tool to identify hazards and then
to assess the risks related to those hazards. It provides examples of hazard and required
steps to assess risk, however the list is not complete. During the application of guideline
it was found out that it is not perfectly suited for assessment of seismic risk.

The Geotechnical Hazard Potential evaluation method aids in fast description of the
general hazard level in the mine, however it has some limitations. First, is uses Barton’s
Q system that is not used in the mine on regular basis. It would be beneficial to implement
other geotechnical parameters to evaluate hazard potential. Second, only few mining
method are available in the guideline. More detailed selection of mining method could
bring more benefit to be more specific.

The Geotechnical Risk Assessment (GRA) approach selection tool (see chapter 4.2) helps
in fast and reliable selection of appropriate methodology of risk assessment. However
some remarks have to be made regarding the number of point assigned to each GRA
category. The final results (selected approach) highly depends on amount of points for
each category where each category can score from 0 to 4 points. Particularly the category
‘available resources’, as the only category with a maximum score of 4 points, has a very

74
big influence on the final score and have to be carefully chosen in order not to introduce
any bias. Furthermore, the deterministic approach is unlikely to score the highest amount
of points to be selected as the most appropriate one, what is questionable.

Last issue discussed here relates to the amount of required geotechnical data available for
analysis. In this study, seismic activity has been selected as the geotechnical hazard for
the assessment, primarily because the amount of data and measurements results easily
available for analysis. This emphasizes the importance of geotechnical monitoring carried
out on regular basis in order to provide sufficient data density.

7 Conclusions
In conclusion, seismic risk in the Pyhäsalmi mine can be considered as low. In majority
of assessment zones the level of seismic risk is found to be very low or low and only in
six assessment zones is moderate. The biggest risk is found in mine openings located at
the northern ore-waste contact zone, near mine infrastructure such as ore passes, fresh air
rescue chamber and access drive. Those areas require special attention and installation of
additional ground support in order to prevent damages in case a severe seismic event takes
place. This result confirms the necessity for further development and implementation of
seismic risk monitoring and risk assessment as an important element of safe mining
operations.

The Geotechnical Risk Assessment guideline that was a basis for the risk assessment
process in this study supports in successful selection of appropriate assessment approach
and aids in selection of tools to assess geotechnical risks. Its applicability can be
considered as high, however some elements could be improved to use its full potential.
Focus should be put on reevaluation of the risk assessment approach selection tool, as
well as to creation of a geotechnical hazards database with recommended assessment
approaches with required geotechnical data and hazard descriptions. This could provide
a basis for fast selection of required data, calculation methods and resources that are
needed in the risk assessment process.

8 Recommendations for future work

This chapter gives recommendations for future work and is divided into two sections:
seismic risk assessment and risk mitigation measures, and further development of the
Geotechnical Risk Assessment guideline.

75
8.1 Seismic risk assessment and mitigation
In order to further develop the seismic risk assessment process in the Pyhäsalmi mine
following steps are recommended:
 Develop a scale of personnel exposure that is well-suited for operating conditions
in the mine. This can be done through detailed measurement of time spent by
workers in each location and measurements of personal protection.
 Put emphasis on keeping a good record of all damages related to seismic events,
so that they could be easier evaluated in the future. This could help to tune the
damage scale to a particular seismic event size.
 Install additional ground support in areas where seismic risk is the highest in order
to prevent damages.

8.2 Geotechnical risk assessment guideline improvement

In order to further upgrade the Geotechnical Risk Assessment guideline following
recommendations are given:
 Include the description of seismic risk assessment approach as part of
development of a systematic collection of assessment approaches with required
geotechnical data and hazard descriptions. This could help to simplify the
selection process of appropriate assessment approach that is well-suited for
specific hazard conditions.
 Apply the guideline internally within a company by mine personnel to test how
the guideline can be included as a systematic duty of mine personnel responsible
for geotechnical risk assessment and how successfully it could help to perform the
assessment. This will result in larger resources allocation.
 Use the guideline to investigate other geotechnical hazards and other risk
assessment approaches. Particularly risk assessment for small scale area, as well
as site specific risk assessment will bring more information on applicability of the
guideline.

76
References
Bednarik, L. i Kovacs, L., 2012. Efficiency analysis of quality threshold clustering
algorithms. Production Systems and Information Enegineering, 6(2013), pp. 15-26.

Bergström, P., 2012. Ground Control Audit - Subsidence, Pyhäsalmi Mine: Internal
company presentation.

Bergström, P., 2013. Ground Control Audit - Seismicity, Pyhäsalmi Mine: Internal
company presentation.

Bergström, P., 2014. Rock mechanics at the Pyhäsalmi Mine, Pyhäsalmi Mine: Internal
company presentation.

Blake, W. & Hedley, D., 2001. Rockburst case histories for north american hardrock
mines, Sudbury: Canadian Mining Research Organization.

de Jongh, R., 2013. Installation report: Pyhäsalmi Mine Oy, Pyhäsalmi: Not published.

Falmagne, V., 2001. Quantification of rock mass degradation using microseismic

monitoring and applications for mine design, Ph.D. thesis: Queen's University,
Ontario, Canada.

First Quantum Minerals, 2013. Pyhasalmi Production Stats. [Online]

Available at: http://www.first-quantum.com/Our-Business/operating-
mines/Pyhasalmi/Production-Stats/default.aspx [Accessed 3 March 2014].

Froehlich, T., 2014. Guideline for geotechnical risk assessment: Geotechnical risk
assessment in the Garpenberg mine, Sweden, Helsinki: Unpublished.

Gibowicz, S. & Kijko, A., 1994. An introduction to mine seismology. 1st ed. San Diego:
Academic Press.

Gleeson, D., 2010. Innovation at depth. International Mining, Volume April, pp. 12-18.

Hakala, M., Lamberg, M. & Kuula, H., 2013. Life of Mine Rock Mechanical Simulation,
s.l.: Unpublished.

Halkidi, M., Batistakis, Y. & Vazirgiannis, M., 2001. On Clustering Validation

Techniques. Journal of Intelligent Information Systems, 17(2-3), pp. 107-145.

Heyer, L., Kruglyak, S. & Yooseph, S., 1999. Exploring expression data: identification
and analysis of coexpressed genes. Genome Res, 9(11), pp. 1106-15.

Hudyma, M., 2004. Mining-Induced Seismicity in Underground, Mechanised, Hardrock

Mines—Results of a World Wide Survey, Nedlands: Australian Centre for
Geomechanics.

Hudyma, M. & Potvin, Y., 2004. Seismic hazard in Western Australian mines. The
Journal of The South African Institute of Mining and Metallurgy, 104(5), pp. 265-
276.

77
Hudyma, M. & Potvin, Y. H., 2010. An Engineering Approach to Seismic Risk
Management in Hardrock Mines. Rock Mechanics and Rock Engineering, 43(6), pp.
891-906.

Hudyma, M. R., 2008. Analysis and Interpretation of Clusters of Seismic Events in

Mines, PhD thesis: The University of Western Australia.

Hudyma, M. R., 2014. Personal correspondation. s.l.:s.n.

Jain, A., Murty, M. & Flynn, P., 1999. Data Clustering: A Review. ACM Computing
Surveys, 31(3), pp. 264-323.

Kaiser, P. K. & Cai, M., 2012. Design of rock support system under rockburst
condition. Journal of Rock Mechanics and Geotechnical Engineering, 4(3), pp. 215-
227.

Kaiser, P., Vasak, P. & Suorineni, F., 2005. Hazard Assessment In Burst-prone Mines -
Seismic Data Interpretation With 3D Virtual Reality Visualization. London,
International Society for Rock Mechanics.

Meyer, S., 2014. Monthly Seismic Report: 01 Jan 2014, 00:00 - 01 Feb 2014, 00:00,
Pyhasalmi: Unpublished.

Mikula, P., Heal, D. & Hudyma, M., 2008. Generic Seismic Risk Management Plan for
Underground Hardrock Mines, Nedlands: Australian Centre for Geomechanics.

Mishra, R. K., 2012. Guidelines for getechnical risk assessment in underground mines,
MSc thesis: Aalto University.

Numminen, M., 2012. Pyhasalmi mine geology - presenation. s.l.:s.n.

Ortlepp, W. D., 1997. Rock fracture and rockbursts: an illustrative study.

Johannesburg: South African Institute of Mining and Metallurgy.

Owen, M., 2004. Exposure model – Detailed profiling and quantification of the
exposure of personnel to geotechnical hazards in underground mines, PhD Thesis:
The University of Western Australia.

Oye, V. & Roth, M., 2005. Source parameters of microearthquakes from the 1.5km
deep Pyhasalmi ore mine, Finland. Stanford, California, Stanford University.

Potvin, Y., 2009. Strategies and tactics to control seismic risk in mines. The Journal of
The Southern African Institute of Mining and Metallurgy, 109(3), pp. 177-186.

Puustjärvi, H., 2006. Pyhäsalmi Modeling Project, Unpublished: Geological Survey of

Finland.

Pyhäsalmi Mine Oy, 2013. Life of mine, Pyhajarvi: Internal company presentation.

Richter, C., 1935. An instrumental earthquake magnitude scale. Bulletin of the

Seismological Society of America, 25(1), pp. 1-32.

78
Sambuelli, L., 2009. Theoretical derivation of a peak particle velocity-distance law for
the predictian of virbrations from blasting. Rock Mechanics and Rock Engineering,
42(3), pp. 547-556.

Trzaska, W. et al., 2010. Laguna in Pyhäsalmi. Acta Physica Polonica B, 41(7), pp.
1779-1787.

Weisstein, E. W., 2012. "Distance" From Mathworld-A Wolfram Web Resource.

[Online] Available at: http://mathworld.wolfram.com/Distance.html
[Accessed 21 03 2014].

Wen, Z., 2013. Probabilistic approach for seismic risk assessment. Helsinki: s.n.

Young, R., Maxwel, S., Urbanic, T. & Feignier, B., 1992. Mining-induced
microseismicity: Monitoring and applications of imaging and source mechanism
techniques. Pure and applied geophysics, 139(3-4), pp. 697-719.

79
Appendices
A Ground support system standard

Figure A-1. Minimum standard for 5.5m wide meshed heading – for wider headings
additional support to be installed (Bergström, 2014).

Figure A-2. Minimum standard for 5.5m wide heading without mesh – for wider headings
additional support to be installed (Bergström, 2014).

80
B Seismic events clustering
Table B-1. Five most populous clusters created using the QTCLUST algorithm with
parameters used for evaluation and comparison in the third stage of clustering
procedure.
Cluster ID QT 5 QT 21 QT 4 QT 14 QT 8
X [m] 8332.5 8270.6 8337.4 8366.6 8372.4
Y [m] 2366.3 2312.2 2373.0 2354.3 2314.8
Z [m] -1232.2 -1124.1 -1260.8 -1337.1 -1265.5
Nr of events 3908 3664 3047 2098 1848
Nr of significant
27 2 6 1 7
events
Number of large
1 0 0 0 0
events
b-value -1.1 -1.0 -1.3 -1.9 -1.1
a/b 1.2 1.4 0.7 0.1 1.1
Median S:P
17.9 16.9 13.7 11.0 8.3
Energy
Histogram peak 1 2013-11-21 2014-02-17 2012-05-04 2013-11-02 2012-12-27
Histogram peak 2 2012-05-04 2014-02-16 2013-10-31 2013-11-07 2013-11-10
Histogram peak 3 2013-11-22 2014-02-20 2013-04-12 2012-12-28 2014-02-15
Histogram peak 4 2013-12-02 2013-10-31 2012-12-20 2013-01-31 2011-01-09
Histogram peak 5 2013-11-28 2013-11-02 2013-09-08 2013-10-26 2013-01-29
ASTH peak 1 314.1 29.1 228.8 88.4 85.7
ASTH peak 2 228.3 26.1 106.2 46.9 82.0
ASTH peak 3 202.2 20.8 80.5 29.2 80.0
ASTH peak 4 112.7 20.6 77.4 14.3 76.1
ASTH peak 5 95.4 20.2 75.3 14.0 45.4
ASTH peak date 1 2012-05-25 2012-11-16 2013-03-26 2011-11-03 2011-01-16
ASTH peak date 2 2012-08-13 2012-12-05 2011-01-01 2012-12-06 2011-01-09
ASTH peak date 3 2011-03-24 2012-12-05 2012-10-29 2011-01-17 2011-02-08
ASTH peak date 4 2011-02-15 2012-12-05 2011-03-14 2011-12-10 2011-02-08
ASTH peak date 5 2011-07-20 2012-11-21 2011-06-22 2012-05-07 2011-01-09
Max magnitude 1 1.0 0.4 0.6 0.0 0.4
Max magnitude 2 0.5 0.2 0.1 -0.1 0.2
Max magnitude 3 0.5 -0.1 0.1 -0.4 0.2
Max magnitude 4 0.5 -0.1 0.1 -0.5 0.1
Max magnitude 5 0.5 -0.2 0.0 -0.5 0.1
Max magnitude
2011-01-28 2012-11-19 2012-05-13 2012-10-11 2013-09-23
date 1
Max magnitude
2012-05-04 2014-01-21 2012-03-04 2011-01-30 2013-09-12
date 2
Max magnitude
2012-12-19 2013-08-23 2012-03-15 2013-07-16 2014-02-15
date 3
Max magnitude
2013-03-16 2014-02-20 2013-09-07 2012-06-15 2011-04-02
date 4
Max magnitude
2013-09-12 2012-11-16 2013-02-05 2012-10-13 2011-11-14
date 5
∑ E [J] 2.17E+06 4.60E+05 4.99E+05 6.46E+04 2.82E+05
∑ App stress [bar] 5933.2 1203.2 4509.9 883.8 1651.8

81
Table B-2. Cluster groups created in the third stage of clustering using SLINK clustering
algorithm.
Max. ∑ App
Cluster Nr of ∑ Energy
X [m] Y [m] Z [m] predicted stress
group events [J]
ML [bar]
S01 8375.4 2353.2 -1124.6 0.6 1927 3.44E+05 8.47E+02
S02 8268.7 2320.7 -1129.4 1.0 6564 1.16E+06 2.90E+03
S03 8366.2 2355.5 -1361.4 0.1 5581 2.30E+05 2.40E+03
S04 8164.9 2159.8 -1175.3 0.3 1052 1.06E+06 6.81E+03
S05 8445.7 2306.2 -1204.6 0.3 1249 2.30E+05 1.97E+03
S06 8319.0 2311.7 -1072.6 0.5 2171 4.17E+05 9.60E+02
S07 8442.2 2330.8 -1084.7 0.8 3818 3.40E+05 1.69E+03
S08 8298.0 2290.2 -1357.2 0.9 1786 7.84E+05 2.60E+03
S09 8307.9 2318.2 -1150.6 0.2 1409 3.87E+05 3.54E+02
S10 8217.5 2304.8 -1315.8 0.6 2489 9.52E+04 5.22E+02
S11 8242.6 2350.6 -1135.5 0.8 333 2.86E+05 4.14E+02
S12 8379.3 2309.1 -1276.8 0.2 5579 1.89E+06 6.37E+03
S13 8426.5 2370.3 -1096.7 0.7 280 1.32E+05 2.93E+02
S14 8465.1 2296.4 -1069.2 0.5 843 5.24E+05 6.55E+01
S15 8367.4 2294.5 -1347.0 0.7 1397 1.00E+06 2.77E+03
S16 8452.7 2253.6 -1252.0 -0.2 58 4.69E+05 3.66E+02
S17 8218.5 2183.9 -1168.7 -0.2 184 4.77E+04 6.45E+02
S18 8311.0 2365.4 -1168.3 0.7 120 1.05E+05 6.05E+01
S19 8329.2 2333.7 -1325.4 0.3 2024 1.46E+05 1.17E+03
S20 8337.8 2373.8 -1257.6 0.9 11097 3.60E+06 1.65E+04
S21 8409.1 2333.8 -1169.7 0.4 1142 1.88E+05 9.14E+02
S22 8241.4 2271.5 -1231.5 -0.3 70 4.52E+04 3.57E+02
S23 8249.6 2326.1 -1207.2 0.5 2856 7.73E+05 2.64E+03
S24 8219.7 2314.9 -1246.2 0.5 3052 5.66E+05 2.81E+03
S25 8468.7 2286.6 -1155.2 -0.1 438 1.07E+05 4.83E+02
S26 8322.2 2218.9 -1021.4 0.0 713 2.72E+05 2.28E+03
S27 8252.4 2326.4 -1302.0 -0.1 874 4.79E+04 3.32E+02
S28 8223.5 2105.5 -1199.5 0.0 736 1.06E+06 5.46E+03
S29 8284.2 2149.1 -1115.5 0.1 1095 3.02E+05 2.86E+03
S30 8212.4 2297.5 -1363.8 0.2 545 3.29E+05 9.27E+02
S31 8262.9 2293.1 -1074.9 0.7 1504 1.09E+06 1.88E+03
S32 8467.9 2186.0 -1165.1 0.0 48 4.07E+05 1.46E+02
S33 8281.8 2103.1 -1252.9 0.1 294 1.63E+06 3.11E+03
S34 8265.2 2258.9 -1417.0 -0.1 60 6.27E+04 1.85E+02
S35 8320.8 2220.1 -938.3 -0.1 277 2.55E+05 2.07E+03
S36 8311.9 2098.0 -1136.7 0.2 465 6.06E+05 3.52E+03
S37 8389.3 2334.5 -1069.9 0.3 1034 1.31E+05 4.78E+02
S38 8450.7 2363.2 -1064.0 0.0 78 2.62E+04 9.56E+01
S39 8352.9 2351.9 -1203.8 1.1 4464 1.16E+06 3.35E+03
S40 8209.6 2269.9 -1403.2 -0.1 381 4.59E+04 4.77E+02
S41 8177.5 2295.6 -1294.7 0.1 871 2.63E+05 1.14E+03

82
 Max. ∑ App
Cluster Nr of ∑ Energy
X [m] Y [m] Z [m] predicted stress
group events [J]
ML [bar]
S42 8244.4 2207.7 -1114.1 0.5 2067 8.02E+05 4.22E+03
S43 8211.4 2149.6 -1131.8 -0.1 534 2.93E+05 2.77E+03
S44 8323.1 2271.9 -993.4 0.3 862 3.64E+05 1.82E+03
S45 8452.6 2337.5 -1151.0 0.3 2004 1.06E+05 1.71E+03
S46 8243.1 2049.2 -1233.5 0.1 255 2.07E+05 1.37E+03
S47 8398.8 2295.5 -1164.5 0.0 209 1.65E+04 9.87E+01
S48 8379.2 2393.6 -1292.9 0.4 316 3.30E+05 2.80E+02
S49 8252.8 2122.2 -1118.1 0.1 232 3.02E+05 1.25E+03
S50 8170.6 2228.0 -1174.6 0.1 1409 4.84E+05 5.73E+03
S51 8201.7 2260.3 -1104.9 0.0 1279 1.12E+05 3.02E+03
S52 8208.3 2313.1 -1159.7 0.1 942 3.03E+05 2.23E+03
S53 8217.0 2306.9 -1081.2 -0.1 105 2.42E+04 1.60E+02
S54 8364.1 2327.3 -1030.8 0.1 526 1.29E+05 8.43E+02
S55 8417.6 2248.9 -1334.1 0.2 181 1.37E+05 7.77E+02
S56 8357.7 2329.2 -989.2 0.0 137 4.25E+04 2.46E+02
S57 8223.1 2040.2 -1201.8 0.2 133 1.92E+05 1.17E+03
S58 8495.1 2317.3 -1091.3 0.0 229 1.92E+04 1.18E+02
S59 8261.9 2034.6 -1279.6 0.0 113 1.60E+05 8.06E+02
S60 8269.2 2087.7 -1101.2 0.1 153 5.55E+05 1.13E+03
S61 8488.3 2235.1 -1233.2 0.1 63 2.05E+05 3.49E+02
S62 8348.6 2097.7 -1218.0 0.3 234 5.02E+06 1.22E+04
S63 8461.2 2179.9 -1098.0 0.3 142 7.27E+05 4.48E+02
S64 8303.6 2052.4 -1219.3 0.2 133 6.65E+05 1.47E+03
S65 8210.3 2363.3 -1094.3 0.3 40 8.68E+04 1.15E+02
S66 8162.4 2279.9 -1328.0 0.3 237 5.60E+04 3.88E+02
S67 8479.9 2370.2 -1104.6 0.1 279 5.16E+04 2.93E+02
S68 8478.8 2173.4 -1222.9 0.2 61 5.45E+05 4.56E+02
S69 8478.3 2284.7 -1324.8 0.1 175 1.05E+05 9.62E+02
S70 8370.8 2092.1 -1154.1 0.1 67 1.97E+05 9.00E+02
S71 8365.0 2208.6 -1028.4 -0.2 102 5.25E+04 4.57E+02
S72 8253.7 2305.4 -1332.6 0.9 1016 2.83E+05 2.41E+02
S73 8365.4 2365.5 -1305.8 0.4 767 8.07E+04 5.49E+02
S74 8321.2 2432.8 -1213.4 0.2 327 2.73E+04 1.60E+02
S75 8230.3 2303.3 -1125.6 0.0 245 2.11E+04 3.13E+02
S76 8328.3 2303.4 -1280.8 0.0 182 1.18E+04 4.81E+01
S77 8394.1 2359.9 -1082.0 0.3 175 1.78E+04 2.15E+02
S78 8301.1 2229.6 -1076.5 0.0 147 9.81E+03 2.06E+02
S79 8415.3 2352.1 -1041.1 0.3 139 1.87E+04 8.56E+01
S80 8220.9 2262.6 -1135.6 -0.6 111 1.66E+04 3.33E+02
S81 8264.6 2150.6 -1087.5 -0.1 110 6.91E+04 4.23E+02
S82 8246.2 2087.5 -1176.5 0.1 66 1.82E+05 7.63E+02

83
 8550.0
N
8500.0 S58
S61 S69
S68 S67
S32 S14
S63 S25
8450.0 S16 S45 S38
S05 S07
S13
S55
S21 S79
8400.0 S47
S37 S77
S12 S01 S48
S70 S15
S71 S54 S03 S73
8350.0 S62 S56 S39
S20
X [m]

S35 S76 S19

S26 S44 S06 S74
S36 S09 S18
8300.0 S64 S78 S08
S33 S29
S60 S81 S34 S31 S02
S59
8250.0 S49 S72 S27
S46 S82 S42 S22 S23 S11
S75 S24
S57 S28 S17 S80 S10
S43 S40 S30 S53 S65
8200.0 S51 S52
S41
S04 S50
S66
8150.0

8100.0
2000.0 2050.0 2100.0 2150.0 2200.0 2250.0 2300.0 2350.0 2400.0 2450.0
Y [m]

Figure B-1. Cluster groups created in the third stage of clustering using SLINK clustering algorithm, plan view.

84
 -900.0
2000.0 2050.0 2100.0 2150.0 2200.0 2250.0 2300.0 2350.0 2400.0 2450.0
S35

S44 S56
-1000.0
S71S26 S54
S79
S14 S06 S37 S38
S78 S31 S53 S07 S77
S81
-1100.0 S60 S63 S51 S58 S65 S13
S29 S42 S67
S49 S75 S02 S01
S36 S43 S80 S11
S70 S32 S25 S09 S45
S47 S52 S21 S18
S82 S04 S17 S50
Z [m]

-1200.0 S57 S28 S05 S23 S39

S64 S62 S74
S68 S22
S46 S61
S33 S16 S24
S20
S59 S76 S12
-1300.0 S41 S48
S27 S73
S69 S10
S55 S72 S19
S66 S15
S08 S03
S30
-1400.0 S40
S34

-1500.0
Y [m]

Figure B-2. Cluster groups created in the third stage of clustering using SLINK clustering algorithm, view looking east.

85
C Seismic Risk Assessment – PASRA
a. Level 1225
Table C-1. Results of the seismic risk assessment using PASRA for mining level 1225.
Closest
MAX σ UCS MAX
GRID MAX REC group with AEC REC/AEC
[Mpa] [Mpa] ML
max REC
C12 40 180 0.09 1.1 S39 2 4.3
D11 35 180 0.31 1.1 S39 5 6.2
D13 40 180 0.11 1.1 S39 2 5.7
D14 40 180 0.09 1.1 S39 2 4.3
D15 50 180 0.23 1.1 S39 5 4.5
E14 40 75 0.34 1.1 S39 5 6.8
E15 40 180 0.31 1.1 S39 5 6.2
E16 50 180 0.28 1.1 S39 2 14.1
E17 60 180 0.32 1.1 S39 2 16.2
F15 30 110 0.23 1.1 S39 2 11.3
F16 50 180 0.31 1.1 S39 4 7.7
F17 50 180 0.23 1.1 S39 2 11.3
G16:2 35 110 0.25 1.1 S39 2 12.7
G16:25 35 110 0.28 1.1 S39 5 5.6
G17 45 180 0.32 1.1 S39 2 16.2
H17 55 180 0.14 1.1 S39 2 7.1
H18 40 180 0.25 1.1 S39 2 12.7
I15 10 75 0.34 1.1 S39 5 6.7
I18 40 75 0.34 1.1 S39 2 16.8
J1 50 180 0.23 1.1 S39 1 22.6
J14 15 110 0.34 1.1 S39 5 6.7
J15 15 110 0.28 1.1 S39 2 14.1
J17 25 180 0.21 1.1 S39 2 10.4
J18 40 180 0.09 1.1 S39 2 4.3
J3 25 75 0.28 1.1 S39 6 4.7
K12 10 75 0.28 1.1 S39 2 14.1
K13 15 180 0.34 1.1 S39 5 6.7
K14 15 180 0.34 1.1 S39 5 6.7
K15 20 180 0.34 1.1 S39 4 8.4
K16 15 180 0.31 1.1 S39 2 15.4
K2 40 180 0.14 1.1 S39 2 7.1
K3 40 180 0.28 1.1 S39 5 5.6
K4 35 110 0.23 1.1 S39 6 3.8
K5 30 110 0.28 1.1 S39 1 28.1
L12 15 75 0.41 1.1 S39 2 20.4
L13:25 10 75 0.34 1.1 S39 5 6.7
L13:4 10 75 0.34 1.1 S39 4 8.4
L3 45 180 0.23 1.1 S39 2 11.3
L4 40 180 0.28 1.1 S39 4 7.0
L4 40 180 0.31 1.1 S39 6 5.1

86
 Closest
MAX σ UCS MAX
GRID MAX REC group with AEC REC/AEC
[Mpa] [Mpa] ML
max REC
L6 35 110 0.23 1.1 S39 1 22.6
L7 25 75 0.20 1.1 S39 1 19.8
L9 30 75 0.23 1.1 S39 1 22.6
M10 40 180 0.28 1.1 S39 5 5.6
M11 40 180 0.31 1.1 S39 5 6.2
M12 40 180 0.23 1.1 S39 4 5.6
M3 55 180 0.23 1.1 S39 2 11.3
M5:4 50 180 0.14 1.1 S39 4 3.5
M5:45 50 180 0.28 1.1 S39 6 4.7
M6 50 180 0.25 1.1 S39 6 4.2
M7 50 180 0.14 1.1 S39 2 7.1
M8 50 180 0.14 1.1 S39 4 3.5
M9 40 180 0.28 1.1 S39 2 13.9
N10 50 180 0.23 1.1 S39 5 4.5
N5 45 180 0.28 1.1 S39 1 27.8
N6 50 180 0.14 1.1 S39 2 7.1
N8 50 180 0.54 1.1 S39 1 54.1
O6 55 180 0.31 1.1 S39 2 15.4
P6 55 180 0.14 1.1 S39 2 7.1
Q6 55 180 0.54 1.1 S39 2 27.1
R6 55 180 0.34 1.1 S39 2 17.1
R7 55 180 0.28 1.1 S39 1 28.1
S4 60 180 0.28 1.1 S39 2 14.1
S5 60 180 0.28 1.1 S39 2 14.1
S7 60 180 0.23 1.1 S39 1 22.6
T3 60 180 0.32 1.1 S39 3 10.8
T4 60 180 0.23 1.1 S39 3 7.5
T5:2 60 180 0.32 1.1 S39 2 16.2
T5:3 60 180 0.23 1.1 S39 3 7.5

87
b. Level 1250
Table C-2. Results of the seismic risk assessment using PASRA for the mining level 1250.
Closest
MAX σ UCS MAX
GRID MAX ML group with AEC REC/AEC
[Mpa] [Mpa] REC
max REC
B10 40 180 0.23 1.1 S39 2 11.3
B11 40 180 0.23 1.1 S39 2 11.3
B7 30 180 0.17 1.1 S39 2 8.5
B8 35 180 0.20 1.1 S39 2 9.9
B9:2 35 180 0.20 1.1 S39 2 9.9
B9:3 35 180 0.20 1.1 S39 3 6.6
C10 30 180 0.17 1.1 S39 2 8.5
C11 35 180 0.20 1.1 S39 5 4.0
C12:2 40 180 0.23 1.1 S39 2 11.3
C12:4 40 180 0.23 1.1 S39 4 5.6
C13 45 180 0.25 1.1 S39 2 12.7
C14 50 180 0.28 1.1 S39 2 14.1
C3 10 110 0.10 1.1 S39 6 1.6
C4 5 105 0.05 1.1 S39 6 0.9
C5 5 110 0.05 1.1 S39 6 0.8
C6 5 110 0.05 1.1 S39 6 0.8
C7 10 110 0.10 1.1 S39 2 4.8
C8 10 75 0.14 1.1 S39 3 4.7
C9 20 75 0.27 1.1 S39 2 13.7
D11 20 110 0.19 1.1 S39 6 3.1
D12:4 10 75 0.14 1.1 S39 4 3.5
D12:4 20 180 0.11 1.1 S39 4 2.9
D13 35 180 0.20 1.1 S39 2 9.9
D14 40 180 0.23 1.1 S39 2 11.3
D15 45 180 0.25 1.1 S39 2 12.7
D2 10 75 0.14 1.1 S39 4 3.5
D8 5 110 0.05 1.1 S39 4 1.3
D8 5 110 0.05 1.1 S39 6 0.8
E1 10 75 0.14 1.1 S39 4 3.5
E14 25 180 0.14 1.1 S39 3 4.7
E15 35 180 0.20 1.1 S39 2 9.9
E17 45 180 0.25 1.1 S39 2 12.7
E2 10 75 0.14 1.1 S39 4 3.5
F1 10 75 0.14 1.1 S39 5 2.8
F13 10 75 0.14 1.1 S39 2 7.1
F16 40 180 0.23 1.1 S39 2 11.3
F17 50 180 0.28 1.1 S39 2 14.1
G1 15 180 0.09 1.1 S39 2 4.3
G17 45 180 0.25 1.1 S39 2 12.7
H1 30 180 0.17 1.1 S39 2 8.5
H17 50 180 0.28 1.1 S39 2 14.1

88
 Closest
MAX σ UCS MAX
GRID MAX ML group with AEC REC/AEC
[Mpa] [Mpa] REC
max REC
H2 10 75 0.14 1.1 S39 5 2.8
I1 45 180 0.25 1.1 S39 2 12.7
I17 10 180 0.06 1.1 S39 5 1.2
I2 40 75 0.54 1.1 S39 5 10.8
I3 10 75 0.14 1.1 S39 2 7.1
J1:2 50 180 0.28 1.1 S39 2 14.1
J1:45 50 180 0.28 1.1 S39 6 4.7
J13 10 75 0.14 1.1 S39 6 2.4
J16 35 180 0.20 1.1 S39 5 4.0
J3:2 30 180 0.17 1.1 S39 2 8.5
J3:4 30 180 0.17 1.1 S39 4 4.2
K12 10 75 0.14 1.1 S39 6 2.4
K15 10 75 0.14 1.1 S39 5 2.8
K2 50 180 0.28 1.1 S39 5 5.6
K3:2 35 180 0.20 1.1 S39 2 9.9
K3:25 35 180 0.20 1.1 S39 5 4.0
K5 20 110 0.19 1.1 S39 2 9.4
L12 10 75 0.14 1.1 S39 5 2.8
L13 10 75 0.14 1.1 S39 2 7.1
L14 10 75 0.14 1.1 S39 2 7.1
L2 55 180 0.31 1.1 S39 2 15.4
L3 60 180 0.34 1.1 S39 5 6.7
L4 60 180 0.34 1.1 S39 4 8.4
L5 30 180 0.17 1.1 S39 2 8.5
L7 45 110 0.41 1.1 S39 2 20.7
L8 30 110 0.28 1.1 S39 4 7.0
L9 30 110 0.28 1.1 S39 3 9.3
M10 35 180 0.20 1.1 S39 2 9.9
M11 35 180 0.20 1.1 S39 5 4.0
M12 30 180 0.17 1.1 S39 6 2.8
M13 10 75 0.14 1.1 S39 2 7.1
M2 50 180 0.28 1.1 S39 2 14.1
M3 55 180 0.31 1.1 S39 2 15.4
M4 55 180 0.31 1.1 S39 2 15.4
M5:2 45 180 0.25 1.1 S39 2 12.7
M5:4 45 180 0.25 1.1 S39 4 6.3
M6 40 180 0.23 1.1 S39 2 11.3
M7 40 180 0.23 1.1 S39 4 5.6
M8 40 180 0.23 1.1 S39 2 11.3
M9:2 35 180 0.20 1.1 S39 2 9.9
M9:4 35 180 0.20 1.1 S39 4 4.9
N12 60 180 0.34 1.1 S39 5 6.7
N13 65 180 0.36 1.1 S39 2 18.2
N5 50 180 0.28 1.1 S39 1 28.1
N8 50 180 0.28 1.1 S39 1 28.1

89
 Closest
MAX σ UCS MAX
GRID MAX ML group with AEC REC/AEC
[Mpa] [Mpa] REC
max REC
N9 50 180 0.28 1.1 S39 5 5.6
O12 65 180 0.36 1.1 S39 2 18.2
O13 70 180 0.39 1.1 S39 5 7.8
P10 60 180 0.34 1.1 S39 1 33.7
P11 65 180 0.36 1.1 S39 2 18.2

90
c. Level 1275

Table C-3. Results of seismic risk assessment using PASRA for mining level 1275.
Closest
MAX σ UCS
GRID MAX REC MAX ML group with AEC REC/AEC
[Mpa] [Mpa]
max REC
E0 25 75 0.34 1.1 S39 5 6.8
F0 35 180 0.20 1.1 S39 5 4.0
F1 15 75 0.21 1.1 S39 2 10.4
G0 25 180 0.14 1.1 S39 5 2.8
H0 25 180 0.14 1.1 S39 2 7.1
H1 25 110 0.23 1.1 S39 2 11.6
I0 35 180 0.20 1.1 S39 2 9.9
I1:2 25 180 0.14 1.1 S39 2 7.1
I1:25 25 180 0.14 1.1 S39 5 2.8
J1:2 30 180 0.17 1.1 S39 2 8.5
J1:45 30 180 0.17 1.1 S39 6 2.8
J2:25 25 180 0.14 1.1 S39 5 2.8
J2:2 25 180 0.14 1.1 S39 2 7.1
J13 10 75 0.14 1.1 S39 5 2.8
K2 50 180 0.28 1.1 S39 6 4.7
K3 80 110 0.73 1.1 S39 5 14.7
K4 25 110 0.23 1.1 S39 5 4.7
K5 25 110 0.23 1.1 S39 6 3.9
K10 25 110 0.23 1.1 S39 6 3.9
K12 10 75 0.14 1.1 S39 4 3.5
K13 10 75 0.14 1.1 S39 5 2.8
L2:2 65 180 0.36 1.1 S39 2 18.2
L2:45 65 180 0.36 1.1 S39 6 6.1
L3 70 180 0.39 1.1 S39 6 6.5
L4:45 60 180 0.34 1.1 S39 6 5.6
L4:25 60 180 0.34 1.1 S39 5 6.7
L5:45 50 180 0.28 1.1 S39 6 4.7
L5:2 50 180 0.28 1.1 S39 2 14.1
L6 35 180 0.20 1.1 S39 5 4.0
L9:4 35 180 0.20 1.1 S39 4 4.9
L9:45 35 180 0.20 1.1 S39 6 3.3
L10 30 180 0.17 1.1 S39 6 2.8
L11 25 180 0.14 1.1 S39 4 3.6
L13 10 180 0.06 1.1 S39 2 2.9
M2 60 180 0.34 1.1 S39 2 16.8
M3 60 180 0.34 1.1 S39 2 16.8
M4 60 180 0.34 1.1 S39 2 16.8
M5 60 180 0.34 1.1 S39 6 5.6
M6 60 180 0.34 1.1 S39 6 5.6
M7:45 60 180 0.34 1.1 S39 6 5.6

91
 Closest
MAX σ UCS
GRID MAX REC MAX ML group with AEC REC/AEC
[Mpa] [Mpa]
max REC
M7:4 60 180 0.34 1.1 S39 4 8.4
M8 55 180 0.31 1.1 S39 5 6.2
M9:45 50 180 0.28 1.1 S39 6 4.7
M9:25 45 180 0.25 1.1 S39 5 5.1
M10 40 180 0.23 1.1 S39 6 3.8
M11 40 180 0.23 1.1 S39 6 3.8
M12 40 180 0.23 1.1 S39 6 3.8
N2 60 180 0.34 1.1 S39 2 16.8
N6 60 180 0.34 1.1 S39 2 16.8
N7 60 180 0.34 1.1 S39 1 33.7
N9 60 180 0.34 1.1 S39 5 6.7
N10 60 180 0.34 1.1 S39 5 6.7
N11 60 180 0.34 1.1 S39 6 5.6
O6 60 180 0.34 1.1 S39 2 16.8
O10 60 180 0.34 1.1 S39 4 8.4
P6 60 180 0.34 1.1 S39 2 16.8
Q6 60 180 0.34 1.1 S39 2 16.8
R6 60 180 0.34 1.1 S39 2 16.8
R7:2 60 180 0.34 1.1 S39 2 16.8
R7:1 60 180 0.34 1.1 S39 1 33.7
R8 60 180 0.34 1.1 S39 1 33.7
S6 60 180 0.34 1.1 S39 2 16.8
S7 60 180 0.34 1.1 S39 1 33.7
T6 60 180 0.34 1.1 S39 2 16.8

92
D Seismic Risk Assessment – QSHRAF
a. Level 1225

Table D-1. Results of seismic risk assessment using QSHRAF for mining level 1225.
MAX σ UCS Closest
GRID MAX PPV MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] cluster group
C12 40 180 0.05 0.3 S62 22 5 6 1.5 17.8 0.9 1000 VL
D11 35 180 0.09 0.3 S62 19 6 6 0.5 38.9 3.3 1000 VL
D13 40 180 0.07 0.2 S68 22 5 6.3 1.5 18.7 1.2 1000 VL
D14 40 180 0.18 0.2 S68 22 5 9 1.5 26.7 4.7 1000 VL
D15 50 180 0.18 0.2 S68 28 6 6 1.5 18.5 3.3 1000 VL
E14 30 75 0.11 0.2 S68 40 6 7.2 1 48.0 5.1 1000 VL
E15 30 180 0.14 0.2 S68 17 6 9.6 1.5 17.8 2.5 1000 VL
E16 50 180 0.25 0.2 S68 28 5 6 1.5 22.2 5.5 1000 VL
E17 60 180 0.18 0.2 S68 33 5 6 1.5 26.7 4.9 3000 VL
F15 30 110 0.32 0.2 S68 27 5 6 0.5 65.5 21.0 1000 VL
F16 50 180 0.30 0.2 S68 28 10 11.8 1.5 21.9 6.5 1000 VL
F17 60 180 0.31 0.2 S68 33 5 6.5 1.5 28.9 9.0 1000 VL
G16:2 35 110 0.39 0.2 S68 32 5 6.6 1 42.0 16.2 1000 VL
G16:25 35 110 0.26 0.2 S68 32 6 6.6 0.5 70.0 18.0 1000 VL
G17 45 180 0.14 0.2 S68 25 5 6.6 1.5 22.0 3.2 1000 VL
H17 55 180 0.09 0.1 S61 31 5 6.2 1.5 25.3 2.3 1000 VL
H18 60 180 0.14 0.1 S61 33 5 6 1.5 26.7 3.8 1000 VL
I15 10 75 0.11 0.1 S61 13 6 6.2 0.5 27.6 3.1 1000 VL
I18 30 75 0.15 0.1 S61 40 5 6.5 1.5 34.7 5.1 1000 VL
J3 25 75 0.31 0.5 S24 33 10 11.2 0.5 74.7 23.0 1000 VL

93
 MAX σ UCS Closest
GRID MAX PPV MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] cluster group
J14 10 110 0.12 1.1 S39 9 6 7 1 10.6 1.3 1000 VL
J15 10 110 0.13 0.3 S05 9 5 6.2 1.5 7.5 1.0 1000 VL
J17 25 180 0.13 0.1 S61 14 5 6.2 1.5 11.5 1.5 1000 VL
J18 40 180 0.23 0.1 S61 22 5 8.6 1.5 25.5 5.8 1000 VL
J1 50 180 0.51 0.5 S24 28 4 5 1.5 23.1 11.8 100 VL
K2 40 180 1.91 0.5 S24 22 5 8.8 1.5 26.1 49.8 2000 L
K3 30 180 1.05 0.5 S24 17 6 10 1.5 18.5 19.5 2000 VL
K4 25 110 0.34 -0.3 S22 23 10 6.6 0.5 30.0 10.3 3000 VL
K5 25 110 0.48 0.5 S23 23 4 10 0.5 113.6 54.4 100 VL
K12 10 75 0.34 1.1 S39 13 5 6 0.5 32.0 10.8 1000 VL
K13 15 180 0.47 0.3 S05 8 6 6 1 8.3 3.9 1000 VL
K14 15 180 2.26 0.3 S05 8 6 5.6 1.5 5.2 11.7 1000 VL
K15 15 180 1.34 0.3 S05 8 10 9.6 1.5 5.3 7.1 1000 VL
K16 15 180 0.20 0.3 S05 8 5 6.3 1.5 7.0 1.4 1000 VL
L3 45 180 0.54 0.5 S23 25 5 6.8 1.5 22.7 12.3 3000 VL
L4:45 40 180 1.02 0.5 S23 22 10 9.4 1.5 13.9 14.2 3000 VL
L4:4 40 180 0.60 0.5 S23 22 10 5.4 1 12.0 7.2 1000 VL
L6 25 110 0.31 0.5 S24 23 4 5.2 1 29.5 9.1 1000 VL
L7 20 75 0.28 0.5 S24 27 4 9.9 0.5 132.0 36.7 100 VL
L9 20 75 0.39 0.9 S20 27 4 9 0.5 120.0 46.5 1000 L
L12 10 75 0.81 1.1 S39 13 5 5.2 1.5 9.2 7.5 1000 VL
L13:25 10 75 0.87 0.3 S05 13 6 5.5 1.5 8.1 7.1 1000 VL
L13:4 10 75 0.55 0.3 S05 13 10 9.2 1.5 8.2 4.5 1000 VL
M3 60 180 0.49 0.5 S23 33 5 6.6 1.5 29.3 14.5 3000 VL
M5:45 50 180 1.11 0.5 S23 28 10 8 1.5 14.8 16.4 2000 VL
M5:4 50 180 1.85 0.5 S23 28 10 5.2 1.5 9.6 17.8 2000 VL

94
 MAX σ UCS Closest
GRID MAX PPV MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] cluster group
M6 45 180 1.23 0.5 S23 25 10 10.5 1.5 17.5 21.4 4000 VL
M7 45 180 0.88 0.9 S20 25 5 5.5 1.5 18.3 16.2 4000 VL
M8 50 180 1.31 0.9 S20 28 10 10.6 1.5 19.6 25.7 4000 L
M9 40 180 4.82 0.9 S20 22 5 8.5 1.5 25.2 121.3 3000 M
M10 40 180 1.28 1.1 S39 22 6 10.5 1.5 25.9 33.2 3000 L
M11 40 180 1.31 1.1 S39 22 6 5.5 1.5 13.6 17.8 1000 VL
M12 40 180 1.57 1.1 S39 22 10 9.5 1.5 14.1 22.2 1000 VL
N5 45 180 3.42 0.5 S23 25 4 5.4 1.5 22.5 76.9 100 L
N6 50 180 1.01 0.5 S23 28 5 5.5 1.5 20.4 20.5 4000 VL
N8 50 180 2.40 0.9 S20 28 4 5.2 1.5 24.1 57.9 100 VL
N10 50 180 2.02 1.1 S39 28 6 6.1 1.5 18.8 38.1 3000 L
O6 60 180 1.34 1.1 S39 33 5 5.4 1.5 24.0 32.2 4000 L
P6 60 180 1.00 1.1 S39 33 5 6.2 1.5 27.6 27.4 4000 L
Q6 60 180 0.39 1.1 S39 33 5 6 1.5 26.7 10.5 4000 VL
R6 60 180 0.17 1.1 S39 33 5 9.5 1.5 42.2 7.3 4000 VL
R7 60 180 0.08 1.1 S39 33 4 10.6 1.5 58.9 4.6 4000 VL
S4 60 180 0.07 1.1 S39 33 5 6 1.5 26.7 1.9 1000 VL
S5 60 180 0.08 1.1 S39 33 5 11.4 1.5 50.7 4.0 1000 VL
S7 60 180 0.21 0.9 S20 33 4 9.3 1.5 51.7 10.9 4000 VL
T3 60 180 0.05 1.1 S39 33 5 10.6 1.5 47.1 2.4 100 VL
T4 60 180 0.06 1.1 S39 33 5 6 1.5 26.7 1.6 100 VL
T5:2 60 180 0.07 1.1 S39 33 5 6 1.5 26.7 1.9 1000 VL
T5:3 60 180 0.07 0.9 S20 33 5 12.5 1.5 55.6 4.2 1000 VL

95
 Peak Particle Velocity, PPV [m/s]

0.00
1.00
2.00
3.00
4.00
5.00
6.00
C12
D11
D13
D14
D15
E14
E15
E16
E17
F15
F16
F17
G16:2
G16:25
G17
H17
H18
I15
I18
J3
J14
J15
J17
J18
J1
K2
K3
K4
K5
K12
K13
K14
K15
K16
L3
L4:45
PPV - Level 1225

L4:4
L6
L7
L9
L12
L13:25
L13:4
M3
M5:45
M5:4
M6
M7
M8
M9
M10
M11
M12
N5
N6
N8
N10
O6
P6
Q6
R6
R7
S4
S5
S7
T3
T4
T5:2
T5:3
96
 Excavation Vulnerability Potential, EVP

20.0
40.0
60.0
80.0
100.0
120.0
140.0

0.0
C12
D11
D13
D14
D15
E14
E15
E16
E17
F15
F16
F17
G16:2
G16:25
G17
H17
H18
I15
I18
J3
J14
J15
J17
J18
J1
K2
K3
K4
K5
K12
K13
K14
K15
K16
L3
L4:45
EVP - Level 1225

L4:4
L6
L7
L9
L12
L13:25
L13:4
M3
M5:45
M5:4
M6
M7
M8
M9
M10
M11
M12
N5
N6
N8
N10
O6
P6
Q6
R6
R7
S4
S5
S7
T3
T4
T5:2
T5:3
97
 Rocbkurst damage potential, RDP

10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
110.0
120.0
130.0

0.0
M9
N5
N8
K5
K2
L9
N10
L7
M10
O6
P6
M8
J3
M12
M6
F15
N6
K3
G16:25
M5:4
M11
M5:45
G16:2
M7
M3
L4:45
L3
J1
damage zone

K14
S7
K12
Q6
K4
L6
F17
L12
RDP - Level 1225

R6
L4:4
K15
L13:25
F16
J18
E16
I18
E14
E17
D14
R7
L13:4
T5:3
S5
K13
H18
D11
D15
G17
I15
E15
T3
H17
T5:2
S4
T4
J17
K16
J14
D13
J15
C12
98
 EVP vs PPV - Level 1225
7.00

6.00

5.00
Peak Particle Velocity, PPV [m/s]

M9

4.00

N5
3.00

N10N8
2.00 K2
M10

1.00
K5
L9
L7
0.00
1.0 10.0 100.0 1000.0
Excavation Vulnerability Potential, EVP

99
 b. Level 1250
Table D-2. Results of seismic risk assessment using QSHRAF for mining level 1250.
MAX σ UCS MAX PPV Closest
GRID MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] [m/s] cluster group
B7 30 180 0.31 0.1 S33 17 5 6.5 1.5 14.4 4.4 1000 VL
B8 35 180 0.19 0.1 S33 19 5 6.5 1.5 16.9 3.1 1000 VL
B9:2 35 180 0.22 0.3 S62 19 5 9.2 1.5 23.9 5.3 2000 VL
B9:3 35 180 0.42 0.3 S62 19 5 6.5 1.5 16.9 7.1 2000 VL
B10 40 180 0.24 0.3 S62 22 5 10.3 1.5 30.5 7.3 1000 VL
B11 40 180 0.12 0.3 S62 22 5 6 1.5 17.8 2.1 1000 VL
C3 10 110 0.09 0.1 S33 9 10 9 1.5 5.5 0.5 100 VL
C4 5 105 0.18 0.1 S33 5 10 6.4 1.5 2.0 0.4 100 VL
C5 5 110 0.71 0.1 S33 5 10 7 0.5 6.4 4.5 100 VL
C6 5 110 0.20 0.1 S33 5 10 6.4 0.5 5.8 1.2 100 VL
C7 10 110 0.31 0.1 S33 9 5 6.5 1.5 7.9 2.5 1000 VL
C8 10 75 0.39 0.1 S33 13 5 9.6 0.5 51.2 19.9 1000 VL
C9 20 75 0.41 0.3 S62 27 5 6 0.5 64.0 26.4 1000 L
C10 30 180 0.28 0.3 S62 17 5 6 1.5 13.3 3.7 1000 VL
C11 35 180 0.10 0.3 S62 19 6 6 0.5 38.9 4.0 1000 VL
C12:4 40 180 0.05 0.3 S62 22 10 7 1.5 10.4 0.6 1000 VL
C12:2 40 180 0.05 0.3 S62 22 5 9.8 1.5 29.0 1.6 1000 VL
C13 45 180 0.05 0.2 S68 25 5 6.6 1.5 22.0 1.1 1000 VL
C14 50 180 0.08 0.2 S68 28 5 5.2 1.5 19.3 1.6 1000 VL
D2 10 75 0.07 0 S28 13 10 6.6 0.5 17.6 1.2 100 VL
D8 5 110 0.14 0.1 S33 5 10 6.6 0.5 6.0 0.8 100 VL
D8 5 110 0.13 0.1 S33 5 10 7 0.5 6.4 0.8 100 VL
D11 20 110 0.06 0.3 S62 18 10 8.2 0.5 29.8 1.7 100 VL

100
 MAX σ UCS MAX PPV Closest
GRID MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] [m/s] cluster group
D12:4 20 180 0.05 0.3 S62 11 10 5.5 0.5 12.2 0.7 1000 VL
D12:4 10 75 0.04 0.2 S68 13 10 5.5 0.5 14.7 0.6 1000 VL
D13 35 180 0.06 0.2 S68 19 5 6 1.5 15.6 0.9 1000 VL
D14 40 180 0.10 0.2 S68 22 5 8.6 1.5 25.5 2.7 1000 VL
D15 45 180 0.08 0.2 S68 25 5 9 1.5 30.0 2.4 1000 VL
E1 10 75 0.08 0.3 S04 13 10 6 0.5 16.0 1.3 100 VL
E2 10 75 0.06 0.3 S04 13 10 6.5 0.5 17.3 1.0 100 VL
E14 25 180 0.07 0.2 S68 14 5 6.7 1 18.6 1.2 1000 VL
E15 35 180 0.09 0.2 S68 19 5 9.6 1.5 24.9 2.1 1000 VL
E17 45 180 0.08 0.2 S68 25 5 10.3 1.5 34.3 2.7 3000 VL
F1 10 75 0.10 0.3 S04 13 6 7 0.5 31.1 3.1 1000 VL
F13 10 75 0.06 0.2 S68 13 5 5.6 0.5 29.9 1.8 100 VL
F16 40 180 0.15 0.2 S68 22 5 5.6 1.5 16.6 2.5 1000 VL
F17 50 180 0.11 0.2 S68 28 5 10.2 1.5 37.8 4.1 14000 M
G1 15 180 0.06 0.3 S04 8 5 6.4 1 10.7 0.6 1000 VL
G17 45 180 0.09 0.2 S68 25 5 6.5 1.5 21.7 1.9 1000 VL
H1 30 180 0.05 0.1 S50 17 5 6.8 1 22.7 1.2 1000 VL
H2 10 75 0.06 0.5 S24 13 6 6 0.5 26.7 1.6 1000 VL
H17 50 180 0.19 0.1 S61 28 5 6.4 1.5 23.7 4.5 1000 VL
I1 45 180 0.07 0.5 S24 25 5 6 1.5 20.0 1.4 1000 VL
I2 40 75 0.08 0.5 S24 53 6 6 1 53.3 4.1 1000 VL
I3 10 75 0.09 0.5 S24 13 5 7.6 0.5 40.5 3.5 1000 VL
I17 10 180 0.27 0.1 S61 6 6 7.8 1.5 4.8 1.3 1000 VL
J1:45 50 180 0.12 0.5 S24 28 10 11 1.5 20.4 2.5 1000 VL
J1:2 50 180 0.29 0.5 S24 28 5 5 1.5 18.5 5.3 1000 VL
J3:2 30 180 0.16 0.5 S24 17 5 10.2 0.5 68.0 11.0 1000 VL

101
 MAX σ UCS MAX PPV Closest
GRID MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] [m/s] cluster group
J3:4 30 180 0.12 0.5 S24 17 10 6 0.5 20.0 2.4 1000 VL
J13 10 75 0.14 -0.2 S16 13 10 8 0.5 21.3 2.9 1000 VL
J16 35 180 0.17 0.1 S61 19 6 6.5 1.5 14.0 2.4 1000 VL
K2 50 180 0.69 0.5 S24 28 6 6.5 1.5 20.1 13.8 1000 VL
K3:2 35 180 0.47 0.5 S24 19 5 6.5 1.5 16.9 7.9 1000 VL
K3:25 35 180 0.37 0.5 S24 19 6 10 1.5 21.6 8.1 1000 VL
K5 20 110 0.28 -0.3 S22 18 5 8 0.5 58.2 16.4 1000 VL
K12 10 75 0.42 0.2 S12 13 10 9.6 0.5 25.6 10.7 1000 VL
K15 10 75 0.16 -0.2 S16 13 6 5.6 1.5 8.3 1.3 1000 VL
L2 55 180 0.69 0.5 S24 31 5 7 1.5 28.5 19.7 3000 VL
L3 60 180 1.00 0.5 S24 33 6 10 1.5 37.0 37.2 3000 L
L4 60 180 1.55 0.5 S24 33 10 10.4 1.5 23.1 35.8 3000 L
L5 30 180 0.57 0.5 S24 17 5 12 0.5 80.0 45.4 1000 L
L7 45 110 1.51 0.9 S20 41 5 6 0.5 98.2 147.8 1000 M
L8 30 110 0.84 0.9 S20 27 10 5.6 0.5 30.5 25.7 1000 L
L9 30 110 0.27 0.9 S20 27 5 12.4 0.5 135.3 36.7 1000 L
L12 10 75 0.70 0.2 S12 13 6 9.1 1.5 13.5 9.4 7000 L
L13 10 75 0.84 0.2 S12 13 5 5.6 1.5 10.0 8.3 1000 VL
L14 10 75 0.26 0.2 S12 13 5 6 1.5 10.7 2.7 1000 VL
M2 50 180 0.55 0.5 S24 28 5 6 1.5 22.2 12.2 3000 VL
M3 55 180 0.80 0.5 S24 31 5 6.5 1.5 26.5 21.1 3000 VL
M4 55 180 1.73 0.5 S24 31 5 6.5 1.5 26.5 45.8 3000 L
M5:2 45 180 0.72 0.5 S24 25 5 5.6 1.5 18.7 13.5 1000 VL
M5:4 45 180 0.92 0.5 S23 25 10 9.8 1.5 16.3 15.0 1000 VL
M6 40 180 2.23 0.5 S24 22 5 5.8 1.5 17.2 38.3 1000 L
M7 40 180 1.28 0.9 S20 22 10 9.6 1.5 14.2 18.2 1000 VL

102
 MAX σ UCS MAX PPV Closest
GRID MAX ML E1 E2 E3 E4 EVP RDP Exposure SEISMIC RISK RATING
[Mpa] [Mpa] [m/s] cluster group
M8 40 180 1.29 0.9 S20 22 5 9.2 1.5 27.3 35.2 4000 L
M9:4 35 180 1.12 0.9 S20 19 10 10.6 1.5 13.7 15.4 3000 VL
M9:2 35 180 2.38 0.9 S20 19 5 10 1.5 25.9 61.6 3000 L
M10 35 180 2.61 0.2 S12 19 5 5.2 1.5 13.5 35.2 2000 L
M11 35 180 0.92 0.2 S12 19 6 5.8 1.5 12.5 11.6 2000 VL
M12 30 180 2.57 0.2 S12 17 10 10.2 1.5 11.3 29.1 4000 L
M13 10 75 0.70 0.2 S12 13 5 5.5 1.5 9.8 6.9 1000 VL
N5 50 180 6.12 0.5 S24 28 4 5.2 1.5 24.1 147.3 100 M
N8 50 180 6.08 0.9 S20 28 4 5.6 1.5 25.9 157.6 100 M
N9 50 180 1.55 0.9 S20 28 6 5.6 1.5 17.3 26.8 3000 L
N12 60 180 0.94 1.1 S39 33 6 7.8 1.5 28.9 27.2 4000 L
N13 65 180 0.32 1.1 S39 36 5 11 1.5 53.0 16.9 4000 VL
O12 65 180 0.92 0.9 S20 36 5 7 1.5 33.7 31.0 4000 L
O13 70 180 0.28 0.9 S20 39 6 11 1.5 47.5 13.5 4000 VL
P10 60 180 1.17 0.9 S20 33 4 5 1.5 27.8 32.6 4000 L
P11 65 180 0.97 0.9 S20 36 4 6 1.5 36.1 35.2 4000 L

103
 Peak Particle Velocity, PPV [m/s]

0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
B7
B8
B9:2
B9:3
B10
B11
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12:4
C12:2
C13
C14
D2
D8
D8
D11
D12:4
D12:4
D13
D14
D15
E1
E2
E14
E15
E17
F1
F13
F16
F17
G1
G17
H1
H2
H17
I1
I2
I3
I17
J1:45
PPV - Level 1250

J1:2
J3:2
J3:4
J13
J16
K2
K3:2
K3:25
K5
K12
K15
L2
L3
L4
L5
L7
L8
L9
L12
L13
L14
M2
M3
M4
M5:2
M5:4
M6
M7
M8
M9:4
M9:2
M10
M11
M12
M13
N5
N8
N9
N12
N13
O12
O13
P10
P11
104
 Excavation Vulnerability Potential, EVP

20.0
40.0
60.0
80.0

0.0
100.0
120.0
140.0
160.0
B7
B8
B9:2
B9:3
B10
B11
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12:4
C12:2
C13
C14
D2
D8
D8
D11
D12:4
D12:4
D13
D14
D15
E1
E2
E14
E15
E17
F1
F13
F16
F17
G1
G17
H1
H2
H17
I1
I2
I3
I17
J1:45
EVP - Level 1250

J1:2
J3:2
J3:4
J13
J16
K2
K3:2
K3:25
K5
K12
K15
L2
L3
L4
L5
L7
L8
L9
L12
L13
L14
M2
M3
M4
M5:2
M5:4
M6
M7
M8
M9:4
M9:2
M10
M11
M12
M13
N5
N8
N9
N12
N13
O12
O13
P10
P11
105
 Rockburst damage potential, RDP

10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0

0.0
100.0
110.0
120.0
130.0
140.0
150.0
160.0
170.0
N8
L7
N5
M9:2
M4
L5
M6
L3
L9
L4
M10
P11
M8
P10
O12
M12
N12
N9
C9
L8
M3
C8
L2
M7
N13
K5
M9:4
M5:4
K2
O13
M5:2
M2
M11
J3:2
K12
L12
L13
K3:25
K3:2
B10
damage zone

B9:3
M13
J1:2
B9:2
H17
C5
B7
I2
RDP - Level 1250

F17
C11
C10
I3
B8
F1
J13
L14
E17
D14
J1:45
F16
C7
D15
J3:4
J16
E15
B11
G17
F13
D11
C14
C12:2
H2
I1
K15
I17
E1
E14
H1
D2
C6
C13
E2
D13
D8
D8
D12:4
G1
D12:4
C12:4
C3
C4
106
 EVP vs PPV - Level 1250
7.00

N5
6.00 N8

5.00
Peak Particle Velocity, PPV [m/s]

4.00

3.00

M9:2
2.00
M4
L7

1.00
L5
L9
0.00
1.0 10.0 100.0 1000.0
Excavation Vulnerability Potential, EVP

107
 c. Level 1275
Table D-3. Results of seismic risk assessment using QSHRAF for mining level 1275.
MAX Closest
MAX σ UCS MAX SEISMIC RISK
GRID PPV cluster E1 E2 E3 E4 EVP RDP Exposure
[MPa] [MPa] ML RATING
[m/s] group
E0 15 75 0.08 0.3 S04 20 6 6.6 0.5 44.0 3.6 1000 VL
F0 25 180 0.08 0.3 S04 14 6 6.2 1.5 9.6 0.8 1000 VL
F1 35 75 0.07 0.3 S04 47 5 6.6 0.5 123.2 8.0 1000 VL
G0 35 180 0.05 0.3 S04 19 6 11 1.5 23.8 1.3 1000 VL
H0 40 180 0.04 0.6 S10 22 5 6.5 1.5 19.3 0.8 1000 VL
H1 25 110 0.05 0.1 S41 23 5 5.6 0.5 50.9 2.7 1000 VL
I0 40 180 0.06 0.1 S41 22 5 11 1.5 32.6 1.9 1000 VL
I1:2 25 180 0.07 0.1 S41 14 5 6.3 1.5 11.7 0.8 1000 VL
I1:25 25 180 0.23 0.1 S41 14 6 6.2 0.5 28.7 6.7 1000 VL
J1:2 50 180 0.26 0.1 S41 28 5 5.2 1.5 19.3 5.0 1000 VL
J1:45 45 180 0.38 0.1 S41 25 10 11 1.5 18.3 7.0 1000 VL
J2:25 30 180 0.36 0.1 S41 17 6 6.4 0.5 35.6 12.7 1000 VL
J2:2 30 180 0.17 0.1 S41 17 5 6 0.5 40.0 6.6 1000 VL
J13 15 75 0.08 1.1 S39 20 6 7 0.5 46.7 3.7 100 VL
K2 50 180 0.31 0.1 S41 28 10 9.4 0.5 52.2 16.4 1000 VL
K3 65 110 0.51 0.5 S24 59 6 6.5 0.5 128.0 64.8 1000 L
K4 30 110 0.38 0.5 S24 27 6 6 0.5 54.5 20.9 1000 VL
K5 35 110 0.21 0.5 S24 32 10 6.1 0.5 38.8 8.1 1000 VL
K10 25 110 0.82 0.2 S12 23 10 6 0.5 27.3 22.3 1000 VL
K12 15 75 0.25 0.2 S12 20 10 6.3 0.5 25.2 6.2 1000 VL
K13 15 75 0.10 1.1 S39 20 6 6 1.5 13.3 1.3 1000 VL
L2:2 65 180 0.39 0.5 S24 36 5 7 1.5 33.7 13.3 1000 VL

108
 MAX Closest
MAX σ UCS MAX SEISMIC RISK
GRID PPV cluster E1 E2 E3 E4 EVP RDP Exposure
[MPa] [MPa] ML RATING
[m/s] group
L2:45 85 180 0.69 0.5 S24 47 10 6.6 1 31.2 21.5 1000 VL
L3 85 180 1.21 0.5 S24 47 10 9.4 1 44.4 53.6 2000 L
L4:45 65 180 0.85 0.5 S24 36 10 10 1.5 24.1 20.5 1000 VL
L4:25 60 180 0.47 0.5 S24 33 6 6 1 33.3 15.7 1000 VL
L5:45 50 180 0.42 0.5 S24 28 10 10.7 0.5 59.4 24.9 1000 VL
L5:2 50 180 0.27 -0.1 S27 28 5 6 0.5 66.7 17.9 1000 VL
L6 45 180 0.35 0.9 S20 25 6 5.4 0.5 45.0 15.6 1000 VL
L9:4 45 180 0.22 0.9 S08 25 10 10.6 0.5 53.0 11.4 2000 VL
L9:45 40 180 0.33 0 S76 22 10 6.4 0.5 28.4 9.5 2000 VL
L10 45 180 0.34 0.2 S12 25 10 6.2 1 15.5 5.3 2000 VL
L11 25 180 0.72 0.2 S12 14 10 6.1 1 8.5 6.1 1000 VL
L13 15 180 0.14 0.2 S12 8 5 7.6 1.5 8.4 1.2 1000 VL
M2 65 180 1.03 0.5 S24 36 5 12 1.5 57.8 59.5 2000 L
M3 65 180 2.47 0.5 S24 36 5 6.5 1.5 31.3 77.3 2000 M
M4 80 180 0.76 0.5 S24 44 5 5.2 1.5 30.8 23.3 2000 VL
M5 60 180 0.36 -0.1 S27 33 10 5.5 1.5 12.2 4.4 2000 VL
M6 60 180 1.49 0.9 S20 33 10 12 1.5 26.7 39.8 4000 L
M7:45 60 180 0.98 0.9 S20 33 10 9.4 1.5 20.9 20.4 2000 VL
M7:4 60 180 1.98 0.9 S20 33 10 5.2 1.5 11.6 22.9 2000 VL
M8 55 180 0.42 0.9 S20 31 6 5.4 1.5 18.3 7.7 2000 VL
M9:45 50 180 0.60 0.2 S12 28 10 13 1.5 24.1 14.3 2000 VL
M9:25 45 180 0.45 0.2 S12 25 6 13 1.5 36.1 16.4 2000 VL
M10 50 180 0.85 0.2 S12 28 10 10.5 1.5 19.4 16.4 3000 VL
M11 60 180 0.93 0.2 S12 33 10 12.2 1.5 27.1 25.3 3000 L
M12 45 180 0.30 0.2 S12 25 10 5.4 1.5 9.0 2.7 1000 VL

109
 MAX Closest
MAX σ UCS MAX SEISMIC RISK
GRID PPV cluster E1 E2 E3 E4 EVP RDP Exposure
[MPa] [MPa] ML RATING
[m/s] group
N2 60 180 0.66 0.5 S24 33 5 6.3 1.5 28.0 18.4 1000 VL
N6 60 180 1.67 0.9 S20 33 5 5.5 1.5 24.4 40.9 4000 L
N7 60 180 1.57 0.9 S20 33 4 8.6 1.5 47.8 75.2 100 L
N9 60 180 1.15 0.2 S12 33 6 6 1.5 22.2 25.7 1000 L
N10 60 180 1.51 0.2 S12 33 6 8.5 1.5 31.5 47.5 1000 L
N11 65 180 0.70 0.2 S12 36 10 6 1.5 14.4 10.2 3000 VL
O6 60 180 0.60 0.9 S20 33 5 6 1.5 26.7 16.1 4000 VL
O10 65 180 1.13 0.2 S12 36 10 6 1.5 14.4 16.3 3000 VL
P6 65 180 0.19 1.1 S39 36 5 6.1 1.5 29.4 5.5 4000 VL
Q6 65 180 0.16 1.1 S39 36 5 6 1.5 28.9 4.5 4000 VL
R6 65 180 0.10 1.1 S39 36 5 6.6 1.5 31.8 3.3 4000 VL
R7:2 65 180 0.13 0.9 S20 36 5 11 1.5 53.0 7.1 4000 VL
R7:1 65 180 0.16 0.9 S20 36 4 10 1.5 60.2 9.8 4000 VL
R8 65 180 0.09 0.9 S08 36 4 5.8 1.5 34.9 3.3 4000 VL
S6 65 180 0.12 0.9 S20 36 5 7.2 1.5 34.7 4.0 4000 VL
S7 65 180 0.20 0.9 S20 36 4 10.8 1.5 65.0 12.8 4000 VL
T6 65 180 0.10 0.9 S20 36 5 6.4 1.5 30.8 3.2 4000 VL

110
 Peak Particle Velocity, PPV [m/s]

0.00
0.50
1.00
1.50
2.00
2.50
3.00
E0
F0
F1
G0
H0
H1
I0
I1:2
I1:25
J1:2
J1:45
J2:25
J2:2
J13
K2
K3
K4
K5
K10
K12
K13
L2:2
L2:45
L3
L4:45
L4:25
L5:45
L5:2
L6
L9:4
L9:45
L10
L11
PPV - Level 1275

L13
M2
M3
M4
M5
M6
M7:45
M7:4
M8
M9:45
M9:25
M10
M11
M12
N2
N6
N7
N9
N10
N11
O6
O10
P6
Q6
R6
R7:2
R7:1
R8
S6
S7
T6
111
 Excavation Vulnerability Potential, EVP

20.0
40.0
60.0
80.0

0.0
100.0
120.0
140.0
E0
F0
F1
G0
H0
H1
I0
I1:2
I1:25
J1:2
J1:45
J2:25
J2:2
J13
K2
K3
K4
K5
K10
K12
K13
L2:2
L2:45
L3
L4:45
L4:25
L5:45
L5:2
L6
L9:4
L9:45
L10
L11
EVP - Level 1275

L13
M2
M3
M4
M5
M6
M7:45
M7:4
M8
M9:45
M9:25
M10
M11
M12
N2
N6
N7
N9
N10
N11
O6
O10
P6
Q6
R6
R7:2
R7:1
R8
S6
S7
T6
112
 Rockburst damage potential, RDP

10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0

0.0
M3
N7
K3
M2
L3
N10
N6
M6
N9
M11
L5:45
M4
M7:4
K10
L2:45
K4
L4:45
M7:45
N2
L5:2
M10
K2
M9:25
O10
O6
L4:25
L6
M9:45
L2:2
damage zone

S7
J2:25
L9:4
N11
R7:1
RDP - Level 1275

L9:45
K5
F1
M8
R7:2
J1:45
I1:25
J2:2
K12
L11
P6
L10
J1:2
Q6
M5
S6
J13
E0
R8
R6
T6
M12
H1
I0
K13
G0
L13
H0
I1:2
F0
113
 EVP vs PPV - Level 1275
7.00

6.00

5.00
Peak Particle Velocity, PPV [m/s]

4.00

3.00

M3

2.00
N6
N10 N7
M6
L3
1.00 M2

K3

0.00
1.0 10.0 100.0 1000.0
Excavation Vulnerability Potential, EVP

114
E Seismic risk mitigation

Level 1225 mitigation

7

6
Peak Particle Velocity, PPV [m/s]

5
M9

N5
3

N8
2 K2

K5
L9
0
1 10 100 1000
Excavation Vulnerability Potential, EVP

115
 Level 1250 mitigation
7

5
Peak Particle Velocity, PPV [m/s]

0
1 10 100 1000
Excavation Vulnerability Potential, EVP

116
 Level 1275 mitigation
7

5
Peak Particle Velocity, PPV [m/s]

0
1 10 100 1000
Excavation Vulnerability Potential, EVP

117