DYNAMICS OF WIND BLASTS IN UNDERGROUND COAL MINES
by
Pradeep Sharma
B Tech, M.Tech (Mining Engineering, Indian School of Mines, Dhanbad, India)
Thesis submitted to The University of New South Wales in fulfilment of the requirements for the degree of Doctor of Philosophy
May 2001 ADDENDUM
Holt, GE, 1989, Water infusion and hydraulic fracturing to control massive roof caving, National Energy Research, Development and Demonstration Program, End of Grant Report No. EG/91/965, available Melbourne: The Australian Mineral Industry Research Association.
Kapp, WA, 1984, Mine Subsidence and Strata Control in the Newcastle District of the Northern Coalfields, NSW, Ph.D thesis, Dept, of Civil and Mining Engineering, The Univ. of Wollongong (unpublished), 1984.
McCarthy, PJ, 1990, Myuna Colliery - Significant incident, Newcom Collieries Pty Ltd (unpublished report).
Sarkar, SK & Singh, B, 1985, Longwall Mining in India, Scientific Mining Publishers, Dhanbad, 1985.
Simson, J, Hebblewhite, B & Fowler, JCW, 1997, Windblast investigations and operational experience at Newstan Colliery, Proc. 1st International Underground Coal Conference, Jointly organised by The University of New South Wales and ACRIL, Sydney, 11-13 June, 1997, (eds.) Hebblewhite, BK, Galvin, JM & Broome, AJ, pp. 133-147.
Stothard, R, 1989, Wallarah Colliery - Significant incident, United Mine Workers Federation of Australia (unpublished report).
ERRATA
Page Description Existing To read as
15 reference Salamon, 1990 Salamon & Oravecz, 1976
17 reference Sarkar, 1982 Sarkar & Singh, 1985
24,25 reference Frith, 1993 Frith, Stewart & Price, 1992
25 reference Hatherly et al., 1996 Hatherly et al., 1995
35 reference Wilson et al., 1976 Wilson,Ucciardi & Armstrong, 1976
36,37 reference Sarkar & Singh, 1987 Sarkar & Singh, 1985 ABSTRACT
Strong, massive roof strata in coal mines, especially in the Newcastle Coalfield, New South Wales, Australia tend to “hang up”, leading to extensive areas of unsupported roof. Subsequent collapse of such roof, often without warning, compresses the air beneath and displaces it into adjacent workings, leading to phenomenon known as wind blast. The risks associated include injury to mine personnel, damage to equipment, disruption to the ventilation system and possibility of methane explosion.
Five longwall panels were monitored at Moonee Colliery, mining the Great Northern seam under massive conglomerate roof. Significant wind blast events associated with the goaf falls were recorded utilising the Wind Blast Monitoring System. The impact of roof caving mechanism on wind blast events including induced caving by hydraulic fracturing has been analysed. Reanalysis of previous wind blast data for Newstan Colliery longwall panels was done for comparison with Moonee Colliery.
The fluid mechanics involved in the compression and expulsion of air during wind blasts have been defined and the overpressure and air velocity time histories utilised to determine the wind blast parameters and define the relationship between them. The characteristics of wind blasts pressure pulses in mine roadways show some similarity with transient phenomenon like air blasts from explosives and shock waves from the failure of pressurised vessels. The peak wind velocities in roadways however exhibit marked deviations from theoretical predictions based on compressible fluid flow and Rankine-Hugoniot relationships applied to goaf air displacement and panel geometry Peak overpressures and wind velocities have shown a positive correlation with the areas of goaf falls with the values peaking off for larger areas. Empirical scaling laws for wind blasts could not be properly defined due to the unpredictable variations in strata caving behaviour and their influence on the magnitude and intensity of the events. However, the monitoring has also provided some insight into the geo-technical aspects associated with caving in longwall panels.
The field investigations have helped in quantifying the values of the wind blasts parameters and assisted the mine management in risk assessments and evaluating the success of new strategies introduced to mitigate the hazard.
i Dedicated to the loving memory of my father
Late Shri Madanjit Sharma
(1930- 1998) ACKNOWLEDGEMENT
I would like to formally acknowledge the help and guidance received during the writing of this thesis and the preceding period of research from the following persons and organisations:
Dr JCW Fowler, research supervisor and mentor under whose guidance and support this work became possible. I was motivated to work under him in this challenging and unique area of research as it had immediate practical applications in the industry. Dr Fowler not only provided able guidance, but also arranged financial support without which it would not have been possible to complete the study. His guidance was also invaluable in writing research proposals and obtaining the Australian Research Council small grants, which partially funded the research and provided me with financial support.
Dr AK Bhattacharya, co-research supervisor, for providing invaluable suggestions and motivating me to be on track and finish the thesis in time
Professor JM Galvin, Head of School of Mining Engineering and Professor Bruce Hebblewhite, Director Mining Research, for awarding me a research scholarship and providing the facilities of the department. Professor Galvin also supported the research proposal and encouraged me to continue and complete my research during my initial period of financial and emotional distress.
Mr Ross Campbell, Manager, Moonee Colliery; Mr Colin Macdonald, Mining Engineer; Mr John Edwards, Geologist; Mr Steve Lamb, Survey Officer and all the staff of Moonee Colliery for their help in conducting the field investigation and providing necessary facilities.
The research work has utilised the field instrumentation and laboratory equipment systems developed earlier and funded by previous research grants. The present work is a continuation of the research program funded by the Australian Coal Association Research Program. All research agencies associated with the research project including The University of New South Wales and the associated research personnel are duly acknowledged.
iii The Regional Engineering College, Rourkela, India, provided me study leave to pursue this research and my colleagues, friends and teachers motivated me to come to Australia, for availing the better research facilities. I also acknowledge the encouragement and guidance of my professors at Indian School of Mines, Dhanbad, especially, Prof RN Gupta, Prof Surek Bordia, Prof VP Singh, Prof SB Shrivastava, Prof SP Banerjee and Prof AK Ghose.
Without the blessings of my parents it would not have been possible to undertake this task. My father, Shri Madanjit Sharma always was a role model for me. His untimely death, during my absence from India, was a big emotional setback for me. My mother, Smt Sharda Sharma and my family provided me the necessary strength to recover from this setback and remain focussed on the task ahead. My elder sister, Dr Anoop Thakur, helped me to venture overseas to pursue my studies, my younger sister, Mrs Shamita Bhagat and sister-in-law Mrs Anuradha Sharma, have constantly encouraged me.
Finally, I acknowledge the help and encouragement of my parents in law, Mr and Mrs RP Prashara, for unhesitatingly looking after my two daughters, Shikha and Neha for nearly 2 years during my stay overseas. I admire their sacrifice and also affection towards my daughters. My wife, Poonam, has also borne the initial period of separation with fortitude. I am grateful to my family for their patience and all the help rendered during the writing of this thesis.
iv LIST OF PUBLICATIONS
The following papers are based on this research work:
Sharma P & Fowler JCW 1998, ‘Longwall caving under massive conglomerate roof- an investigation into wind blast and associated safety issues’ Proc. 5 South American Congress on Rock Mechanics, ISRMRegional Symposium., Santos, Brazil, November, 1998.
Sharma P & Fowler JCW 1999, ‘Impact of roof rockmass properties on wind blast: laboratory modelling and field investigations in Australian coal mines’, Proc. 8th Australian New Zealand Conference on Geomechanics, Hobart, Tasmania, February, 1999.
Sharma P & Fowler JCW 1999, ‘Longwall mining under massive conglomerate roof: Case studies of Australian underground coal mines’, Site Characterisation Practice, Proc. International Conference on Rock Engineering for Site Characterisation, Bangalore, December, 1999. TABLE OF CONTENTS
ABSTRACT...... i
ACKNOWLEDGEMENT...... iii
LIST OF PUBLICATIONS...... v
LIST OF FIGURES...... xi
LIST OF TABLES...... xviii
LIST OF SYMBOLS...... xx
CHAPTER 1...... 1
INTRODUCTION...... 1
1.1 WIND BLAST PROBLEM IN MINES...... 1
1.2 BACKGROUND OF WIND BLAST RESEARCH IN AUSTRALIA...... 2 1.2.1 UNSW Wind Blast Project...... 2 1.3 RESEARCH OBJECTIVES...... 3
1.4 CURRENT RESEARCH SCOPE...... 3
1.5 STRUCTURE OF THESIS AND CONTENTS OF CHAPTERS...... 6
CHAPTER 2...... 8
WIND BLASTS IN MINES...... 8
2.1 INTRODUCTION...... 8 2.2 SAFETY ISSUES CONCERNING WIND BLASTS IN MINES...... 9 2.2.1 Safety of mine personnel...... 9 2.2.2 Disruption of the mine ventilation system...... 10 2.2.3 Damage to plant and equipment...... 11 2.2.4 Expulsion of methane from the goaf and possibility of an explosion...... 11 2.3 WIND BLAST INCIDENTS IN AUSTRALIA...... 13 2.4 WIND BLAST INCIDENTS OVERSEAS...... 14 2.4.1 Uni ted States of America...... 14 2.4.2 South Africa...... 15 2.4.3 China...... 16
vi 2.4.4 India...... 16 2.5 FACTORS AFFECTING WIND BLAST OCCURRENCE...... 17 2.5.1 Stress distribution around a typical longwall panel...... 18 2.5.2 Caving characteristics of a strong, massive roof in longwall workings...... 19 2.5.3 Height of caving zone...... 22 2.5.4 Impact of longwall panel width on caving behaviour...... 23 2.5.5 The magnitude of a wind blast...... 26 2.5.6 The intensity of a wind blast...... 27 2.6 WIND BLAST CASE STUDIES...... 27 2.6.1 Wallarah Colliery...... 28 2.6.2 Myuna Colliery...... 28 2.6.3 Endeavour Colliery...... 32 2.7 REDUCTION OF WIND BLAST HAZARD...... 34 2.7.1 Reducing magnitude...... 34 2.7.2 Improving caving...... 35 2.7.3 Induced caving...... 35 2.7.3.1 Hydraulic fracturing/Water infusion...... 36 2.7.3.2 Blasting...... 36 2.7.4 Reducing goaf fall height...... 37 2.1 A. 1 Restricting mining height...... 37 2.7.5 Packing the goaf...... 37 2.7.6 Reducing wind blast intensity...... 38 2.7.7 Reducing the potential for methane explosions...... 38
CHAPTER 3...... 39
AUSTRALIAN RESEARCH ON WIND BLASTS...... 39
3.1 INTRODUCTION...... 39 3.2 WIND BLAST MONITORING SYSTEM...... 39 3.3 PHYSICAL MODEL DEVELOPMENT...... 43 3.3.1 Computerised control and data acquisition...... 44 3.4 NUMERICAL MODEL DEVELOPMENT...... 47
CHAPTER 4...... 49
DYNAMICS OF WIND BLASTS IN COAL MINES...... 49
4.1 INTRODUCTION...... 49
4.2 Fluid dynamics of compressible airflow...... 50 4.2.1 Compressibility factor...... 51 4.2.2 Flow through a convergent nozzle...... 53 4.2.3 Shock waves...... 53 4.2.4 Rankine - Hugoniot relations...... 54 4.2.5 A diabatic friction flow...... 59 4.2.6 Pressure pulses due to surface blasts...... 60 4.2.7 Air pressure pulses due to underground blasts...... 65
4.3 WIND BLAST DYNAMICS IN UNDERGROUND MINE WORKINGS...... 66 4.3.1 Characteristics of wind blast overpressure and velocity time histories...... 66 4.4 ANALYTICAL MODELS OF WIND BLAST...... 69 4.4.1 Bounded model...... 70 4.4.2 A ir leakage model...... 72 4.4.3 Bursting reservoir model...... 73 4.4.3.1 Scaling laws for blast waves from bursting pressure systems...... 75 4.4.4 Pressurised reservoir model...... 76
CHAPTER 5...... 80
WIND BLAST STUDY AT MOONEE COLLIERY...... 80
5.1 INTRODUCTION...... 80 5.2 DETAILS OF MOONEE COLLIERY...... 81 5.2.1 Geology and tectonic setting...... 81 5.2.2 Insitu stress field at Moonee Colliery...... 86 5.2.3 Panel design criteria...... 89 5.2.4 Wind blast management plan...... 89 5.2.5 Mon itoring...... 90 5.2.5.1 Microseismic monitoring...... 91 5.2.5.2 Roof support leg pressure monitoring...... 92 5.3 WIND BLAST MONITORING...... 92 5.3.1 Introduction...... 92 5.3.2 Location of instrumentation...... 93 5.3.3 Installation of instrumentation...... 96 5.3.4 Wind blast monitoring procedure...... 98 5.3.5 A nalysis of field data...... 101
5.4 RESULTS OF MONITORING IN LW1 AND LW2 PANELS...... 103 5.4.1 Overpressure and wind velocity time histories...... 104 5.4.2 Analysis of overpressure and Wind velocity data...... 110 5.4.2.1 Characteristics of the overpressure time history...... 110 5.4.2.2 Characteristics of wind velocity time history...... Ill 5.4.2.3 Relationships between overpressure and wind velocity time histories...... 114 5.4.2.4 Attenuation of wind blast intensity with distance from goaf fall...... 117
viii 5.4.2.5 Relationship between wind blast intensity and overall roof fall area...... 117 5.4.3 Roof failure andprecusors to wind blast events...... 120 5.4.3.1 LW1 panel first fall...... 126 5.4.3.2 LW2 panel first fall...... 126 5.4.3.3 LW3 panel first fall...... 127 5.5 INDUCED CAVING BY HYDRAULIC FRACTURING...... 129 5.5.1 Hydraulic fracturing procedure...... 130 5.5.2 Impact of hydraulic fracturing on the rooffailure mechanism...... 133 5.5.2.1 Relative magnitude of stresses acting on the roof strata...... 133 5.5.2.2 Hydraulic fracturing and roof failure process...... 134 5.5.3 Wind blast events associated with hydraulic fracturing...... 135 5.5.3.1 LW4a panel first fall...... 139 5.5.3.2 LW4b panel first fall...... 140 5.5.4 Comparative analysis of overpressure and wind velocity data...... 140 5.5.4.1 Characteristics of the overpressure time history...... 140 5.5.4.2 Characteristics of wind velocity time history...... 141 5.5.4 3 Relationships between overpressure and wind velocity time histories...... 144 5.5.4.4 Relationship between wind blast intensity and overall roof fall area...... 147 5.5.5 Conclusions regarding the impact of hydraulic fracturing...... 149
5.6 Empirical relationship between wind blast parameter and
SCALED DISTANCE...... 153
CHAPTER 6...... 158
COMPARATIVE STUDY OF WIND BLASTS IN NEWSTAN AND MOONEE COLLIERIES...... 158 6.1 INTRODUCTION...... 158 6.2 GEOLOGICAL AND TECTONIC SETTING AT NEWSTAN COLLIERY...... 159 6.2.1 Regional and longwall panel geology...... 159 6.2.2 Stress field at Newstan...... 164 6.3 WIND BLAST EVENTS AT NEWSTAN COLLIERY...... 164 6.3.1 Caving mechanism at Newstan Colliery...... 165 6.4 REANALYSIS OF NEWSTAN WIND BLAST DATA...... 169 6.4.1 Overpressure and wind velocity time histories...... 169 6.4.2 Analysis of overpressure and wind velocity data...... 171 6.4.2.1 Characteristics of the overpressure time history...... 174 6.4.2.2 Characteristics of wind velocity time history...... 174 6.4.2.3 Relationships between overpressure and wind velocity time histories...... 174 6.4.3 A ttenuation of wind blast intensity with distance from goaffall...... 179
6.5 COMPARISON OF RESULTS...... 179
6.6 Conclusions...... i8i
ix CHAPTER 7 182
CONCLUSIONS AND RECOMMENDATIONS REGARDING WIND BLASTS...... 182
7.1 GEO-TECHNICAL ASPECTS RELATING TO WIND BLASTS...... 182
7.1.1 Caving at Moonee Colliery...... 183
7.2 FLUID-DYNAMIC ASPECTS OF WIND BLASTS...... 184 7.2.1 A coustic effect...... 184 7.2.2 Overpressure characteristics...... 185 7.2.3 Wind velocity characteristics...... 186 7.2.4 Impact of cushioning...... 187 7.2.5 Wind blast events associated with hydraulic fracturing of roof...... 187
7.3 RECOMMENDATIONS...... 188
7.3.1 Recommendations for further work...... 190 7.3.1.1 Monitoring of wind blasts...... 190 7.3.1.2 Physical modelling...... 190 7.3.1.3 Numerical modelling...... 191 7.3.1.4 Research in caving mechanism...... 191
REFERENCES...... 192
APPENDICES...... 203
Appendix A...... 203
Wind Blast Data Logger, Sample file format for overpressure and velocity
CHANNEL,INTERPRETATION AND CORRECTIONS...... 203
Appendix B...... 215
EXAMPLES OF PROCESSED OVERPRESSURE AND VELOCITY TIME HISTORIES, NEWSTAN & MOONEE
Collieries...... 215
Appendix C...... 223
WIND BLAST VELOCITY AND OVERPRESSURE PARAMETERS, MOONEE COLLIERY...... 223
Appendix D...... 234
Example of wind blast notes, Moonee Colliery...... 234
Appendix E...... 238
EXAMPLE OF HYDRAULIC FRACTURING NOTES, MOONEE COLLIERY...... 238
Appendix F...... 242
EXAMPLE OF MICRO-SEISMIC OPERATOR’S REPORT, MOONEE COLLIERY...... 242
Appendix G...... 246
WIND BLAST VELOCITY AND OVERPRESSURE PARAMETERS, NEWSTAN COLLIERY...... 246
Appendix H...... 249
Calibration and wind tunnel testing of sensor pods, Wind Blast Data Logger...... 249
x LIST OF FIGURES
Figure 1.1 The Newcastle Coalfield, New South Wales...... 5
Figure 2.1 The stress distribution around a typical longwall panel...... 21
Figure 2.2 Failure mechanism for bedded roof: shear, and snap through...... 21
Figure 2.3 Longwall and shortwall geomechanics...... 25
Figure 2.4 Distribution of microseismic events relative to longwall face position, Gordonstone
Mine...... 25
Figure 2.5 Location plan , significant wind blast, 3 Panel Left, Wallarah Colliery...... 29
Figure 2.6 Plan of site of significant wind blast, Wallarah Colliery...... 30
Figure 2.7 Plan of site of significant wind blast, Myuna Colliery...... 31
Figure 2.8 Plan of site of significant wind blast, Endeavour Colliery...... 33
Figure 3.1 The Wind Blast Data Logger...... 41
Figure 3.2 The Sensor Pods...... 41
Figure 3.3 The Hand Held Interface...... 42
Figure 3.4 The Wind Blast Physical Model...... 42
Figure 3.5 Layout of Wind Blast Physical Model...... 45
Figure 3.6 Wind velocity time history for physical model...... 45
Figure 4.1 Normal Shock in steady flow...... 55
Figure 4.2 Explosive Shock front...... 55
Figure 4.3 Rankine -Hugoniot curve for various density ratios...... 57
Figure 4.4 Shock front velocity, particle velocity and gas front velocity, as a function of
scaled distance from TNT charges in air...... 57
Figure 4.5 Moody diagram for frictional flow of air...... 61
Figure 4.6 Idealised profile of a blast wave from a condensed high explosive...... 64
Figure 4.7 Nomogram for air blast overpressure as function of scaled distance...... 64
Figure 4.8 Typical overpressure time histories for wind blasts in longwall panel...... 67 xi Figure 4.9 Wind velocity time histories for wind blasts in longwall panel 67
Figure 4.10 The Bounded model for wind blast in mines...... 71
Figure 4.11 Idealized profile for a blast wave from a pressurized vessel burst...... 71
Figure 4.12 Relationship between ambient and reservoir pressure ratio and “blowdown” time for
reservoir volume and orifice ratio 100:1...... 78
Figure 4.13 Relationship between ambient and reservoir pressure ratio and “blowdown” time for
reservoir volume and orifice ratio 1000:1...... 78
Figure 5.1 Generalised stratigraphic sequence, Moon Island Beach Sub-Group and
overlying Munnorah Conglomerate formation, Newcastle Coal measures...... 84
Figure 5.2 Section through the Moon Island Beach Sub-Group passing near to Moonee Colliery.....84
Figure 5.3 Stratigraphic sequence at Borehole JWA 30 and typical development section.
Great Northern seam, Moonee Colliery...... 87
Figure 5.4 Longwall panel no. 2 and sensor pods locations at time of initial goaf fall,
Moonee Colliery...... 94
Figure 5.5 Typical longwall panel layout and maingate A heading layout, Moonee Colliery...... 97
Figure 5.6 Effect of smoothing the wind blast velocity time history...... 102
Figure 5.7 Wind velocity time history for goaf fall no. 1 in LW3, Moonee Colliery...... 102
Figure 5.8 Overpressure time history in tailgate, goaf fall no. 6, LW2, Moonee Colliery...... 108
Figure 5.9 Overpressure time histories around face and outbye in maingate, goaf fall no. 6,
LW2, Moonee Colliery...... 108
Figure 5.10 Overpressure time history for goaf fall no. 10,maingate LW2, near to face,
Moonee Colliery...... 109
Figure 5.11 Wind velocity time history of goaf fall no. 11, tailgate LW2, near to the face,
Moonee Colliery...... 109
Figure 5.12 Relationship between impulse and peak overpressure...... 112
Figure 5.13 Relationship between max. rate of rise of overpressure and peak overpressure at Pod 2
location...... 112
xii Figure 5 .14 Relationship between the maximum rate of rise of overpressure and peak
overpressure at Pod 4 location...... 113
Figure 5.15 Relationship between maximum excursion and peak wind blast velocity at
Pod 4 location...... 113
Figure 5.16 Relationship between maximum excursion and peak wind blast velocity
at Pod 2 location...... 115
Figure 5.17 Relationship between maximum rate of rise of wind blast velocity with peak wind blast
velocity at Pod 3 location...... 115
Figure 5.18 Relationship between peak wind blast velocity and peak overpressure
at Pod 4 location...... 116
Figure 5.19 Relationship between peak wind blast velocity and peak overpressure
at Pod 2 location...... 116
Figure 5.20 Attenuation of peak overpressure with distance along the maingates
of longwall panels...... 119
Figure 5.21 Relationship between peak overpressure and roof fall area at Pod 2 location...... 119
Figure 5.22 Relationship between impulse and goaf fall area at Pod 2 location...... 121
Figure 5.23 Relationship between peak wind velocity and goaf fall area at Pod 2 location...... 121
Figure 5.24 Support leg pressures recorded for fall no. 6 at LW2, Moonee Colliery...... 122
Figure 5.25 Support leg pressures recorded for fall no. 6 at LW2 (reverse), Moonee Colliery...... 123
Figure 5.26 Goaf falls in longwall panels, Moonee Colliery...... 128
Figure 5.27 Hydraulic fracturing sequence in LW3 panel, Moonee Colliery...... 131
Figure 5.28 Hydraulic fracturing sequence in LW4 panel, Moonee Colliery...... 132
Figure 5.29 Relationship between impulse and peak overpressure for all falls classes,
at Pod 2 location Moonee Colliery...... 142
Figure 5.30 Relationship between impulse and peak overpressure for all falls classes,
at Pod 4 location Moonee Colliery...... 142
Xlll Figure 5.31 Relationship between the maximum rate of rise of overpressure and peak overpressure
at Pod 2 location...... 143
Figure 5.32 Relationship between the maximum rate of rise of overpressure and peak overpressure
at Pod 4 location for all fall classes, Moonee colliery...... 143
Figure 5.33 Relationship between maximum excursion and peak wind blast velocity at
Pod 4 location for all fall classes, Moonee Colliery...... 145
Figure 5.34 Relationship between maximum excursion and peak wind blast velocity
at Pod 2 location for all fall classes, Moonee Colliery...... 145
Figure 5.35 Relationship between maximum rate of rise of wind blast velocity with
peak wind blast velocity at Pod 2 location for all fall classes, Moonee Colliery...... 146
Figure 5.36 Relationship between peak wind blast velocity and peak overpressure at
Pod 2 location for all fall classes, Moonee Colliery...... 146
Figure 5.37 Relationship between impulse and goaf fall area for all fall classes, Moonee Colliery ... 148
Figure 5.38 Relationship between maximum excursion and goaf fall area for all fall classes,
Moonee Colliery...... 148
Figure 5.39 Relationship between peak overpressure and roof fall area at Pod 2 location,
Moonee Colliery...... 150
Figure 5.40 Relationship between peak overpressure and goaf fall area for all fall classes with area less
than 10,000 square metres, Moonee Colliery...... 150
Figure 5.41 Relationship between peak velocity and goaf fall area for all falls classes with
area less than 10,000 square metres. Moonee Colliery...... 154
Figure 5.42 Relationship between peak overpressure and scaled distance at Pod 4 location, Moonee
Colliery...... 154
Figure 5.43 Relationship between peak overpressure and scaled distance at Pod 2 location, Moonee
Colliery...... 156
Figure 5.44 Attenuation of peak overpressure with distance along the maingates of
longwall panels, Moonee Colliery...... 156
xiv Figure 6.1 Newstan Colliery longwall panels 160
Figure 6.2 Generalised stratigraphic sequence, Lambton Sub-Group and overlying
Kotara Formation, Newcastle Coal Measures...... 161
Figure 6.3 Section through Young Wallsend and The West Borehole coal, passing
near to Newstan Colliery...... 161
Figure 6.4 Longitudinal section through the roof strata, maingate 7, Newstan Colliery...... 163
Figure 6.5 Location of the wind blast events in the longwall panel in relation to the
height of channel above seam, Newstan Colliery...... 167
Figure 6.6 Location of the wind blast events in relation to the subsidence profile
and channel thickness, Newstan Colliery...... 170
Figure 6.7 Location plan of wind blast sensor, Newstan Colliery...... 170
Figure 6.8 Overpressure time history of a significant wind blast, panel 8, Newstan Colliery...... 172
Figure 6.9 Wind velocity time history of a significant wind blast, panel 8, Newstan Colliery...... 172
Figure 6.10 Overpressure time history of a less significant wind blast, panel 7, Newstan Colliery ... 173
Figure 6.11 Wind velocity time history of a less significant wind blast, panel 7, Newstan Colliery . . . 173
Figure 6.12 Relationship between impulse and peak overpressure at Pod 1 location,
Newstan Colliery...... 175
Figure 6.13 Relationship between impulse and peak overpressure at Pod 2 location,
Newstan Colliery...... 175
Figure 6.14 Relationship between maximum rate of rise of wind blast velocity
and peak wind blast velocity at Pod 1 location, Newstan Colliery...... 176
Figure 6.15 Relationship between the maximum rate of rise of overpressure
and peak overpressure at Pod 1 location, Newstan Colliery...... 176
Figure 6.16 Relationship between maximum excursion and peak wind blast
velocity at Podl location, Newstan Colliery...... 177
Figure 6.17 Relationship between maximum excursion and peak wind blast
velocity at Pod2 location, Newstan Colliery...... 177
XV Figure 6.18 Relationship between peak wind blast velocity and peak overpressure,
at Pod 1 location, Newstan Colliery...... 178
Figure 6.19 Relationship between peak wind blast velocity and peak overpressure,
at Pod 2 location, Newstan Colliery...... 178
Figure 6.20 Attenuation of peak overpressure with distance along the maingates
of longwall panels, Newstan Colliery...... 180
Figure 6.21 Attenuation of peak velocity with distance along the maingates of
longwall panels, Newstan Colliery...... 180
Figure B. 1 Raw overpressure time histories of wind blast event at LW8, Newstan Colliery...... 216
Figure B.2 Smoothed overpressure time histories of wind blast event at LW8, Newstan Colliery.....216
Figure B.3 Integrated overpressure time history of wind blast event at LW8, Newstan Colliery...... 217
Figure B.4 Air density correction factor time history of wind blast event at LW8, Newstan Colliery.217
Figure B.5 Differentiated overpressure time history of wind blast event at LW8, Newstan Colliery.,218
Figure B.6 Smoothed wind velocity time history of wind blast event at LW8, Newstan Colliery.....218
Figure B.7 Overpressure time history, near face, for wind blast event at LW4b, Moonee Colliery__219
Figure B.8 Wind velocity time history, outbye in maingate (A hdg), for wind blast event
at LW4b, Moonee Colliery...... 219
Figure B.9 Overpressure time history, in tailgate, near face, for wind blast event at LW4a,
Moonee Colliery...... 220
Figure B. 10 Wind velocity time history, in tailgate, near face, for wind blast event at LW4a,
Moonee Colliery...... 220
Figure B. 11 Overpressure time history, in maingate, near face, for wind blast event at LW3,
Moonee Colliery...... 221
Figure B. 12 Wind velcoity time history, in maingate, near face, for wind blast event at LW3,
Moonee Colliery...... 221
Figure B. 13 Overpressure time history, in tailgate, near face, for wind blast event at LW3,
Moonee Colliery...... 222
xvi Figure B. 14 Wind velocity time history, in tailgate, near face, for wind blast event at LW3,
Moonee Colliery...... 222
Figure E. 1 First underground hydraulic fracturing, LW3, Moonee Colliery...... 241
Figure E.2 First vertical hole hydraulic fracturing, LW3, Moonee Colliery...... 241
Figure F.l Prediction of roof fall, LW2, microseismic monitoring at Moonee Colliery...... 245
Figure H. 1 The University of Queensland wind tunnel results...... 250
Figure H.2 Wind tunnel testing of the wind blast sensor pod...... 251
xvii LIST OF TABLES
Table 2.1 Damage caused by various amplitudes of shock waves...... 12
Table 2.2 Wind blasts in Datong Coal Mines, China...... 17
Table 4.1 Dynamic pressure compressibility factor at various air speeds...... 52
Table 4.2 Sachs-scaled nondimensional, blast parameters...... 63
Table 4.3 Initial pressure pulses produced by explosive charges...... 65
Table 4.4 Physical parameters and their dimensions for vessel blast scaling law...... 76
Table 5.1 Moonee Colliery longwall operational details...... 82
Table 5.2 Physical damage caused by wind blast, Moonee Colliery,
longwall panel no. 1, first goaf fall...... 83
Table 5.3 Summary of geotechnical test results, CWA 42 and 43, Moonee Colliery...... 88
Table 5.4 Wind Blast Monitoring System trigger levels at Moonee Colliery...... 98
Table 5.5 Wind blast parameters, LW1 and LW2 panels, Moonee Colliery...... 104
Table 5.6 The Beaufort scale of wind speeds...... 105
Table 5.7 Monitoring of goaf falls at longwall panel no. 1, Moonee Colliery...... 124
Table 5.8 Monitoring of goaf falls at longwall panel no. 2, Moonee Colliery...... 125
Table 5.9 Physical damage caused by wind blast, Moonee Colliery,
LW2 panel, first goaf fall...... 127
Table 5.10 Physical damage caused by wind blast, Moonee Colliery,
LW3 panel, first goaf fall...... 129
Table 5.11 Monitoring of goaf falls at longwall panel no. 3, Moonee Colliery...... 136
Table 5.12 Monitoring of goaf falls at longwall panel no. 4a, Moonee Colliery...... 137
Table 5.13 Monitoring of goaf falls at longwall Panel no. 4b, Moonee Colliery...... 138
Table 5.14 Comparison of wind blast parameters with fall Class, Moonee Colliery...... 151
Table 6.1 Measurement of insitu stress, Lake Macquaire District...... 165
xviii Table 6.2 Significant wind blast events at Newstan...... 166
Table 6.3 Wind blast parameters at Newstan and Moonee Collieries...... 181
Table A. 1 Zero correction equivalent calculations...... 213
Table C. 1 Incidences of roof falls and wind blast velocities, LW1, Moonee Colliery...... 223
Table C.2 Incidences of roof falls and wind blast overpressures, LW1, Moonee Colliery...... 224
Table C.3 Incidences of roof falls and wind blast velocities, LW2, Moonee Colliery...... 225
Table C.4 Incidences of roof falls and wind blast overpressures, LW2, Moonee Colliery...... 226
Table C.5 Incidences of roof falls and wind blast velocities, LW3, Moonee Colliery...... 227
Table C.6 Incidences of roof falls and wind blast overpressures, LW3, Moonee Colliery...... 228
Table C.7 Incidences of roof falls and wind blast velocities, LW4a, Moonee Colliery...... 229
Table C.8 Incidences of roof falls and wind blast overpressures, LW4a, Moonee Colliery...... 230
Table C.9 Incidences of roof falls and wind blast velocities, LW4b, Moonee Colliery...... 231
Table C. 10 Incidences of roof falls and wind blast overpressures, LW4b,Moonee Colliery...... 232
Table C. 11 Schedule of wind blast pod locations LWl-4b, Moonee Colliery...... 233
Table G. 1 Incidences of roof falls and wind blast velocities, LW7-9, Newstan Colliery...... 247
Table G.2 Incidences of roof falls and wind blast overpressures, LW7-9, Newstan Colliery...... 248
XIX LIST OF SYMBOLS
Symbol Parameter
A Area Av Area, Vent C,CX Velocity, sound D Diameter e Energy per unit mass E Energy f Friction factor
g Acceleration due to gravity h Enthalpy, height H Height, Thickness Impulse m Mass Mx Mach Number, upstream of shock My Mach Number downstream of shock
P Pressure, overpressure P Scaled pressure Po Ambient pressure Pi, P2, P„ Pf Pressure, initial, final
Po Stagnation pressure, Px Pressure, upstream of shock
Py Pressure, downstream of shock P Peak overpressure R Distance R Scaled Distance R Specific gas constant R Resistance
Re Reynolds number t Time T Temperature
T1 O Temperature, ambient
xx Symbol Parameter
u Internal energy
Us Velocity, peak particle
U Velocity, shock
V Volume
V Velocity vP Velocity, Particle vx Velocity, upstream of shock
Vy Velocity, downstream of shock w Weight
z Distance
Z Compressibility factor
Greek Symbols
y Specific heat ratio p Coefficient of viscosity
V Kinematic viscosity
p Density
Po Density, ambient
Px Density, upstream of shock
Py Density, downstream of shock
0 Temperature
CTl Major principal stress
ct2 Intermediate principal stress
a3 Minor principal stress
CTh Minor horizontal principal stress
Major horizontal principal stress
av Vertical stress
XXI CHAPTER 1
INTRODUCTION
1.1 WIND BLAST PROBLEM IN MINES
Coal mining under strong and massive roof strata can lead to extensive areas of roof “hang up” i.e. the roof fails to cave in. These areas eventually tend to collapse, suddenly and often without warning, rapidly compressing the air beneath and forcing it out of the goaf through the adjacent mine workings, which phenomenon is termed wind blast. The air velocities and overpressure pulses associated with a significant wind blast event are of sufficient magnitude so as to endanger the mine personnel, cause disruption of the ventilation system and damage to plant and equipment. Wind blasts have been known to have led to explosions in underground coal mines of the accumulated methane expelled from the goaf
Wind blasts have been a serious problem in some collieries in the Newcastle Coalfield in New South Wales (NSW), Australia, especially those mining the West Borehole seam under massive channel conglomerates to the north west of Lake Macquaire and those collieries south of Lake Macquarie which mine the Great Northern seam under the Teralba Conglomerate. The incidences of wind blast is likely to increase as a result of trend away from pillar extraction to longwall mining due to the larger areas of roof exposed. Longwall panels, where face length is restricted to less than “critical width” because of strata control, structural geology or subsidence considerations, are prone to wind blasts as the roof often tend not to cave regularly i.e. “hang up”.
Wind blast phenomena also occurs in underground metalliferrous mines. The recent (November, 1999) disaster, resulting in four fatalities at Northparkes Mine in NSW, which employs block caving to mine copper and gold ore, illustrates the gravity of the problem.
1 1.2 BACKGROUND OF WIND BLAST RESEARCH IN AUSTRALIA
The United Mineworkers’ Federation of Australia requested support from the Joint Coal Board, in Feb 1990, to initiate action to control and if possible, to eliminate the dangers associated with wind blasts (Joint Coal Board, 1990). After communications with the New South Wales Chief Inspector of Mines, the Board directed that the phenomenon be considered independently.
Subsequently, in early 1991, the Chief Inspector, under an initiative from the New South Wales Minister for Minerals and Energy, formed a committee to investigate the phenomenon. An appraisal was conducted of the areas of prediction, prevention and protection (Wind Blast Committee, 1991). At present, wind blast that results in injuries requiring first-aid treatment, is a reportable incident, as defined by the Coal Mines (Underground) Regulation 1999 under Clause 34, Part 2, Division 5.
1.2.1 UNSW wind blast project
Research on wind blast commenced in School of Mining Engineering, University of New South Wales in early 1991. The impetus to undertake research came after two incidents at Wallarah Colliery in 1989 and in Myuna Colliery in 1990 respectively, both underground coal mines in the Lake Macquaire district of New South Wales.
The initial phase of the wind blast project was funded by a grant under the National Energy Research, Development and Demonstration (NERDD) Programme and with financial contributions from three participating companies, namely Newcom Collieries Pty Ltd, Elcom Collieries Pty Ltd , and Coal Allied Operations Pty Ltd. The second phase was funded by a grant from the Joint Coal Board Health and Safety Trust. The research findings (Fowler, 1997; Fowler & Torabi, 1997) include details of Australian case histories, description of the field instrumentation and monitoring at Wallarah, Cooranbong Collieries during 1992 and Newstan Colliery during 1995/97 and physical modelling results. Also included were the preliminary results from the above collieries.
2 The present phase of research is funded by the Australian Coal Association Research Program and some of the research findings are included in the final end of grant report (Fowler & Sharma, 2000).
1.3 RESEARCH OBJECTIVES
The aim of this research was to develop a fundamental understanding of the wind blast phenomenon resulting from massive roof failure in underground coal mines and thus provide a basis on which to develop strategies to mitigate the hazard.
The research objectives were primarily dictated by the needs of the mine management and the research directed accordingly. The primary objective was to define the fluid mechanics involved in the compression and expulsion of air during wind blasts in longwall panels in underground coal mines.
This objective included the following.
1. Real time monitoring of overpressure and wind velocity time histories for wind blast events in longwall panels, to generate a database of events and classify them;
2. Study of the influence of longwall panel layout design and panel orientation on wind blast events;
3. Study of the goaf strata caving mechanism and its impact on generation of wind blast events and risk assessment.
The secondary objectives were to provide the industry with improved, practical design tools with which to optimise panel design and layout in order to minimise the hazards associated with wind blasts
1.4 CURRENT RESEARCH SCOPE
Wind blast phenomena occurs in both underground coal mines and metalliferrous mines. However the terms of present study, in this thesis, were confined to longwall panels in underground coal mines, in particular to the study of wind blasts at Moonee Colliery.
3 Moonee Colliery, owned and operated by Coal Operations Australia Limited, mines the Great Northern seam by the longwall retreat system of mining. The Colliery is located on the New South Wales coast at Catherine Hill Bay, approximately 20 kilometres south of Newcastle and 100 km north of Sydney, within the Lake Macquaire district of the Newcastle Coalfield as shown in Figure 1.1. Many significant wind blast have occurred at Moonee Colliery since January 1998, during the mining of the first five longwall panels.
The mine introduced hydraulic fracturing of the roof in Longwall no. 3 panel for “caving on demand”. However, the wind blast hazard still remains, the strategy of the mine management being to control rather than eliminate the hazard as due to prior development, it was committed to the longwall layout. The author has been associated with wind blast research at Moonee Colliery since the problem appeared and the main findings are incorporated in this thesis. Moonee Colliery continues to experience significant wind blasts to date.
Some of the most severe wind blasts in underground coal mines have been associated with violent pillar failures for example those at Muswellbrook No. 2 Colliery in the Hunter Coalfield of NSW and at Coalbrook North Colliery, South Africa. The research issues for obviating these types of wind blasts are related to pillar design and thus lie outside the terms of the present study which was confined to those wind blasts associated with the failure of the roof element alone.
Studying the caving behaviour of massive roof was also not the main objective of the present study. However, the geomechanics aspects of roof caving associated with wind blasts have been dealt with as appropriate, without involving any numerical modelling or geo-technical instrumentation as these were beyond the scope of the study.
Physical modelling of the wind blast phenomena is restricted to that of thin roof element due to technical constraints in modelling a thick roof element while the wind blast phenomena studied in field relates to failure of a thick roof element. Moreover, for the simulation in the scaled down physical model to be directly relevant to the
4 Figure 1.1 The Newcastle Coalfield, New South Wales
(adapted from Armstrong & Mische, 1998)
5 mine, the Reynolds number of the flow should be similar to that in the field. However, the physical modelling has given some insight to the phenomena.
1.5 STRUCTURE OF THESIS AND CONTENTS OF CHAPTERS
Literature review has revealed that wind blast phenomena is not confined to underground coal mines and incidents have occurred at metalliferous mines too. However, the literature review in chapter two, emphasises on underground coal mines. The safety issues concerning wind blast in coal mines are also reviewed in the second chapter. It also deals extensively with the factors affecting wind blasts and in particular to those related to wind blasts in longwall panels of coal mines including the geo-technical aspects of wind blasts.
The wind blast project started at the School of Mining Engineering, The University of New South Wales, long before to the author’s involvement. Substantial work had been done in the past including the development of the wind blast monitoring system (Fowler, 1997). The present work has utilised the field instrumentation developed earlier, after the necessary refurbishment and modifications. The third chapter reviews the past research work at UNSW and gives the description of the various research tools developed to study the wind blast phenomenon and currently used by the author and others.
Real time monitoring of wind blasts have lead to the recording of the overpressure and wind velocity time histories associated with each significant event at both inbye and outbye locations with respect to the longwall face position. The transient phenomenon being studied involves compressible fluid flow and also ‘weak’ shock pulses generated as a result of the compression of the air under the falling roof element. The fourth chapter reviews the dynamics of wind blasts in underground coal mines and the various analytical models. Similarities between the time histories of wind blast overpressure and strong shock pulses or air blasts due to explosives or pressurised vessel failures are also discussed.
6 The field investigation at Moonee Colliery forms the substantial work done during the current phase of research. It involved the monitoring of wind blasts in the five longwall panels since January 1998. Moonee Colliery has successfully implemented hydraulic fracturing of roof to induce caving “on demand”. The roof caving behaviour is consequently modified and its impact on wind blasts has been studied. Analysis of the substantial data collected at Moonee and the relationships between the various wind blast parameters such as peak overpressure, impulse, peak wind velocity, goaf fall area etc are defined in chapter five.
Prior to Moonee Colliery, significant wind blasts had been monitored at longwall panels only in Newstan Colliery during 1995 and 1997. Out of the 23 events recorded in the 4 panels monitored, only eight were classified as significant wind blasts, i.e. of sufficient intensity to pose risk of personal injury or of damage to the mine ventilation system (Fowler & Torabi, 1997). Further detailed analysis of Newstan data set has been done using the methodology and techniques that have been applied to the Moonee data set. The wind blasts at Newstan were localised to section of the panels and the values of the wind blast parameters much lower compared to Moonee. A comparative study of wind blasts at Newstan and Moonee Collieries reveals the impact of differences in geology as discussed in chapter six.
Finally, chapter seven presents the main contributions of the research on the dynamics of wind blasts in underground coal mines towards understanding both the fluid dynamics aspects of wind blasts and the related geo-technical issues. These have also helped the Moonee Colliery management in successfully managing the hazard of wind blast.
7 CHAPTER 2
WIND BLASTS IN MINES
2.1 INTRODUCTION
Wind blasts occur in both underground coal and metalliferous mines. The recent, November 1999, disaster in Northparkes mine, NSW, Australia, an underground copper and gold mine exemplified the danger of wind blasts in mines (Brown, 1999). However in this thesis, wind blasts in underground coal mines and especially in longwall panels are addressed.
Longwall mining under massive conglomerate roof in the Newcastle Coalfield of New South Wales, Australia, has resulted in wind blasts of sufficient intensity to raise serious concerns regarding the safety of mine personnel, disruption to the ventilation system and potential expulsion of methane from the goaf. Coal mines in the Newcastle Coalfield which have experienced significant wind blasts include Wallarah, Myuna, and Cooranbong Collieries (Fowler, 1997), Endeavour Colliery and, more recently, Newstan and Moonee Collieries. At Endeavour Colliery, a significant wind blast associated with a major goaf fall preceded an explosion (Anderson, 1997). Figure 1.1 gives the location of these collieries.
The high incidence of wind blasts in the Newcastle Coalfield of New South Wales is due to the particular geology of the coalfield, a dominant feature of which is the presence of massive, strong conglomerate beds whose basal sections often lie in close proximity to the coal seams. The conglomerates, in common with the other clastic sediments, show considerable variation in both their lateral and vertical extents throughout the coalfield. They have been proven to extend over areas in excess of 200 square kilometres in irregular lenticular sheets, several discontinuous lenses often occurring on one stratigraphic horizon (Booking, Howes & Weber, 1988).
This chapter reviews the safety issues related to wind blasts in mines, the incidence of past wind blasts in Australia and overseas as reported in the literature and includes
8 several case histories. The goaf caving characteristics under massive roofs in longwall mines is also addressed. Possible strategies to ameliorate the effects of wind blast are also reviewed.
2.2 SAFETY ISSUES CONCERNING WIND BLASTS IN MINES
2.2.1 Safety of mine personnel
High wind velocities can result in mine personnel being blown over and injured or hit by air entrained objects. Measured or inferred velocities during a wind blast have substantially exceeded the value of 20 m/s which has been tentatively adopted as the threshold for the toppling of mine personnel in an upright stance (Fowler, 1997). Direct blast injury may result from air overpressure and it has been suggested (Fowler, 1997) that there are three parameters which are relevant in this context: peak overpressure, pressure rise time and impulse. The human body can withstand blast overpressures levels of 35 kPa or more before it is likely to cause organ damage. However a person may sustain injuries by being blown over or struck by flying debris at such overpressure levels. The eardrum is the organ in human body most sensitive to damage by rapid pressure changes. Acoustic trauma is a condition of sudden aural damage resulting from short term exposure or even from one single exposure and impulsive overpressure rise due to explosions is often responsible for this condition. The threshold overpressure value for eardrum rupture is of the order of 35 kPa and for lung damage 100 kPa (Gierke, 1966). This is based upon blast data and is specific to overpressure time histories associated with the arrival of shock wave following an explosion. The threshold may be a bit conservative if applied to wind blasts as they exhibit a much slower rise time.
Wind blasts resulting in injury or death have been reported from United States of America, South Africa, China and India as detailed in section 2.4 of this chapter.
Some of the worst accidents due to wind blasts, involving multiple fatalities, have been reported from coal mines of the Datong Coal Mining Administration in China. In Wa Jing Wan Mine, 14 persons were killed and 19 injured by a wind blast in October
9 1961, caused by the “instantaneous collapse” of over 163 000 square metres of hanging roof (Song & Xu, 1992).
2.2.2 Disruption of the mine ventilation system
Wind blasts cause disruption of the mine ventilation system due to damage or destruction of the mine stoppings, overcasts and other ventilation structures. Disruptions in ventilation circuits may result in dangerous methane accumulations and also in prolonged production delays.
The damage is due to the plant and equipment being subjected to dynamic loading from the overpressures associated with wind blast The overpressures can be augmented by reflection. For normal reflection of shock waves, the reflection coefficient, defined as ratio of the (total) reflection overpressure, to the overpressure of the incident shock is related to the Mach number (dimensionless index giving ratio of velocity to local speed of sound) of the incident shock (Kinney, 1962). For very strong shocks the reflection coefficient approaches the limiting value of 8.0 in air. If real gas effects such as dissociation and ionisation of air molecules are accounted for the ratio could be much higher (Baker, 1973). For weak shocks, where the Mach number approaches unity, the reflection coefficient approaches the value 2.0. The reflection mechanism for strong explosive shocks is complex and includes normal reflection, oblique reflection and Mach stem formation, a spurt-type effect which occurs when a shock front impinges on a surface at low incidence and which takes the form of a shock approximately perpendicular to the surface (Kinney, 1962).
The characteristics of wind blast pressure pulse are discussed in Chapter 4, section 4.2. If, however, for the purpose of calculation, the wind blast pressure pulse is taken to have the same characteristics as a weak shock wave, the instantaneous value of the reflected overpressure pulse is calculated to be equal to twice the peak overpressure. During one of the wind blast events at Moonee Colliery, (described in chapter 5, section 5.2) the maximum recorded peak overpressure was 34 kPa (Fowler & Sharma, 2000) which gives a value of 68 kPa in case of a reflected overpressure pulse. This is.equivalent to an instantaneous dynamic force of 1122 kN on a stopping in a roadway 10 of width 5.5 metres and height 3.0 metres. This is more than enough to destroy any stopping other than designed to be an explosion proof type which are capable of withstanding pressure loadings of 150-350 kPa.
2.2.3 Damage to plant and equipment
Most mining equipment is sufficiently robust not to be directly damaged by a wind blast. However, damage has been reported to ventilation stoppings and doors, overcasts, water barriers and occasionally, main fans (Joint Coal Board, 1990; Caffery, 1990). As mentioned in previous section 2.1.2, the damage is due to the plant and equipment being subjected to dynamic loading from the overpressures associated with the wind blast. Mumford and Lihou (1979) have shown the likely damage from various peak overpressures caused by shock waves resulting from TNT explosions from data compiled by Peter (1968) and presented in Table 2.1. However, the dynamic load imposed on a structure by strong blast wave is complex. The structure may be subjected to instantaneous dynamic load imposed on the front face by the reflected overpressure, by stagnation overpressure (combined effect of shock overpressure and blast wind) after the reflected shock dies out and finally by the drag pressures on the back of the structure at various times during the blast wave decay process. It may also be subjected to instantaneous side and top loads by the peak overpressure (Kinney, 1963).
2.2.4 Expulsion of methane from the goaf and possibility of an explosion
Wind blast may increase the hazard of explosion by expelling the goaf air which may contain methane and mixing it with raised coal dust. This methane /air mixture from the goaf can reverse the ventilation flow on the intake side of a longwall panel and may penetrate beyond the “hazardous zone” defined by statutes into areas where non- intrinsically safe and non-flameproof equipment may be located. It is also of concern that such equipment may be in the process of being shut down by safety devices designed to trip the power supply in the event of a wind blast and hence, be at its most hazardous (as a potential ignition source) at the very time while being inundated by displaced air/methane (Fowler, 1997).
11 McPherson (1995a & 1995b) concluded that adiabatic temperature increase during compression of air during large roof falls, was a potential source of ignition of coal dust/methane explosion following a wind blast. At Moura no. 4 Mine in Queensland, in 1986, twelve miners were killed in an explosion which is considered to have been preceded by a wind blast (SIMTARS, 1990). An explosion in Endeavour Colliery, NSW on 28 June, 1995 also followed a large wind blast in the 300 pillar extraction panel (Anderson, 1997).
Table 2.1
Damage caused by various amplitudes of shock waves (adapted from Mumford and Lihou, 1979; after Peter, 1968)
Pressure Damage (kPa) 0.2 Occasional breaking of large glass windows already under strain. 0.7 Breakage of windows, small, under strain. 1.0 Typical pressure for glass failure 2.0 “Safe Distance” (probability 0.05 no serious damage beyond this point) 2.0 Some damage to house ceilings, 10 % window glass broken 2.8 Limited minor structural damage. 3.5-6.9 Large and small windows usually shattered; occasional damage to window frames. 4.8 Minor damage to house structure. 5.2 Breakage of small windows , not under strain. 6.9 Partial demolition of houses, made uninhabitable. 9.9-13.8 Corrugated asbestos shattered. Corrugated steel or aluminium panels, fastenings fail. 9.0 Steel frames of clad building slightly distorted. 13.8 Partial collapse of walls and roof houses. 18.8-20.7 Concrete or cinder block walls, not reinforced, shattered. 15.9 Lower limit of serious structural damage. 17.3 50 % destruction of brickwork of house. 20.7 Heavy machines (1350 kg) in industrial building suffered little damage. 20.7 Frameless, self framing, steel panel building demolished. Steel frame building distorted. 20.7-27.6 Rupture of oil storage tanks. 27.6 Cladding of light industrial building ruptured. 34.5 Wooden utilities poles snapped. Hydraulic press (18000 kg) slightly damaged. 34.5-48.3 Nearly complete destruction of house. 48.3 Loaded train wagons overturned. 48.3-55.2 Brick panels 0.2 -0.3 metres not reinforced, failed by shearing or flexture. 62.1 Loaded train/box-cars completely demolished. 69.0 Probable total destruction buildings. Heavy (3200 kg) machine tools moved and badly damaged. Very heavy (5500 kg) machine tools survived.
12 2.3 WIND BLAST INCIDENTS IN AUSTRALIA
Perhaps the earliest recorded incidence of wind blasts in Australian mines relates to the nine fatalities which occurred in the South Mine at Broken Hill, New South Wales, on 18 July, 1895. Just before the roof of a stope caved in, the underground manager had arrived and after assessing the perilous state of the roof, ordered twenty men to leave. When the roof fell, it caused a wind blast through a lower level and 9 men who were sitting there, were battered to death against the top and sides of the mine (Carrol, 1977).
The most recent fatal incident in Australia, resulting in 4 deaths, was that at Northparkes mine, New South Wales on 24th November 1999 (Brown, 1999). The mine utilises the block caving method, to mine the massive low grade copper and gold deposit. Sudden collapse of the stope roof caused the wind blast “with a force greater than the worst hurricane, killing the four men before they could blink.” Two of the dead were in a vehicle when caught by the blast in the main decline shaft and tossed about, hitting the wall. The air blast shredded the steel 4WD vehicle, strewing its wreckage a kilometre along the decline
Wind blasts resulting in injury or death have also been reported in both pillar extraction and longwall panels in both New South Wales and Queensland underground coal mines.
The Joint Coal Board (JCB), in its 1990 report on the wind blast phenomenon stated that “ a majority of colliery managers acknowledge that wind blast is a high incidence phenomenon, but as it has a low accident frequency, there is very little recording of the wind blasts when they occur”.
For the Newcastle Coalfield, the JCB report cites the following instances where mine personnel were blown over by wind blasts during pillars extraction operations in a 15 month period.
18 November, 1988 Awaba
13 23 March, 1989 Nattai South
27 November, 1989 Wallarah
15 February, 1990 Myuna
(date not mentioned) Wallsend Borehole
Further case histories involving wind blasts in are presented in section 2.4 and 2.6 of this chapter.
First falls in longwall panels have often been the source of wind blasts of significant intensity, as experienced in the first longwall panel at Southern Colliery, Queensland which resulted in damage to several ventilation stoppings and lifted the explosion lid off the main fan (Caffery, 1990). However, once regular caving is established, subsequent falls do not generally give rise to such events. Moonee and Newstan Colliery longwall panels have proved to be an exception and wind blast events have occurred throughout all or most of the length of several contiguous panels. Case studies at Moonee and Newstan Collieries are described in detail in chapters 5 and 6 respectively.
2.4 WIND BLAST INCIDENTS OVERSEAS
Mining under massive roof conditions has resulted in wind blasts in the United States of America, South Africa, China and India.
2.4.1 United States of America
A goaf fall at Kopperston No. 4 mine in West Virginia in September, 1989 caused a “hurricane force” wind that killed one miner and injured several others by blowing them against mine machinery (Anon, 1989). The Wind Blast Committee (1991) reported that a visit to Elk Run Coal Mine in West Virginia had revealed that a strong roof in a pillar extraction panel had hung up, its eventual failure resulting in a massive wind blast as a consequence of which pillar extraction in the seam involved had been discontinued.
Production economics have led to wider longwall panels being adopted and with multiple entries which coincidentally provide good dissipation for potential wind blasts. 14 Consequently, the literature review did not reveal any wind blast problem in longwall workings under massive strata. Coal bumps have been the major cause of concern for longwall workings under strong massive roof strata (Wu & Karfakis, 1993).
2.4.2 South Africa
The highest recorded loss of life due to the collapse of mine working occurred at Coalbrook North Colliery in January, 1960 (McPherson, 1980) when Bord and Pillar workings collapsed over an area exceeding one square mile. The resulting wind blast reached the surface, blowing out a wall in the fan house.
In their book “Rock Mechanics in Coal Mining”, Salamon and Oravecz (1976) have discussed the South African longwall mining experience under a massive dolerite sill at Hlobane, Durban Navigation and Sigma Collieries. From longwall mining experience at Durban Navigation Collieries, it was established that:
1. The extraction of the seam resulted in the caving of the roof strata only between the seam and the base of the sill; 2. The dolerite sill bridged over the extracted area without failure and that it underwent only elastic deflection; 3. As a result of the arrested caving and subsequent reconsolidation of the caved strata under its own weight an extensive cavity was formed at or near the base of the sill; 4. The maximum surface subsidence observed reflected only the elastic deflection of the sill, and that it was a small proportion of the extracted height (less than ten percent); 5. After completion of mining in the panel there was little further change in the displacement picture, indicating that dolerite does not creep appreciably.
When the failure of the competent dolerite stratum takes place in the presence of some bed separation at its base, the resulting collapse could endanger the safety of the mining operation (Salamon, 1990). If the gap arises when height of caving (Hc) is less than height of competent strata above the seam (Ho), such delayed subsidence of the upper beds may represent a serious hazard. However if He is significantly greater than Ho, then the hazards are likely to be less acute since the rock pile and the beds lying on it 15 cushion the most severe effect of a sudden failure. The occurrence of incomplete caving and the associated gap formation was considered a potentially dangerous situation. A sudden failure of the dolerite bed could cause, apart from other dynamic effects, an inrush of inflammable gases leading to a hazardous situation.
Sigma Colliery demonstrated that it was safe to conduct longwall mining with caving under the massive dolerite sill provided it is restricted to areas where the normal shale/sandstone parting between the dolerite and the seam is not less than eight to ten times the height of the seam. Consequently, longwall mining is not adopted, where geo mining conditions as described above, could lead to wind blasts in South Africa.
2.4.3 China
As mentioned in section 2.1.1, some of the worst accidents due to wind blasts, involving fatalities, have been reported from coal mines of the Datong Coal Mining Administration in China. Niu and Gu (1982) reported that over the 30 years period to 1982, 27 serious accidents due to falls of large areas of roof occurred including that involving multiple fatalities at the Wa Jing Wan Mine in 1961.
The Jurassic coals at Datong are characterised by their strong, massive sandstone and conglomerate roofs which fail suddenly over large areas, crushing pillars and creating destructive wind blasts (Song & Xu , 1992). Table 2.2 indicates the damage suffered by some coal mines due to roof collapse and resulting wind blasts at Datong Mining Area.
2.4.4 India
The geo-mining conditions in some Indian coalfields are similar to that in Australia. In the coalfields of central India and also in isolated locales of Jharia and Ranigunj coal basins sometimes coal measure rocks are comprised of uniformly massive sandstones. Bord and Pillar is the predominant mining method in Indian coalfields. In longwall mining, the overlying roof under massive strata conditions is very difficult to cave especially if the depth of mining is shallow. The main roof above the parting plane either sags down or rest on the caved material much further back from the face line or breaks
16 beyond the face line and hangs in cantilevers. Breakage of main roof ahead of the face line subjects the face supports to severe loading. Many failures of longwall faces due to collapse have been reported due to inadequate ground control and support capacities (Sarkar, 1982). In 1990, Churcha Colliery commenced longwall extraction under a massive dolerite sill. A 155 metres wide panel was retreated about 116 metres when heavy face weightings began and at 198 metres of retreat the whole face was lost. Wind blast events were experienced in the longwall panel at Churcha Colliery prior to the face collapse (Gupta & Ghose, 1992).
Table 2.2
Wind blasts in Datong Coal Mines, China (after Song & Xu, 1992) Mine Panel Date Goaf Area of Remarks Area roof collapse (m2) (m2) Wa Jing Quing Yang July 63 010 22 310 One person killed, twelve injured Wan Wan shaft 1960 in roadway. Six surface cracks of 14402 lengths 60 to 70 metres and width 0.1 to 0.4 metres and surface subsidence 0.3 to 0.7 metres occurred Wa Jing Quing Yang October 184 047 163 000 Fourteen persons killed, nineteen Wan Wan shaft 1961 injured, severe damage to 14832 equipment and ventilation system. Yan Ya 11301 October 30 001 17 370 150 chocks were damaged, 1961 subsidence basin of area 7850 m2 and depth 4 metres was formed Ma Ji 2402 June 151 280 125 302 10 stoppings destroyed. Mine car, Liang 1975 a switch and a fan were damaged or moved.
2.5 FACTORS AFFECTING WIND BLAST OCCURRENCE
For underground coal mines, the following five factors have been identified, some of which were considered necessary in order for a significant wind blast to occur (Fowler, 1997):
1. A strong massive roof, such as some sandstones and conglomerates, which does not cave readily;
17 2. Consequently, an extensive area of roof “hanging up”;
3. Little coal left in the standing goaf in forms of stooks or pillar remnants;
4. Initial total extraction in a previously unworked area;
5. Few roadways intersecting the goaf.
In this thesis, the wind blast phenomenon in longwall panels is studied and accordingly, the emphasis in the following literature review is towards longwall mining.
The caving behaviour of roof over a longwall panel is related to the stress distribution around the panel; strength, discontinuity and thickness of the immediate roof/rock mass; geological factors including depth of cover; stress interaction effects due to adjacent workings/goafs; longwall panel design, layout and orientation vis a vis geological discontinuities and interaction of face supports with the immediate roof.
2.5.1 Stress distribution around a typical longwall panel
The stress distribution around a typical longwall panel is illustrated in Figure 2.1 (Whittaker, 1982). The impact of the stress distribution and their effect under different strata conditions is as follows (Singh & Singh, 1982):
1. Vertical stress constitutes the main motive force inducing roof rock deformation. The abutment stress in advance of the face varies in between 2 to 7 times the geostatic pressure. Its magnitude depends upon the competency of the formation. Strong roofs sustain abutment loading without any damage and are normally free from vertical cracks or “preliminary Assuring” observed in weaker formations;
2. The abutment zone follows a relaxed zone in and around the face where the vertical stress is reduced to negligible. The immediate roof is however subjected to lateral stress due to unbalanced stress condition. In coalfields, where the ratio of lateral/horizontal thrust to vertical stress is high, outward horizontal flow of the mass near the face and opening of inherent cracks can result. Under the influence of horizontal stress, the spalling of coal face, delamination of the roof rock beds, and opening of the transverse joints is reported under different conditions. The effect of
18 the horizontal stress depends upon cohesive strength or cohesion along the contact planes, joint character and filling undulations along bedding planes, degree of lamination and the lamination thickness and resistance offered by the supports or goaf fillings. Differential movements of roof beds induce delamination due to which, a massive or strong roof bed is reduced to thin incompetent formation;
3. Under a normal condition, the coal face ‘yields’ by a certain degree in advance of the face. The ‘yield’ zone is the area between the faceline and the point where maximum front abutment pressure occurs and occupied by strata that has passed its failure limit (Wilson, 1982). The width of the yield zone ranges from 0.45 to 2.25 times the mining height, and it is widest at the centre of the panel (Peng & Chiang, 1984). The convergence in addition to the face yield, induces shear stress in the roof rock mass. This type of shearing force is a common experience in a beam or a cantilever similar to a longwall roof. The shearing strength of a rock mass is much higher than the tensile strength and the roof cantilever or a beam is more likely to fail under tension. However in presence of prominent joint planes with orientation nearly parallel to the face, failure under shear is frequent, causing face collapse and accidents under strong roof conditions;
4. The flank strata pressure redistribution (side abutment pressure) have an influence on the stability of adjacent gate roadways. With retreat mining, the flank abutment has a major bearing on the positioning of the development headings for the adjacent longwall. For longwall workings in U.K, a critical zone, 0.01 to 0.06 times depth as measured from ribside, was defined, which is of high strata stress inducted by adjacent mining and can create stability problems to coal headings (Whittaker & Singh, 1979).
2.5.2 Caving characteristics of a strong, massive roof in longwall workings
Massive roofs comprising conglomerates and sandstones are often difficult to cave in total extraction panels and give rise to dynamic phenomena such as periodic weighting and wind blasts. Where there is no interburden between the working horizon and the 19 overlying massive strata or where the thickness of interburden is less than about twice the extracted seam height, wind blast have occurred in narrow longwall panels. However, where a greater thickness of friable interburden is encountered, it significantly modifies the caving characteristics of the massive strata, tending to obviate wind blast (Simpson, Hebblewhite & Fowler, 1997).
Under massive strata conditions, typical in parts of Australia, India and the USA, parting plane controlled caving is more likely, rather than the bulking factor controlled caving typical of British conditions where the roof strata caves directly behind the face support. In the former, the caving height is determined by the location of some dominant parting plane such that there will be insufficient caved waste from the immediate roof to provide support to the main roof behind the face line. If the main roof were to break off ahead of the face line, the face would be subject to severe loading (Smart, 1989). Galvin (1983) used a combination of elastic thin plate theory and empirical calibration to develop an equation to calculate the panel dimensions required to induce dolerite sill failure for South African conditions. Wilson (1986) used the elastic beam theory to determine the bridging capability of massive beams. The “Voussoir” arching, mechanism has been used by Beer and Meek (1982), in the analysis of safe spans of extraction under inclined or horizontal bedded deposits for a range of plate dimensions and material properties. Wold and Pala (1986) applied this method to strata conditions at Ellalong Colliery, New South Wales with accurate predictions. For a first fall dimension of 127 metres x 150 meters the plate thickness at failure, as a function of the plate aspect ratio, Young’s modulus and compressive strength, were derived from the design curves of Beer and Meek. For a range of Youngs’s moduli, a plate thickness of 4.7 to 5.7 metres was indicated at failure. Depending on the spans, thickness, Young’s modulus, compressive strength and shear strength of the plate, two main type of failure i.e. snap through and shear are envisaged, as shown in Figure 2.2. Shear failure at the abutments would tend to occur at small span/thickness ratios and low shear strength. Snap-through failure tends to occur at larger span/thickness ratios as a result of either plate buckling or compressive crushing at the centre or abutment hinge points.
20 Corner peak pressure due
of strota loading
pressure(p)
dtttZS Coved waste
due to single pone! working
Figure 2.1 The stress distribution around a typical longwall panel (after Whittaker, 1982)
SHEAR FAILURE AT ABUTMENTS
VOUSSOIR MECHANISM SNAP-THROUGH FAILURE
Figure 2.2 Failure mechanism for bedded roof: shear and snap-through
(after Wold & Pala, 1986 )
21 Massive lithologies, such as sandstones and conglomerates in the Sydney Basin, reduce the subsidence factor (the ratio of the surface subsidence to the seam working height). In Sydney basin, the vertical movement is 30 % to 60 % of the thickness mined (Holla, 1986).’Subcritical’ extraction, with Width (W) to Depth (H) ratio in the Newcastle area ranging from 0.3 to 0.8 is common, and the resulting subsidence may be 10 - 50 % of the maximum possible value. Maximum possible subsidence for ‘supercritical’ extraction is usually about 60 % of the seam’s working thickness in the Sydney basin. For deep longwall mining beneath massive sandstone at Ellalong Colliery, low subsidence factors of 26 % of extracted thickness, were experienced for supercritical extractions (Holla & Hughson, 1987). The caving behaviour of conglomerates in the Newcastle Coalfield has been observed both by subsidence monitoring (Kapp 1984; Creech 1995), and by drilling over the longwall panels, at Teralba Colliery, for example. Observations reveal that the conglomerate retains its rigidity, failing in a manner which generates open voids, and it influences caving behaviour even up to 100 metres above the worked seam. At Newstan Colliery, where the conglomerate spanning the panel has a thickness of more than 35 metres, subsidence was reduced from the expected over 1 metre to around 0.2 metre. The occurrence of numerous large seismic events prior to and following a wind blast and their source location indicated that instability within the conglomerate channel was the cause and the actual block of material likely to be a basal piece from this unit (Creech, 1996).
2.5.3 Height of caving zone
The caving zone across a panel width is believed to assume an arch shape with the highest point in the panel centre, decreasing parabolically towards the sides (Peng & Chiang, 1984). The caving height is dependent on geological conditions such as the location of the thick competent strata and the bulking factor of the caved strata. Caving stops when a self-supporting stratum is encountered or the overlying strata are fully supported by the caved fragments. Additionally, caving height is also affected by the seam extraction thickness and overburden depth. If a strong stratum is encountered in the roof, the caving zone will reach its maximum height after the face has retreated some
22 distance outbye and the overhanging stratum on reaching a critical span, fails when it is no longer strong enough to maintain self support. The caving height assumes importance for support design/load specifications, magnitude of abutment stresses and also magnitude of wind blasts, if caving is arrested due to strong overhanging stratum in close proximity of the seam with inadequate goaf packing by the collapsed immediate roof.
2.5.4 Impact of longwall panel width on caving behaviour
Severe periodic weightings have been reported in longwall panels at Clarence Colliery, mining the Katoomba seam with massive sandstone units 25 metres and 50 -70 metres above it (Mills & Grady, 1998). The problem was more severe for the 200 metres wide face in comparison with the 160 metres wide face, with the downward movement of the major unit 50 -70 metres above the seam along vertical fractures assuming significance in the former instance.
Numerical modelling (Peng & Chiang, 1984) reveal the effect of panel size on front and side abutment pressures in a longwall panel. With no caving in the goaf, the peak front and side abutment pressures, either in the roof or in the coal, increase linearly with the panel width. If caving in the goaf is considered, there is a critical width for a fixed stratigraphic sequence above which roof caving will reduce the maximum front abutment pressure in the coal near the panel centre. It will however not affect the maximum front abutment pressure, near the T-junction or corner, which continue to increase with panel width. Hence with increase in panel width, the peak front abutment will move from panel centre towards the T-junction corners.
Numerical modelling using the finite difference code FLAC (Gale & Nemick, 1998) for the South Bulga Mine, where the longwall mining panels under massive sandstone experiences severe periodic weighting, reveal the caving and fracture mechanism. The absence of weak bedding planes in the upper roof and the moderate strength of the rock prevents frequent formation of fractures in the roof. Major sub-vertical fractures develop at less frequent intervals forming large caving blocks above the longwall face.
23 It was predicted that significant fracturing above the longwall face and supports may occur every 10-20 metres. The geometry of these blocks was defined by;
1. Failure along a weak layer in the roof above or ahead of the face followed by; and 2. A fracture network forming at this zone and extending down to meet the longwall face.
The extent and nature of the fracture zone around total extraction longwall panels, are linked to the seam thickness or height of extraction. However, physical and numerical modelling studies have not been completely validated from field measurements.
Currently, the combined extent of caving and fracturing above the seam is considered to be in the region of 50 times the seam thickness for a 200 metres wide longwall panel, with an approximate linear decrease with decreasing panel width (Styles, Bishop & Toon, 1992). The depth of fracturing in the floor is significantly less than for the roof While attempts to delineate the fracture zones around total extraction panels have been made using extensometers (Peng & Chiang, 1984), microseismic monitoring holds the potential of being able to determine and delineate the full extent of fracture zone, in real time and at relatively low cost.
Figure 2.3 illustrates the fracture profile associated with longwall and caving shortwall geomechanics (Frith, 1993). Normal longwall geomechanics involves strata breakage ahead of the face in an arcuate pattern. This was also confirmed by seismic monitoring at Gordonstone Mine in Queensland (Hatherly, Xun, McKavanagh, Dixon & Devey, 1995) and shown in Figure 2.4. In caving shortwall, the massive strata is expected to bridge the goaf, or to fail well behind the face. High convergence rates are expected where the breakage line is very close to the face representing a transition between these two goafing mechanism.
At longwall panel 5, Newstan Colliery, NSW, having a width of 225 m, excessive periodic weighting was observed. In order to obviate the strata control difficulties, the subsequent panel was split longitudinally into two blocks each 95m wide between
24 RETREAT OlRECTiON
I 1 II I 1 I
0 - Looded lone
Figure 2.3 Longwall and shortwall geomechanics (after Frith, 1993 )
1050 1050
1000 Tailgate Maingate 1000 950 - J 950
900 - 900 :• a I 850 - •c h 850
800 - LW face 800 E J ---- O) a 750 - 750 c . • . ».*. ** 7 CD • • 700 - 700 •
Q 4 650 - 650 I LW103 600 - T T T T 600 gateroads gateroads 550 - 550
500 - 500
450 - 450
400 -- 400 -200 -150 -100 -50 0 50 100 150 200 250 300 350 Distance (m)
Figure 2.4 Distribution of microseismic events relative to longwall face position, Gordonstone Mine (after Hatherly et al., 1996).
25 gateroad centrelines. In the narrowed 6 and 7 longwall panels, wind blast activity was found to occur exclusively where the conglomerate channel was thick enough to bridge the goaf and where it occurred within 7 m of the working roof or approximately twice the extraction height (Simpson, Hebblewhite & Fowler, 1997). The friable immediate roof caved readily behind the longwall chock supports, but failed to fill the goaf void, resulting in significant gaps above and behind the chocks. However, it bulked up sufficiently to fill the goaf void when the base of the conglomerate channel was sufficiently above the seam horizon, cushioning any failure of the overlying channel strata, thereby obviating wind blast. Face lengths for subsequent panels were increased to 127 m between gateroad centrelines. However, for part of the extraction of longwall panel 8 and 9, with similar geology, wind blast remained a hazard. Chapter 6 of this thesis deals with the Newstan Colliery case study in more detail.
2.5.5 The magnitude of a wind blast
The magnitude of a wind is related to the energy released by the event. The overhanging massive roof element has a potential energy (mgh) given by product of its mass (m), acceleration due to gravity (g) and height of the workings (h). This potential energy converts to other forms of energy during the fall of the roof and the resulting wind blast. The mass of the roof element (pAH) is the product of its density (p), plan area (A) and thickness (H). Consequently the potential energy of the roof element is equal to pAHgh. As the goaf volume (v) is the product of the plan area (A) and height of working (h), the potential energy can be expressed as equal to pHgv. Consequently energy effect of the roof fall is a function of the goaf volume for a given thickness of roof element. Part of this energy is manifested in the wind blast, the magnitude of which is related to the amount of air displaced from beneath the falling roof element.
It is considered that the following factors affect the magnitude of a wind blast (Fowler & Torabi, 1997):
1. Geological factors such as the structure of the roof strata, mechanical properties of both the rock fabric and of the discontiniuities, influencing the way the roof fall;
26 2. Geometrical factors such as the plan area, thickness of the falling roof element and the distance through which it falls along with plan area of the standing goaf
2.5.6 The intensity of a wind blast
The intensity of a wind blast relates to its damage effect upon the working place and refers to the severity of the event as observed at the place of observation. Wind blast parameters which are considered important in assessing the intensity include the following (Fowler & Torabi, 1997):
1. Peak dynamic pressure (peak wind blast velocity), rate of rise of velocity, velocity rise time and maximum excursion (air flow distance) relevant for drag sensitive elements such as mine plant and equipment;
2. Peak overpressure, rate of rise of overpressure, overpressure rise time and impulse (the total area under the positive phase of the graphs of overpressure with time) relevant for overpressure sensitive elements such as eardrums of mine personnel.
For a wind blast of a given magnitude, the intensity at any location will depend mainly upon the geometry of the panel. For a given panel geometry, an increase in wind blast magnitude will generally result in an increase in intensity.
2.6 WIND BLAST CASE STUDIES
The occurrence of wind blast, prior to the recent Coal Mines (Underground) Regulation 1999, is not a reportable incident and so information on such incidents, other than those associated with reportable personal injury, is not plentiful in the public domain. This section presents some case histories of wind blast incidents in Australian coal mines working the Great Northern seam by the Bord and Pillar mining system and which have resulted in reportable injuries.
27 2.6.1 Wallarah Colliery
Wallarah Colliery is located within the Lake Macquaire district of the Newcastle Coalfield, 100 km north of Sydney and operated presently by Coal Operations Australia Pty Ltd. A wind blast occurred during the extraction operations in 3 Panel Left, North- East 5 headings on 27 November, 1989 (Fowler, 1997). The extent of the workings in 3 Panel Left at the time of the incident are shown in Figures 2.5 and 2.6.
The mine extracts the Great Northern seam by the Wongawalli (lift and fender) system of pillar extraction, under a conglomerate roof. At the time of the incident, coal cutting was underway in a lift in the fender and nine personnel were in the area of cut-through and the split. Four miners in vicinity of the continuous miner were injured with the Deputy, Peter Kuba (PK) seriously injured after being blown into the rear of the continuous miner. The other five, who were in the area between B and C heading escaped injuries despite being knocked over.
It was reported that there were two phases to the wind blast: an initial flow of air outbye from the 3 Panel Left followed by a flow inbye towards the 3 Panel Left goaf. The mine deputy received injuries during the latter “suck back” phase.
2.6.2 Myuna Colliery
Myuna Colliery is also located within the Lake Macquarie district of the Newcastle Coalfield and operated by Powercoal Pty Ltd. It mines the Great Northern, Wallarah and Fassifern seams mostly from first working due to limitations imposed by overlying lakes and foreshores. However, the 428 Panel, where a significant wind blast associated with a major goaf fall occurred on 15 February, 1990, forms part of pillar extraction area situated under a massive conglomerate roof towards the north-eastern comer of the Great Northern workings (Fowler, 1997).
Figure 2.7 illustrates the extent of workings at the time of incident. A group of 15 pillars had been extracted and an area of almost two hectares was ‘hanging up’. Twelve persons were in the section and the deputy, on hearing roof noise and judging the fall to be imminent, instructed all inbye personnel to take refuge against the pillar in 10 cut-
28 V
Figure 2.5 Location plan, significant wind blast, 3 Panel Left, Wallarah Colliery (after Fowler, 1997) W///////////////////X V//////////A H V///////////A - V////////////, 1SV3
HldON 30 £ 60 3 e 2.6 Plan of site of significant wind blast, Wallarah Collii 69992
lO o a met res
Figure 2.7 Plan of site of significant wind blast, Myuna Colliery (after Fowler, 1997)
31 through between 2 and 3 headings. Seven persons took refuge at position marked “A” and one behind the nearby Eimco. Insufficient time was available to warn the remaining four personnel in headings outbye. The extent of the roof fall involved an area of approximately one hectare and reached an estimated height of 10 metres. All brattice ventilation stoppings along the goaf edge and five plaster board stoppings in the section were blown out but no major equipment was disturbed and no machinery damaged.
All personnel, excluding the vehicle driver, reported being knocked down by the wind blast or by the suck back, some being pushed long the floor and 11 were taken to hospital having sustained some degree of peppering, cuts, abrasions and contusions and were released after treatment (Fowler, 1997).
2.6.3 Endeavour Colliery
Endeavour Colliery, formerly known as Newvale No. 2, is also located within the Lake Macquaire district of the Newcastle Coalfield and owned by Powercoal Pty Ltd. It currently mines the Great Northern seam and the 300 Panel, which witnessed a wind blast associated with a major goaf fall and followed with an explosion, is situated at the western extremity of the workings adjacent to the Munmorah Colliery.
The 300 Panel employed partial extraction using the bord and pillar method. Figure 2.8 illustrates the extent of working at the time of the incident. The Panel was bounded on one side by unmined coal, on the second by the main “buff’ headings and flanking barrier pillars and on the third by barrier pillars adjoining Munmorah Colliery. It was linked to the rest of the workings by a five heading development. Goaf areas had been formed at the time of the incident in the sequence marked 1 to 6. The immediate roof of zones 1 to 4 is reported as having collapsed but the falls had been noted as “irregular and shallow” (Anderson, 1997). An area of 30 metres by 100 metres was standing immediately adjacent to the face within zones 5 and 6. Mining operations had been temporarily suspended while a minor repair to the continuous miner was underway. Eight persons were present in the section and four in vicinity of the continuous miner. Judging an imminent roof fall, one miner took refuge behind the continuous miner while three ran outbye. They were in vicinity of the intersection of 22 cut-through and 3 32 Figure 2.8 Plan of site of significant wind blast, Endeavour Colliery (after Anderson, 1997)
33 heading when knocked to the ground by the wind blast. Both shuttle car drivers had remained in their seats and one was knocked from his machine and other was “peppered” with coal particles and had his helmet thrown off. The remaining two crew were further outbye in 20 cut-through and were unaffected by the wind blast.
Within seconds of the wind blast an explosion occurred. The explosion’s violence was of an order of magnitude greater than the wind blast (Anderson, 1997). The two men outbye also reported being thrown 2-3 metres to the ground by the explosion and suffered burn injuries. Intense heat was felt and visibility was virtually nil. Fortunately, all eight personnel survived and were able to grope their way in three groups to fresh air near 9 cut-through about 13 pillars from the face.
From the available evidence it appears that the methane explosion was due to methane expelled from the goaf during the wind blast and the area of ignition lay in 21 cut through between 1 and 3 headings which is one pillar outbye the general goaf line. The most likely source of ignition has been identified as the shuttle car cable at the intersection of 21 cut-through and 2 heading.
Damage was restricted to stoppings but was not confined to 300 Panel and the few roadways leading to its entry. Significant damage also occurred to plaster board stoppings in 8 West Panel, some 2 km inbye.
2.7 REDUCTION OF WIND BLAST HAZARD
It is possible to reduce the wind blast hazard by either reducing its magnitude, its intensity or eliminating such incidences altogether.
2.7.1 Reducing magnitude
The wind blast hazard could be reduced or eliminated by reducing the area of roof that tends to overhang either by panel design or by artificially promoting caving. Other potential intervention strategy includes reducing the height of fall by either restricting extraction height or by packing the goaf.
34 2.7.2 Improving caving
The caving characteristics of massive roof in longwall workings has been reviewed in section 2.5. The potential strategy to improve roof caving could include change in panel dimensions, in particular face width. However geo-technical or subsidence considerations may impose constraints upon such changes. Improvements in caving and face loading can result from the proper orientation of the panel layout with respect to the structural discontinuities and to the direction of the principal stress. Overseas experience (Wilson et al., 1976; Kidybinski, 1962; Fritzshe, 1962) indicate that in case of massive sandstone roofs, the best caving conditions occur when the major joint set is aligned parallel with or at a slight angle (of the order of 10 degrees) to the extraction line. Shepherd (1996) proposed a triangular design grid accounting for angular variables of faults, stress and joints to optimise design layouts. In a stress field characterised by high horizontal stresses, vertical joints are of lesser importance than horizontal defects. The rate of retreat could also influence the caving, as it is known that conditions deteriorate when a face stands for several days. Proper selection of supports and setting pressures could also promote good caving and face conditions. The literature for caving under massive units however, is primarily concerned with high abutment stresses, periodic weighting and poor face conditions associated with wider longwall panels. The knowledge of actual fracture zones and failure mechanisms ahead of the face and in the floor is also limited.
2.7.3 Induced caving
Blasting and hydraulic fracturing of the roof are the two techniques used to artificially induce the roof to cave in. The treatment could be applied in one of the two ways: either as a method of pre-treatment prior to the commencement of total extraction in a panel or alternatively, as a method of post treatment where an extensive area of roof is already hanging up. Foster (1979) reported the development of a support with an incorporated hydraulic mechanism that initiates fractures in the roof at the rear of the face support roof canopy.
35 2.7.3.1 Hydraulic fracturing/Water infusion
Niu and Gu (1982) reported the application of water infusion to promote successful caving at Datong Mining Administration Area of China. The mineralogy of the sandstone was such that it was susceptible to loss of strength on immersion in water. Here infusion rather than hydraulic fracturing may have been the most significant factor in the failure process.
Two water injection trial were undertaken during 1988-89 at Newstan Colliery (Holt, 1989) During the second trial, roof caving was successfully induced as evidenced by reduced longwall chock pressures. Goaf at Newstan Colliery was not accessible for visual verification of caving of the massive conglomerate channel due to the caved immediate friable roof. The successful use of hydraulic fracturing at Moonee Colliery is discussed in Chapter 5 of this thesis.
2.7.3.2 Blasting
Attempts to induce caving by blasting have been reported in longwall mining with caving under massive sandstone, in India. A longwall panel of 80 metres face length was worked at Bijuri Colliery during 1978 and 1980 (Sarkar & Singh, 1987). The depth of working was 66 metres. After the face had advanced 20 metres without the roof caving, efforts were made initially to blast the roof in the goaf by shotfiring. After the face had been worked up to 60 metres of advance with no substantial caving, it was stopped and six holes were drilled over the panel from the surface and charged with a number of decks of explosives. Holes were fired in two instalments. The operation ensured complete filling in the goaf but due to non-feasibility of repeating such operations due to high costs, the face was abandoned. Blasting was also resorted to at the 104 metres face length mechanised longwall panel at Pathakhera Colliery in 1982. After 25 metres of face advance, an attempt was made to induce caving by firing twenty five 5.5 metres deep holes in the goaf at the back of the face. This resulted in local falls, which were not sufficient to fill the goaf and only after the first major weight, after an advance of 75 metres from the start line, did complete filling of the goaf occur. In the second panel also shot firing had to be resorted after 56 metres of face advance which 36 led to some superficial caving. The first major weighting ensuring complete filling of the goaf occurred after 64 metres of face advance (Sarkar & Singh, 1987).
Attempts were made to induce caving by blasting at Moonee Colliery, New South Wales at longwall panel no. 2 after the face had retreated more than 200 metres. Four holes were drilled from the surface and charged in decks. The attempt failed to induce caving.
2.7.4 Reducing goaf fall height
This could be achieved by either restricting the mining height or packing the goaf.
2.7.4.1 Restricting mining height
Physical modelling (Fowler, Torabi & Daly 1995) revealed that in the simulated roadway, during the phase when the roof element was falling at its terminal velocity, wind velocity was lesser in magnitude to the peak velocity. Also, the duration of the event is proportional to the fall height. However, in the real coal mine situation the terminal velocity is never achieved as the height of fall is less than critical (fall height at which the roof element reaches terminal velocity) and the roof element accelerates till it reaches the floor. McPherson (1995b), by analytical modelling, reported higher peak pressures in the goaf for lower initial fall heights for the same thickness of roof elements, reasoning that resistance to air displacement is greater for smaller extraction heights.
The values of the parameters of wind blast at Newstan Colliery were much lower compared with those of Moonee Colliery although the height through which the roof element fell was greater. However, the thickness of the friable immediate roof was greater and consequently, the fall of roof element at Newstan was cushioned by the debris in the goaf which also reduced the distance the roof element fell. The issue of the impact of height of fall on wind blast is further discussed in the Chapters 5 and 6 of this thesis while discussing the Moonee Colliery and Newstan Colliery case studies.
2.7.5 Packing the goaf
Packing the goaf is not a solution compatible with mechanised longwall retreat mining as currently practised. Hydraulic stowing, for example, would involve a different system of mining to be employed and involve extra expenditure.
37 2.7.6 Reducing wind blast intensity
For a given mass flow rate, an increase in cross sectional area of the roadway would reduce the velocity. Consequently a potential strategy would be to increase the total cross sectional area of the workings that intersect the goaf. Increasing roadway and cut through width is generally not practicable and mining height is often determined by geo mechanical factors and the considerations of coal quality. Consequently, wind blast intensity in the working place can be reduced by increasing the number of openings which intersect the goaf. “Leakage roadways” connected to goaf with a thin stopping which gets destroyed in the event of a wind blast, facilitating an alternate dissipation path for the pressurised air has been reported by Song and Xu (1992).
2.7.7 Reducing the potential for methane explosions
During a wind blast, the incursion of methane /air mixtures into the workings place may be of a short duration and difficult to detect if it is subsequently sucked back following the wind blast and by the ventilation flow (provided that the circuit has not been disrupted by the wind blast). Possible intervention strategy could be to redefine the extent of the hazardous zone on intake side of the panel.
When roof collapses of large areas are anticipated in a mine where flammable gas or dust may be present, the potential for an explosion exists due to the adiabatic compression of the air in the goaf and the resulting high temperatures (McPherson, 1995a). While such an occurrence would be rare, intervention strategies could include the modifying of the caving behaviour of the roof through induced caving.
38 CHAPTER 3
AUSTRALIAN RESEARCH ON WIND BLASTS
3.1 INTRODUCTION
The Australian initiative to undertake research on wind blast arising from roof falls in coal mines came about as the result of incidents at Wallarah Colliery in November 1989 (Stothard, 1989), which resulted in serious injury to a deputy, and at Myuna Colliery in February 1990 (McCarthy, 1990) which put eleven miners at risk.
This chapter presents the main work undertaken by the wind blast research group at the School of Mining Engineering, University of New South Wales over the period January 1991 to July 1997 i.e. the period prior to the author’s involvement in the wind blast project.
The work included the design and development of the Mkl Wind Blast Monitoring System (Fowler, 1994), a laboratory physical model (Fowler & Torabi, 1997) and the preliminary development of a wind blast numerical model. This was followed by a comprehensive series of laboratory tests using the physical model (Fowler, Torabi & Daly, 1995) and field monitoring at underground coal mines including Wallarah, Cooranbong and Newstan Collieries. However, significant wind blast events were recorded only at Newstan colliery. Some of the research tools developed during the above period have been utilised by the Author and others, after necessary modifications, in the current phase of research.
3.2 WIND BLAST MONITORING SYSTEM
The wind blast monitoring has utilised a system specially developed to record the air overpressures and velocities during wind blasts caused by massive roof falls. The equipment comprises a wind blast data logger, four sensor pods and a hand held interface. The equipment is certified and approved for use in hazardous locations in
39 underground coal mines in both New South Wales and Queensland, Australia. The Wind Blast Data Logger (WBDL) is an intrinsically safe apparatus built into a flameproof enclosure. Four sensor pods are deployed to monitor absolute pressure, differential pressure and, hence, wind velocity in the ranges 0 to 200 kPa, ±14 kPa and ± 150 m/s respectively (after recent refurbishment). The hand held interface is an intrinsically safe module used to program the wind blast data logger and to transfer data from the latter to a personal computer. Figure 3.1, 3.2 and 3.3 illustrate the Wind Blast Data Logger, Sensor Pods and the Hand Held Interface respectively.
The cylindrical body of each transducer pod is in the form of the static head of a pitot tube, chamfered at the nose. It contains two differential pressure transducers (Honeywell Micro Switch model 143PC03D). The one in the nose of the pod responds to the stagnation (total) pressure while the second, at the cross port, responds to the static pressure. The difference between the output signals of the transducers is directly proportional to the differential (velocity) pressure and to the square of the velocity. This difference is passed through a 5th order low pass filter with an approximate corner frequency of 500 Hz. The purpose of this filter is to prevent aliasing effects caused by the channel sampling frequency of 1 KHz.
The differential pressure transducers originally specified were Honeywell MicroSwitch 163PC01D37 which provided a differential pressure range of ±1 kPa and, consequently, a wind velocity range of ± 40 m/s. However, monitoring at the first longwall panel at Moonee Colliery revealed that the range was inadequate and the opportunity was taken, between monitoring longwall no. 1 and longwall no. 2 panels, to retrofit transducers to the specifications given earlier.
The pods had initially been calibrated in a wind tunnel to 40 m/s. After refurbishment they were recalibrated at the Defence Science and Technology Organisation’s Aeronautical and Maritime Research Laboratory in Melbourne, which provides the highest velocity compared to any other facility in Australia. The deviation from the theoretical was less than 2 % at air speeds up to of 95 metres per second, the upper limit for the wind tunnel. (Fowler & Sharma, 2000).
40 Figure 3.2 Sensor pod
41 Figure 3.3 Hand Held Interface
'vv *
Figure 3.4 Wind Blast Physical Model
42 Each pod is also equipped with a direct pressure transducer (SenSym SCX30AN) to measure the absolute pressure from which overpressure can readily be calculated.
The associated electronics in each pod is powered from the Wind Blast Data Logger via zenner barriers and each returns, via the 500 Hz low pass filters, two DC current signals, one proportional to the velocity pressure and the other to the static pressure.
The data acquisition unit is the ‘heart’ of the system and continuously monitors each of the current loop circuits at a sampling frequency of 1000 scans per second, recording an event only when the preset trigger levels are exceeded (Fowler, Torabi & Daly, 1996).
The triggers comprise an air velocity level, an absolute pressure level and a duration (delay time). Different trigger levels may be set for each pod. Upon triggering, the data acquisition unit stores differential and absolute pressure information from all four pods in binary (eight bit) format for the two seconds preceding the triggering and for the following 6 seconds. Also stored are the integrated absolute pressure for each of the four pods, the case temperature and the date and time of the event. Sample file format for overpressure and velocity channel, displaying the wind blast “raw data” for one pod, is printed in a tabular mode in Appendix A.
The unit can store data from up to 15 eight second events and there are two modes of operation. In the fixed mode, the first 15 events are preserved and all subsequent events ignored. In the roll-over mode, the first 15 events are recorded but subsequent events are stored by overwriting the earlier ones.
Data is stored in a non-volatile mode such that when the mains power is removed from the WBDL all stored information from previous significant events is retained. The unit is also provided with battery back up which provides the ability to operate for up to 16 hours after the mains power disconnection.
3.3 PHYSICAL MODEL DEVELOPMENT
In order to investigate the wind blast phenomenon, a physical laboratory model was constructed as shown in figure 3.4. The model, at a scale of 1:125, comprises
43 representations of solid coal, 12.5 m square pillars, and 4.75 m wide by 2.5 m high headings together with cut-throughs and goaf areas "sandwiched' between an aluminium base plate and a Perspex lid. The goaf fall is represented by a group of four square pistons contained within a piston box representing an area 56.5 metres square. The head of each piston is a hollow aluminium box which was originally provided with a valve incorporating an occlusion disk containing three pressure relief holes. The holes could be partially or totally occluded from outside the piston box by means of a special tool. The pistons are capable of being loaded internally with lead inserts and lead shot ballast and externally with steel weights such that they are capable of exerting a maximum pressure of 10 kPa. The base plate which forms the underside of the box is equipped with four ports which function as pressure tappings and four others through which extension rods can pass in order to transfer the vertical motion of the pistons to displacement transducers or accelerometers. The abutting pistons are suspended electro- magnetically and released in computer controlled time sequences. The schematic plan and front elevation of the wind blast model is shown in figure 3.5. The model is instrumented with Hewlett Packard displacement transducers (LVDTs) and Environmental Equipment’s piezoelectric accelerometers, responding to the changes in relative vertical position and acceleration of the pistons respectively. It also has Honeywell Micro Switch pressure transducers, detecting changes in pressure above and below the pistons, and Airflow Developments’ miniature pitot tubes and associated Honeywell MicroSwitch differential pressure transducers responding to the flow of air in the model openings.
3.3.1 Computerised control and data acquisition
Control of the model and acquisition of data is arranged by an Apple Macintosh computer equipped with a National Instruments NB-MIO-16XL-18 input/output board installed in the Nubus slot.
The NB-MIO-16XL-18 is a multifunction analogue, digital and timing input/output board. It includes a 16 bit analogue to digital converter, 16 multiplexed inputs, two 12-
44 Figure 3.5 Schematic plan (top) and front elevation (bottom) of the wind blast physical model (after Fowler, Torabi & Daly, 1995).
(Test no. 122S| Air velocity in the simulated roackvay (m/s) Secondary phase
-10 -
-20 -
-30 - 1000 2000 ■Time (ms)
Figure 3.6 Time history of air velocity in a simulated roadway, wind blast physical model (after Fowler, Torabi & Daly, 1995).
45 bit digital to analogue converters, 8 digital input/output lines and three 16-bit counters/timers.
The data acquisition and control program, written in National Instruments Labview version 2.2 running under Macintosh System Software version 7.5.5, controls the release of the pistons, which simulate the goaf fall, in time controlled sequence and acquires subsequent output from the LVDTs, accelerometers, gauge pressure transducers and differential pressure transducers. Labview software is based on the concept of virtual instruments or VI, a software emulation of test equipment that creates, analyses and displays data much as physical instruments do. It facilitates data capture and analysis without requiring access to dedicated test equipment.
Its programming language called G, is a graphic language where the manipulation of objects replaces the more conventional writing of code. It is used to create applications for data acquisition and management, signal and transient analysis, and process control.
The data acquisition and control VI does the following functions:
Sending timed signals via four of the digital input/out lines on the NB-MIO-16XL-18 board to change the state of the electromechanical relays on the SC-2062 signal conditioning board, and hence switch off the currents to each of the electromagnets.
Just prior to the sending of timed signals, it starts to acquire data from the pressure transducers, LVDTs and accelerometers via fifteen of the analogue inputs on the NB- MIO-16XL -18 board and its analogue to digital converter.
It subsequently displays on the computer screen, over an eight second period, binary outputs corresponding to the vertical displacement for the two pistons, acceleration for two pistons, differential pressures at each of the eight pitot tubes and the three gauge pressure transducers and saves the raw binary output data on the disk.
A suite of Vis has also been created to manipulate and display the binary data facilitating the study of the detailed histories of piston displacement and acceleration and air pressures and velocities in both the time and frequency domains.
46 Figure 3.6 illustrates the wind velocity time history illustrating the three phases of wind blast as observed by physical modelling (Fowler, Torabi & Daly, 1995; Fowler & Torabi, 1997). Due to the limitations of the physical model, the study was confined to fall of thin roof elements simulating thickness less than 4 to5 metres and thus limits the wind blast modelling for field studies in longwall panels where the falling roof elements have thickness varying from 6 to 10 metres. Moreover, the wind blasts monitored in the field exhibit two phases as discussed in Chapter 4 section 4.3.1 as opposed to three phases from the physical model in Figure 3.6.
3.4 NUMERICAL MODEL DEVELOPMENT
The numerical model under development (Fowler, 1997) is based upon the time stepping finite element procedure. It takes into account the dimensions of the goaf, the mine layout in the vicinity of the goaf and the mechanics of roof collapse. Account is also taken of the inertia and viscosity of the air which is expelled from the goaf. Pressure changes are taken to be adiabatic and ideal gas conditions. The falling roof is modelled as a “leaky piston” and the user of the software has to specify a roof fall mass per unit plan area, and a parameter defining the extent to which air leaks through the roof as it falls. The values of the leakage parameter (which depends upon the degree to which the roof breaks up as it falls) will have to be determined by the correlation of numerical prediction with field data.
In mathematical terms, the problem amounts to the solution of the transient Navier- Strokes equations for compressible flow, in conjunction with the conservation of mass condition and Newton’s second law of motion for the falling roof. Once ideal gas and adiabatic conditions are assumed, the unknown functions of position and time to be computed reduce to the air velocity vector, air density, roof height above the floor of the goaf, roof fall velocity and density of air in the void above the falling roof.
The analysis under development is two dimensional in plan assuming that air velocity and density depend upon time and plan position only. Time is divided into a large number of small steps. The basic operation consists of computing what is happening at the end of time step from the situation at the beginning, starting with the first time step 47 at the beginning of which all velocities are zero and pressure is atmospheric. The finite element analysis at each time step takes account of the spatial interdependence of air velocities. This spatial interdependence is the result of inertial and viscous effects.
The peculiar difficulty of the time stepping finite element procedure is that the values of the unknown functions at the end of a time step depend not only upon the known values at the beginning but also upon computed values. An iterative method of computation is adopted. At each iteration, the values of the unknown functions are recomputed from their previous computed values and from known values at the beginning of the time step. For the first iteration, an arbitrary choice of the values of the unknown functions at the end of the time step must be made. For each time step, the iteration is continued until successive changes in the computed values of the unknown functions are negligible and the iteration is then said to have converged.
The program based on the Peaceman-Rachford operator splitting technique, has shown to model a simple geometry adequately (Fowler, 1997). It is envisaged that upon its completion the numerical model will be employed to simulate complex roof falls and help provide insight into the dynamics of interaction between the roof elements and the air during a roof fall.
48 CHAPTER 4
DYNAMICS OF WIND BLASTS IN COAL MINES
4.1 INTRODUCTION
The dynamics of goaf air compression due to collapse of mine roof has been exhaustively treated by analytical modelling, McPherson (1980,1995) and by physical modelling, Fowler (1997). Wind blast dynamics within the mine panels and workings have direct impact on mine personnel and mine structures.
The destructive effects of wind blasts in mines have been reviewed in Chapter 2 of this thesis. While damage has generally been confined to the extent of panel dimensions, damages as far as the main fan housing at the surface, have also been reported. The damage has been associated with the air pressure pulse (pressure wave) and also the drag forces due to the high air velocities associated with wind blasts. Measured velocities in longwall panel roadways in the near field i.e. before they have been much attenuated with distance, reach “hurricane level” but do not exceed the sonic levels i.e. Mach values <1. (Fowler & Torabi, 1997; Fowler & Sharma, 2000). However, the celerity or rate at which the wind blast event is propagated through the mine is equal to the speed of sound. (Fowler & Torabi, 1997)
In the case of explosions in air, the shock wave has a steep shock front followed by a rapidly decreasing pressure (Baker, 1973). The term “air blast” is used for pressure waves in air emanating from explosions where “concussion” is the inaudible part of the blast having a frequency content below 20 Hz. In contrast, gas deflagrations and coal dust explosions, where initial combustion and pressure rises are relatively slow processes, do not normally result in shock formation (Baker & Tang, 1991). In controlling the maximum pressure for coal dust explosions and gas deflagrations Scaled vent area, Av/v2/3, where Av is vent area and v is enclosure volume is an important parameter (Bartknecht, 1981). Also, the duration of depressurisation is
49 vent area and reservoir volume for isentropic flow through a nozzle from a pressurised reservoir (Saad, 1985).
In this chapter, fundamentals of compressible fluid flow are reviewed briefly. The dynamics of wind blasts, the analytical models and the analogies with compressible air flow and blast waves occasioned by bursting pressure vessels and explosions are also presented.
The formulations of analytical and empirical models of compressible airflows and pressure pulses associated with wind blasts can lead to a better appreciation of the associated dangers and their zone of influence. This would assist in interpreting the results obtained from the wind blast monitoring at Moonee Colliery and presented as a case study in Chapter 5.
4.2 FLUID DYNAMICS OF COMPRESSIBLE AIRFLOW
As density p in a compressible fluid is not constant, an additional parameter is introduced compared to incompressible flow. Consequently, additional laws/relations (equations of flow) are required to solve the problem. The Eulers equations in vector notation are as follows (Schreier, 1982):
The continuity equation is:
^ + v(/?v) = 0 (4.1) dt
where, V= V(x,y,z,t) is the velocity of the compressible fluid.
The momentum equation is:
(4.2) where, pg is the body force. 50 The conservation of energy equation is:
(4.3) -V-* VT = p — + V(/rV) Dt V ' where, k is the thermal conductivity and e the energy per unit mass of the system. In terms of specific enthalpy h, the equation takes the form:
Dh Dp p = —- + q (4.4) Dt Dt where, q is the heat added or subtracted per unit volume per unit time.
The Eulers equations stated above neglect viscosity. The Navier-Stokes equations consider the impact of viscosity and are stated as follows (Saad, 1985):
PJh =~VP " y V^(V V)]+ 1V' + V * 1^V * V1 (4.5) where, p is the first coefficient of viscosity.
4.2.1 Compressibility factor
Under high pressures and corresponding high densities, the behaviour of a real gas deviates from that of an ideal gas that obeys the perfect gas law. The extent of deviation is indicated by means of a compressibility factor Z, which is defined as (Saad, 1985):
/?v (4.6) RT where, p = absolute pressure v = specific volume R= specific gas constant of a particular gas T= absolute temperature
As the pressure is reduced, the compressibility factor approaches unity. The properties of the gas may also change drastically when it flows at high speed. At low Mach
51 numbers it can be treated as though it were incompressible fluid with negligible error. Stagnation properties (when the fluid is decelerated to zero velocity in a steady adiabatic process), provide a convenient reference state in studying properties of a flowing stream. The stagnation i.e. total pressure is same at all points in the flow, provided the flow is isentropic. Stagnation pressure (p0) is a function of Mach number (M), specific heat ratio (y) and static pressure (p) as expressed by:
r-1 Po (4.7)
The binomial expansion of the right hand side of the equation and rearranging in form of a compressibility factor gives:
P& ~P kT 2-r = 1 + — + ~— M4 + (2-r)(3-2rV + 0.5PV2 4 24 192 (4.8)
where, 0.5pV2 is the dynamic pressure. The right hand side of the equation 4.8 indicates the correction factor for compressibility. Table 4.1 gives the dynamic pressure compressibility factor at various air speeds. At low velocities the compressibility factor is unity.
Table 4.1 Dynamic pressure compressibility factor at various air speeds (after Saad,1985) (p= 1 atm, p= 1.2042 kg/m3, sonic speed 344.14 m/s)
Mach Velocity Compressibility number m/s factor 0.1 34.41 1.0025 0.2 68.83 1.0100 0.3 103.24 1.0227 0.4 137.66 1.0406 0.5 172.07 1.0641 0.6 206.49 1.0933 0.7 240.90 1.1286 0.8 275.32 1.1704 0.9 309.73 1.2192 1.0 344.14 1.2756
52 If a system is insulated so that there is no heat transfer, it is called adiabatic and the processes associated with it are adiabatic processes. A process will be classified as isentropic when the entropy (measure of the probability of a system to be in a particular microscopic state) remains constant. In a reversible adiabatic process, the entropy change of a fixed mass system is due to heat interactions only and the process is isentropic (Saad, 1985). For flow in a duct of varying cross- sectional area, if the flow is adiabatic and isentropic, the static properties of the flow change owing to variations in the cross-sectional area but the stagnation properties remain unchanged.
4.2.2 Flow through a convergent nozzle
If a large reservoir is pumped upto a certain pressure p0 while the temperature in the reservoir is kept constant and a converging nozzle is attached to the reservoir, it will be seen that as the pressure p0 is raised the exhaust velocity increases until it reaches a certain value and then can no longer be increased.
For one dimensional flow of a compressible fluid in a duct of varying cross-sectional area, the velocity (V) and area (A) are related as follows (Saad, 1985):
dV _ 1 dA (4.9) V ~ M2 -1 A where, M is the Mach value.
This equation illustrates that change of area produces opposite effects on subsonic and supersonic flows. In the case of isentropic flow through a convergent nozzle, the limiting velocity is the local velocity of the speed of sound. However, in a converging- diverging nozzle it is possible to attain velocities exceeding the speed of sound.
4.2.3 Shock waves
A pressure difference exists across a compression pulse and when this pressure difference is sufficiently large, there is also an increase in entropy across the pulse (Saad, 1985). In that case, the flow is no longer isentropic and the wave is a called a compression shock wave. The latter is an abrupt disturbance that causes discontinuous 53 and irreversible changes in the fluid properties such as pressure, temperature, density and speed, which changes from supersonic to subsonic. In a normal or one-dimensional shock, the change of properties occur in the same direction as that of the flow.
In an explosion, the shock front of the blast wave generated represents an inverse of the situation of a normal shock in a moving stream. Instead of a stream moving into a shock plane, it is the shock that moves into the medium (Kinney, 1962). However if the shock plane is taken as the frame of reference, it is again the medium which enters the shock. This mathematical treatment of referring to a moving system of co-ordinates reduces the study of transient explosive shock to a simpler problem in steady flow. Figure 4.1 and 4.2 illustrate the normal shock in steady flow and explosive shock front respectively.
For a stationary shock front with subscripts x and y referring to conditions just upstream and downstream of the shock it can been shown (Saad, 1985) that pressure ratio across the shock is:
P y (4.10) P X
4.2.4 Rankine - Hugoniot relations
According to Saad (1985) the continuity, momentum and energy equations relating to normal shock wave are:
P,K=PyVy (4 11)
P,~P,= PyV\~PxV\ (412)
= = (4.13) 2 y-\px 2 y-\py y-1 where, T0 is stagnation temperature, subscripts x and y referring to conditions just upstream and downstream of the shock.
54 Velocity diagram Pressure diagram
Figure 4.1 Normal shock in steady flow (after Kinney, 1962)
Medium after shock at py Ty £ Cx a, Up_ Medium before =u,-uu shock of Px. Tx
velocity diagram pressure diagram Direction of shock movemenf
Figure 4.2 Explosive shock front (after Kinney, 1962)
55 By combining the continuity equation and the momentum equation the velocity downstream of a shock can be eliminated:
Py ~ PX = PXV ^ - PyV \ = pV 1 - (4.14) y y
The velocity upstream of the shock can then be expressed in terms of pressures and densities and similarly for velocity downstream.
1/2 (p,-p,)p, (4.15) (Py-P,)Py
Substituting for Vx and Vy into the energy equation and also, since the enthalpy (sum of the internal energy u, and energy associated with flow work pv, i.e. the product of pressure and volume) of perfect gas can be expressed as:
h — c T = -l—E- (4.16) r-1 p where, cp is constant pressure specific heat, T is absolute temperature.
The terms in energy equation can be rearranged so that pressure ratio across the shock can be expressed in terms of the density ratio called the Rankine -Hugoniot relation:
r +1 Py , Pj_ = Izl p* (4.17) p* £±! _ r-1 Px Figure 4.3 illustrates the Rankine -Hugoniot curve for various density ratios. Also shown is the relationship between pressure and density for isentropic process according to the relation:
56 Figure 4.3 Rankine - Hugoniot curve for various density ratios (after Saad 1985)
rn/s
5000 2000 1000 Shock I root 500
Oelonal ion gas
Particle velocity
0.1 0.2 R/W
Figure 4.4 Shock front velocity, particle velocity and gas front velocity as function of scaled distance from TNT charges in air (adapted from Persson, Holmberg & Lee 1994, after Weibull 1947, & Granstro, 1956)
57 p2_ = (4.18) P\ KPiJ
At low pressure ratios, the Rankine-Hugoniot curve and the isentropic curve differ only slightly from each other, so that a weak shock appears like an isentropic process. At large pressure ratios which characterise strong shocks, the density ratio reaches the limiting value (y+1 / y-1) or 6.0 for y =1.4.
For the shock front represented in figure 4.1 moving with velocity Vx into an atmosphere where the acoustic (sound) speed is Cx, the pressure jump associated with the shock is from atmospheric pressure px to shock-generated pressure py. This corresponds, in the steady-flow counterpart as represented in figure 4.1, to a pressure jump generated in normal shock by a stream with Mach number Mx = VJCX.
From the Rankine-Hugoniot relationships, it can be shown that the pressure increment, or the overpressure of the explosive shock front, is given by (Kinney, 1963):
, = „ -P =p 2r(y,:-lJ = 7(A/,2-lj (4.19) P Py Px Px r + 1 6 Px (4.20) P =
Also the particle velocity or ‘blast wind’ is in the same direction as the shock front and given by the (Kinney, 1963):
y, = M,C,-MyCy = Cx\rfx-My{Ty!Tx)n\ (421>
The particle velocity ratio V/Cx can also be expressed in terms of Mx, the Mach number of the explosive front or in terms of the pressure ratio p/px for air (Kinney, 1963):
v, 2(M/-l) 5(M/-l) (4.22) cx (/ + iR 6mx rK_\2 = {}lApyl_P,-l)2 = 25jpy/px-l)2 (4.23) SJ (r + iipyiPxhir-1)
58 Equations 4.19 and 4.20 indicate that any pressure pulse with finite overpressure is necessarily at supersonic velocity. As a shock wave is propagated through air, as in the case of air blast, far away from its source, it weakens and the shock velocity decreases with increasing distance from the source so that, at great distances it is essentially propagating at the speed of sound. The acoustical laws would then apply and the pressure in the front would decrease as the inverse of the distance. The entire time history would approach essentially a constant form, changing only in amplitude as it advances. The acoustic approximation or the acoustic asymptotes for air blasts are further discussed in section 4.1.5 of this chapter. The transmission of shock waves or blast waves in air is inherently a non-linear process, whereas acoustic wave propagation can be adequately handled by linear theory (Baker, 1973). Acoustic waves involve only infinitesimal pressure changes, produce no finite changes in particle velocity, move at sonic velocity and do not “shock up”.
The general equations for air blast transmissions in three-dimensional cases are difficult to solve, both analytically or numerically, Most of the solutions are limited to one dimensional cases of: linear (plane wave) applicable to shock tubes, spherically symmetric flow applicable to blast waves in air, and cylindrically symmetric flow. The analytical solutions for strong blast wave shocks have been given by many researchers including Taylor (1950), Brode (1955) and approximate analytical method for weak shock regime by Bach and Lee (1970). A large number of versions of ‘hydrocodes’ are available to numerically solve the complex differential equations describing the generation of blast waves and their transmission through various media. They give reasonably accurate predictions of blast wave properties for high explosives especially for symmetric explosion sources. Moreover, their strength lies in their ability to predict properties in regimes where measurements are difficult or impossible or for which, no suitable measurement techniques exist (Baker & Tang, 1991).
4.2.5 Adiabatic friction flow
Adiabatic conditions can prevail if the flow is so rapid that heat interaction is negligible
The entropy of the system can still increase because of friction. Friction is associated 59 with the turbulence and viscous shear of molecules of the gas and also the movement of the molecules near the walls of the conduit. Irreversibility associated with friction causes a decrease in the stagnation pressure and this in turn affects the flow characteristics. The friction factor f, is a function of the Reynolds number Re, of the flow and the roughness of the boundaries s, and can be obtained from the Moody diagram, which gives the relationship among friction factor, the Reynolds number and relative roughness e/D, where D is the conduit diameter (Figure 4.5). At large values of Reynolds number and for a fixed value of relative roughness, the friction factor becomes independent of the Reynolds number. A gas flowing at subsonic velocity tends to expand as result of friction and accelerate to sonic velocity. Conversely, if the flow is supersonic, the gas will be decelerated and approach Mach 1.
4.2.6 Pressure pulses due to surface blasts
A blast wave emitted from a bench blast or unconfined blasts in air will travel through the air for long distances, with an initial supersonic shock velocity, which quickly decreases to the speed of sound. Figure 4.4 illustrates the shock front velocity and particle velocity, as a function of scaled distance. The cube root scaling law states that self similar blast waves are produced at identical scaled distances when two different sized explosive charges of similar geometry and of the same explosive are detonated in the same atmosphere (Hopkinson, 1915; Sach, 1944). The scaled distance R is given by the equation:
n= R (4.24) fVl/3 where, R= distance from a explosive charge (m) W= weight of charge (kg)
An idealised profile of a blast wave from a condensed high explosive is illustrated in Figure 4.6. For a spherical wave front, the overpressure will decay as the inverse of the distance squared. The duration of the blast wave increases with distance from the center of explosion, and reaches a limiting value (and also vanishes) as the shock front
60
fluids
of
flow
boo
1985)
go frictional
—
4f Saad, for
go
61 from Factor,
Roughness,
Diagram
Friction (adapted Relative Moody
4.5
Figure
olds Number, degenerates into a sound wave (Kinney, 1962). At great distances, the shock waves from even very large energy blast sources would rapidly decrease to sound level values and the acoustical law for an inverse decrease in pressure with increase in distance from blast source could be applied. However, wind velocity and temperature inversions do cause unexpected blast overpressures due to focusing effects.
For small pressures, the front pressure (overpressure) for an unconfined charge is given by (Eriksson, 1987):
^>/3 (4.25) p - 0.7------R where, p is in bars, W in kg and R in m.
Pressure time histories have been described by various researchers assuming simple linear decay (Flynn, 1950), and complex formulations with multiple parameters based on exponential decay (Ethridge, 1965; Brode, 1956). The parameters associated with the shock front include shock velocity, peak values of incident and reflected overpressures, dynamic pressure, density and temperature. The parameters are usually presented in dimensionless form and Sachs-scaled parameters used. The basic front parameter is scaled peak overpressure Ps. All other front parameters for the incident wave can be determined once it is defined, by using the Rankine-Hugoniot conditions and a simple integration scheme (Baker, 1973). For weak shocks, or large scaled distances, the acoustic approximations become asymtotic and can be expressed as (Baker, 1973):
_ , Ps , 0.0824 (4.26) p,=\+—=i+—=— Y R — y +1 — 0 0494 U =1 + ^-P, =1+ - (4.27) 4y R ff'=1+rzlP' = l+2m (4.28) s / R
p _ 01153 (4.29) R Ps _ 0,0824 u' = (4.30) y R 62 where, the parameters are as in Table 4.2 and Energy E is for spherical charges and the subscript o stands for ambient conditions.
The mechanism of air blasts from the detonation of explosive charges according to Siskind et al (1980) are determined by the following four important causes:
1. The air pressure pulse; 2. The rock pressure pulse; 3. The gas release pulse; 4. The stemming release pulse.
The air pressure pulse is caused the direct rock displacement at the rock face and in the cratering zone; the rock pressure pulse is seismically induced and usually of such low amplitude that it can be neglected; the gas release pulse is caused by the detonation products expanding through vents in fractured rock and the stemming release pulse is caused by gas escaping from the blown out stemming. The audible component of air blast or the noise is mostly due to last two mechanisms. Figure 4.7 illustrates the relationship between overpressure and scaled distance for air blasts.
Table 4.2 Sachs-Scaled Nondimensional Blast Parameters (after Baker, 1973) Parameter Symbol Equation
Peak overpressure Ps Ps/Po
Overpressure P's Ps/Po
Density ~p p/p 0
Temperature u 0/0 0 Peak particle velocity Ts us/aQ Shock velocity U U/ao
Scaled distance R (R p01/3)/E 1/3
63 r
Primary Shock
Overpressore,
PosJclve Phase Specific 1mpu1se. 1
Secondary Shock
Durac1 on .
f Arrival, c
Time
Figure 4.6 Idealised profile of a blast wave from a condensed high explosive (after Baker & Tang, 1991 )
Scaled Distance R/w'/3 ft lbs1/3
Structural damage
Most windows breaks
Some v.mdows breaks
Scaled Distance R/W
Figure 4.7 Nomogram for air blast overpressure as a function of scaled distance (adapted from Persson, Holmberg & Lee, 1994 )
64 4.2.7 Air pressure pulses due to underground blasts
Research on pressure pulses produced by blasts in underground mine workings is limited in contrast to studies on the effects of blast waves generated by explosives detonated above ground. An unconfined explosive charge when exploded in free space causes the resultant blast wave to move spherically outward in all directions. The overpressure generated would obey the cube- root scaling law. In contrast, when an unconfined charge is detonated in an underground opening, the blast wave is reflected from the roof, rib and floor. It maybe attenuated in certain areas and amplified in others. Hanna and Zabetakis (1968) conducted large scale unconfined blasts in an underground mine, measuring overpressure-time histories. The pressure pulses were of longer duration compared to those obtained from unconfined surface blasts. Wall reflections affected the magnitude of the pulses appreciably with the magnitude of pulses from reflected shocks nearly twice that of pulses from shocks transmitted directly through the tunnel. Table 4.3 gives the initial pressure pulses recorded at various distances from the explosive charges. The cube root scaling law for peak overpressures obtained in free air is valid for underground shots only in the near field (within about 1 tunnel diameter distance). At greater distances, the pressures obtained along the direct paths from the charges deviated from scaling law and were appreciably higher.
Table 4.3 Initial air pressure pulses produced by explosive charges (after Hanna and Zabetakis, 1968)
Explosive Weight of Pressure Charge (psig) (Lb.) 50 ft 100 ft 150 ft 200 ft
TNT 1020 - 22.8 18.0 15.0 Modified Amatol 2000 28.5 14.6 15.0 10.6 TNT 240 22.5 14.0 7.4 7.7 TNT 60 6.7 2.9 2.7 1.9
In contrast, gas deflagrations and coal dust explosions in enclosures, where initial combustion and pressure rise are relatively slow processes, do not normally result in shock formation (Baker & Tang, 1991). Scaled vent area, Av/v273, where Av is vent area 65 and v is enclosure volume is an important parameter in controlling the maximum pressure for coal dust explosions and gas deflagrations (Bartknecht, 1981).
4.3 WIND BLAST DYNAMICS IN UNDERGROUND MINE WORKINGS
A wind blast in mine workings may exhibit three distinct phases as determined by physical modelling (Fowler & Torabi, 1997):
1. A primary phase characterised by an overpressure and a high velocity flow of air outward from the fall, both of which exhibits a peak and corresponds to the period during which the caving roof element is accelerating;
2. A secondary phase characterised by a residual airflow of lesser velocity than in the primary phase and corresponding to the period during which the roof element is falling at its terminal velocity;
3. A tertiary phase characterised by an underpressure and a flow of air towards the fall (i.e. suck back).
Evidence of the secondary phase has yet to be observed in the mine situation. It may be that the conditions of a real roof fall in a coal mine are such that the primary phase is always prematurely terminated by the falling roof element hitting the floor with the result that the secondary phase (corresponding to the roof element achieving its terminal velocity) does not develop.
4.3.1 Characteristics of wind blast overpressure and velocity time histories
Typical overpressure and wind velocity time histories for wind blasts in gate roads of longwall mine workings are shown in figures 4.8 and 4.9 respectively along with the terminology used to describe some of the characteristics (Fowler & Sharma, 2000). For overpressure time history, it is observed that:
1. During the compression phase, the absolute pressure is greater than the ambient
pressure;
66 Peak Overpressure
Maximum Rale ol Rise Of Overpressure (tangent to curve)
Maximum Rate of Fall ol Overpressure (tangent to curve)
Overpressure Hall Width '
Impulse (area under positive phase ol curve)
Negative (Suction or Rarelaction) Phase
Ambient
Overpressure Overpressure Rise Time Fall Time Positive (Compression) Phase
Peak Underpressure
Elapsed Time (seconds)
Figure 4.8 Typical overpressure time histories for wind blasts in longwall panel.
Peak Wind Blast Velocity
Maximum Rate ol Rise ol Wind Blast Velocity (tangent to curve) Maximum Rate ol Fall ol Wind Blast Velocity (tangent to curve)
Wind Blast >- Velocity Hall Width
Maximum Excursion
(area under positive phase of curve) Negative Phase (Suck Back)
Velocity Rise Time Wind Blast Velocity Fall Time
Peak Suck Back Velocity Positive (Wind Blast) Phase
Elapsed Time (seconds)
Figure 4.9 Wind velocity time histories for wind blasts in longwall panel.
67 2. The peak overpressure represents the difference between the maximum value of the absolute pressure and the ambient atmospheric pressure;
3. The overpressure rise time is the duration between the time at which the overpressure begins to rise and the time that the peak overpressure is attained;
4. The overpressure fall time is the duration between the time that the peak overpressure is attained and the time at which the overpressure becomes negative;
5. The impulse is the integral of the overpressure time history over the duration of the compression phase and is represented by the area under the positive phase of the overpressure time history;
6. The overpressure half width is the time for which the overpressure is equal to or greater than half the peak overpressure;
7. The maximum rate of rise of overpressure is the slope of the tangent to the rising segment of the overpressure time history at the point of inflection;
8. The maximum rate of fall of overpressure is the slope of the tangent to the falling segment of the overpressure time history at the point of inflection;
9. During the rarefaction (negative or suction) phase, the absolute pressure is less than the ambient atmospheric pressure;
10. The peak underpressure is the difference between the minimum value of the absolute pressure and the ambient atmospheric pressure.
For a steady state of flow, the changes in the velocity and in the dynamic pressure accompany the changes in overpressure. The dynamic pressure is proportional to the square of the velocity and to the density of the air. For wind velocity time history, as illustrated in figure 4.9, it is observed that:
1. During the positive (wind blast) phase , the wind blows away from the goaf fall;
68 2. The peak wind blast velocity corresponds to almost the same time as the peak overpressure;
3. The wind velocity rise time is the duration between time at which wind first begins to blow and the time that it attains its peak velocity;
4. The wind velocity fall time is the duration between the time that the peak wind blast velocity is attained and the time at which the “suck back” begins;
5. The maximum excursion is the integral of the wind velocity time history over the duration of the positive phase. It is the area under the curve representing positive phase of the wind blast time history. Physically it represents the maximum displacement of a molecule of air away from the goaf fall during a wind blast;
6. The wind blast velocity half width is the time for which the wind blast velocity is equal to or greater than half the peak wind blast velocity;
7. The maximum rate of rise of wind blast velocity is the slope of the tangent to the rising segment of the wind blast time history at the point of inflection;
8. The maximum rate of fall of wind blast velocity is the slope of tangent to the falling segment of the wind blast time history at the point of inflection;
9. During the negative (suck back) phase, which immediately follows the positive phase, the wind blows towards the goaf and attains a peak suck back velocity;
10. The negative phase is not always observed in the mine workings, during a wind blast.
4.4 ANALYTICAL MODELS OF WIND BLAST
McPherson (1980, 1995a & 1995b) has done analytical modelling of large roof falls with adiabatic compression of the goaf gases. The modelling was directed to examine high pressures and temperatures that could result from such goaf falls leading to methane or
69 dust explosions. The two analytical models namely the bounded model and air leakage model, proposed by McPherson are as follows.
4.4.1 Bounded model
In this model, the air is completely contained under the falling mass with no leakage as shown in figure 4.10. The movement of the block follows Newton’s second law of motion assuming no frictional effects. The acceleration of the falling block in an undamped and sealed system can be determined as:
du (431) ~dt where, pr = density of the roof material, (kg/m3), P0 = air pressure over the falling block, (Pa), u = velocity of falling block, (m/s), hr = thickness of the falling block, (m), C = constant, g = acceleration due to gravity, (assumed to be 9.81 m/s2), z= distance between the falling block and the floor, (m).
The system described by above equation would in reality be damped due to shear resistance at the sides of the falling block and the due to the turbulence generated within the compressed air. McPherson (1995a) considered a damping factor given by the following equation: DF = 1.80 + 5 x 10'5 A,+ 2/z„ (4.32) where, As is the surface area of the sides of the falling block, (m2).
To consider damping, the equation (4.31) was modified.
(4.33) du (c ^ g-\ —~Po T------(±)DF u ~dt U J Kp,
where ± means that the system damping will oppose the motion of the falling block.
70 Figure 4.10. The Bounded model for wind blast in mines, (after McPherson 1995)
P rlma r y Shoe k __ Secondary Shock
Time of Arrival
Figure 4.11 Idealized profile for a blast wave from a pressurized vessel burst (after Baker & Tang 1991)
71 The temperature rise will be given by the adiabatic process equation
y-i f_p' (4.34) T = t UJ where T = Absolute temperature under the falling block (K).
4.4.2 Air leakage model
The Bounded model did not take into account the dissipation of air through the leakage paths connected to the goaf McPherson (1995b) extended the above analytical model for air leakage. It was assumed that rate of leakage is affected by the resistance to air movement within the goaf under the falling roof, airway resistances including shock losses at the entries from the collapse zone to the airways. The rate at which the air mass will be displaced will depend on the pressure difference, and the total resistance of the leakage paths. Based on the square law (McPherson, 1993) the equation can be stated as:
—10.5 dm* (P~Pq) (4.35) Pa dt P, where, pa = air density, (kg/m3), ma= air mass, (kg), Rt= total resistance, (m'4).
The loss of air would lead to a corresponding loss of pressure following general gas laws:
(4.36) dP = dm a —^ where, R = gas constant, (287.04 J/Kg °C for air), v= volume of space beneath the falling block, (m3). Numerical solution of the above equations (McPherson 1995b) using a time stepping procedure revealed that thicker the falling block, the higher are the peak pressure values.
72 In the goaf, the pressure reaches a peak and then converges to value corresponding to the weight of the falling block (prghr). However the temperature, after an initial slow rate of rise, escalates rapidly towards the end of the descent of the block (towards the last 0.1 metre). This is due to more work being done against a decreasing mass of air under the descending roof Another interesting outcome of the simulation was that initial peak pressures and rate of temperature rise tend to increase with decreasing roof dimensions. The explanation lies in the fact that air escapes more readily from beneath the falling roof in smaller plan areas resulting in a higher rate of descent. However, the simulation was unstable for smaller plan areas.
For thicker roof elements (>3.5 metres), the void above the block will have negligible effect and the rate of fall and the corresponding rate of air displacement from the goaf into the surrounding workings will be modulated principally by the pressure of the air below the falling block and the resistance of the airways connected to the goaf In contrast, the rate of fall of a thin element is governed principally by the differences in pressure above and below the block implying that air will be expelled from the goaf at near constant mass flow rate, independent of the layout (Fowler & Torabi, 1997).
4.4.3 Bursting reservoir model
The characteristics of air blast waves developed by pressure vessel bursts and those developed by detonation of condensed-phase high explosives such as TNT show some basic differences. An idealised profile of a wave from a TNT explosion is shown in Figure 4.6. The rise to peak shock pressure Ps is essentially instantaneous at shock front arrival, and pressure decay is quite rapid, falling below atmospheric pressure due to inertial effects and finally recovering to ambient level more slowly. A secondary shock with smaller amplitude is often evident. The secondary shock is caused by a rarefaction wave which propagates back into the explosion products as the primary shock is transmitted to surrounding air. On reaching the center, the rarefaction reflects as the secondary shock. The primary wave is usually characterised by its amplitude Ps, positive phase duration td and positive phase specific impulse is.
73 Blast waves from pressurised gas vessel bursts show similarity to those from condensed explosive charges, but their lower energy densities and larger sources required for total energies also change the wave characteristics. Figure 4.11 shows the idealised profile for a blast wave from a pressurised vessel burst. The negative phase is much more pronounced than for TNT explosions and the second shock is much stronger, compared to the first shock, because the rarefaction wave is more pronounced (Baker & Tang, 1991).
The total source energy E can be calculated by assuming isentropic expansion of the fluid from its initial state to atmospheric pressure and calculating the change in internal energy u, and then multiplying by the mass of gas. In equation form, then the internal energy is:
2 U\~U2 = \pdv (4.37) 1
E = m (u\ - u2) (4 38) where, p is the absolute pressure , v is specific volume , and m is mass of fluid.
For a perfect gas, the explosion source energy based on isentropic expansion, is given by Widermann (1986) (4.39) P\v\ 1 - f p° ] O' -1) < P\ >
where, pi and Vi are initial absolute pressure and vessel volume, p0 is the ambient pressure. For real gases, the isentropic expansion energy is given by
E = (l-/3)E, (440)
where, p = b p0 and b is a covolume parameter measured experimentally or fitted to an equation of state. 74 Once explosion source energy is calculated, the TNT equivalency can be determined and standard TNT blast curves could be used to predict blast wave properties. However for close-in effects, TNT equivalency is not appropriate (Baker & Tang, 1991).
4.4.3.1 Scaling laws for blast waves from bursting pressure systems
The scaling laws for blast waves from bursting pressure systems was developed by Esparza and Baker (1977) and can be stated as:
"A p=(p'p J
RPo1'3 PL h={‘, aoP:n/E',}) =f >r»— > E1'3 ’p0 ao Po'/l IE'n) (4.41)
L = (', ao p 212 IE'12) j where, the parameters are defined in Table 4.4
Each dimensionless parameter on the left of the equation is a scaled blast wave characteristic. This way of presenting scaling law indicates that each scaled blast wave characteristic is some different (and unknown) function fk of the five dimensionless ratios within brackets on the right. Some of the dimensional parameters are identical to those developed by Sach (1944), for scaling the effects of varying ambient air conditions on blast waves from high explosive sources.
A wind blast in mine workings, is a pressure pulse and not a shock wave like that emanating from an explosion of a high explosive or as observed in shock tubes or sudden failure of pressurised vessels. The rate of rise of overpressure is quite low compared to that in explosions. There is no evidence that as the wind blast event propagates through the workings, shock wave conditions develop, though the possibility of such an eventuality cannot be excluded. Also the possibility of shock wave conditions
75 developing in the goaf do exist due to reinforcement and resultant steepening of compression waves in air generated by the falling roof element (piston). However, the celerity or rate at which the wind blast event is propagated in the far field, through the mine, outbye of the source (goaf) is equal to the speed of sound.
Table 4.4
Physical parameters and their dimensions for vessel blast scaling law
(after Baker et al, 1983) Symbol Dimensions Description
P F/L2 Blast wave overpressure ta T Time of blast wave arrival
td T Duration of blast wave
Is FT/L2 Positive phase specific impulse
Pi F/L2 Vessel failure pressure
Po F/L2 Air ambient pressure
ai LA' Sound speed within vessel
ao LA Sound speed in ambient air
Yi - Ratio of specific heats in vessel R L Distance from centre of blast source
Vi - Length Ratios E FL Total source energy
4.4.4 Pressurised reservoir model
In shortwall caving mechanics discussed in Chapter 2, the roof in such narrow longwall panels tends to cave behind the face line occasionally exhibiting feather-edging, characterised by low angle roof fracturing that forms an irregular breaker line and “thin” slabs of massive roof detaching. The geometry of the connecting roadway with the caved goaf could be approximated to that of a roadway with a converging nozzle connected to a pressurised chamber. While the peak overpressures generated would be functions of the goaf and roof element dimensions, rock mass properties and vent areas, the duration of the peak overpressure fall time could be related to the “blowdown” time for a pressurised tank with an initial pressure p; and initial temperature Tj. Assuming
76 isentropic expansion, and flow through convergent nozzle for air (y=1.4) as given by the following equation (Saad, 1985):
(4.42)
where, t = Time, (seconds) v = Volume of reservoir,(m3) A = Area of orifice, (m2) T = Temperature, (K) pf = Final pressure, (kPa) Pi = Initial pressure, (kPa).
Assuming isothermal conditions, blowdown time will be given by the following equation (Saad, 1985):
- 0.086 v /« — (4.43) Pi aJt
Figures 4.12 and 4.13 illustrate the relationship between “blowdown” time and final pressure to initial pressure ratios assuming isentropic and isothermal expansion for a assumed reservoir volume to orifice ratio 1:100 and 1:1000 respectively and initial temperature 293° K. The blowdown time taken is greater in the isothermal than in isoentropic case. In case of wind blast in longwall panels, as the event lasts only a few seconds and the process quite rapid, isentropic expansion is more likely. The figures illustrate that the increase in duration of the event could be related to the formation and subsequent “blowdown” of a pressurised reservoir inbye of the longwall face. For a given orifice area, the dimensions of the reservoir formed inbye affects the duration of the event.
The rate of outflow through the vent is calculated by applying various equations that have been developed for orifice flow. Different equations are necessary however,
77 Blowdown time
■ Isentropic a Isothermal
100 (m)
Time (sec)
Figure 4.12 Relationship between "blowdown" time and final pressure to initial pressure ratios for reservior volume and orifice ratio 100:1
Blowdown time
■ Isentropic a Isothermal
S 0.6 1000 (m)
3 04
Time (sec)
Figure 4.13 Relationship between "blowdown" time and final pressure to initial pressure ratios for reservior volume and orifice ratio 1000:1
78 depending on the pressure differential across the vent (p - pa), where pQ is the external pressure. When the pressure inside the reservoir is less than a critical value given by (Saad, 1985),
(4-43) Pcrlt=Po V+x)
the flow through the vent is subsonic. When the pressure is above pcnt, the flow through the vent is at the sonic velocity and is described as “choked”. The flow velocity cannot exceed the sonic velocity. Then for internal pressure above pcnh the flow through the vent is independent of the external pressure.
79 CHAPTER 5
WIND BLAST STUDY AT MOONEE COLLIERY
5.1 INTRODUCTION
Moonee Colliery is one of the three underground mines owned and operated by Coal Operations Australia Limited. The Colliery is located on the New South Wales coast at Catherine Hill Bay, approximately 20 kilometres south of Newcastle and 100 km north of Sydney as shown in Figure 1.1, within the Lake Macquaire district of the Newcastle Coalfield and currently mines the Great Northern seam. The mine employs the retreat longwall system of mining with a pair of development headings on each flank. The panel widths, between the gateroad centrelines were 95 metres for Longwall Panel no. 1 (LW1) to LW4B. The chain pillar widths were generally 35 metres between riblines other than for the tailgate of LW1, where 55 metres wide pillars were adopted. The depth of overburden varies across the area from 90 metres to 160 metres. LW1 has a cover depth of approximately 160 metres and consequently, the ratio of the width of the longwall block to depth is 0.6:1. Table 5.1 summarises the main performance and operational details for Moonee Colliery.
Several significant wind blasts occurred at Moonee Colliery in January 1998 during the mining of LW1. The first event on 10 January at 12.50 am was caused by the initial goaf fall in the panel which took place after the face had retreated 200 metres. Table 5.2 summarises the extent of damage caused to the elements of the mine including the ventilation structures. It was fortunate that the day being a Sunday, no mine personnel were present in the panel. The subsequent goaf falls 2, 3 and 4 were apparently, much smaller events with the damages being minimal. However the fifth goaf fall in the panel, which occurred at 10.00 am on Thursday 22 January, 1998 gave rise to a wind blast which knocked over and injured six of the 19 crew men who were engaged in regular maintenance work, and blew down stoppings as well as damaged an overcast. The UNSW Mkl Wind Blast Monitoring System was subsequently deployed and
80 the monitoring programme has since been extended to include later longwall panels. The mine introduced hydraulic fracturing of the roof in LW3 panel in an attempt to induce “caving on demand”. However, the wind blast hazard still remained, the strategy of the mine management being to control rather than eliminate the hazard as, the prior development committed it to the longwall layout for panels up to LW4B.
This chapter presents the case study of Moonee Colliery, the outcomes of the wind blast monitoring from LW1 panel to LW 4B panel, over the time period ranging from January 1998 to November 2000. The data acquired from the monitoring are tabulated in Appendix C.
5.2 DETAILS OF MOONEE COLLIERY
5.2.1 Geology and Tectonic Setting
A generalised stratigraphic sequence of the uppermost sub-group of the Newcastle Coal measures, the Moon Island Beach, together with the immediately overlying Munmorah Conglomerate Formation of the Narrabeen Group, are given in Figure 5.1. The constituent components comprise coal, conglomerate, sandstone, shale and rocks of pyroclastic origin, the latter are of variable grain size up to coarse sand and termed “tuff’.
The Moon Island Beach Sub-Group crops out around the eastern, northern and north western shores of Lake Macquaire and includes three major Coal Members (The Fassifern, Great Northern and Wallarah) all of which are currently worked in different mines. They constitute high-volatile thermal coals with medium to high ash and very low sulphur and are the main source of coal for power generation in the Newcastle area. An east-west section through the Moon Island Beach Sub-Group is given in Figure 5.2.
The Great Northern Coal has been extensively mined and is currently the most commercially important seam within the Newcastle Coal measures. It comprises a hard, dull coal high in inherent ash and containing minor laminae of bright coal and occasional mudstone bands up to 50 millimetres thick, particularly near the roof and
81 Table 5.1 Moonee Colliery Longwall Operational Details (adapted from Australia’s Longwalls, March 2000, Joint Coal Board Statistical Review, 1999) Company Coal Operations Australia Ltd State/Coalfield NSW/Newcastle Seam (working thickness) Great Northem(3.1) Depth of cover 90m -160 m Recoverable reserves(raw) 10 Mt Longwall operations (weekly) 3 x 8 hrs shifts, 5 days, 13 unit shifts Other operations (weekly) 3 x 8 hrs shifts, 5 days, 22 unit shifts Raw coal output 1999 Longwall face 1 063 500 t Other 202 600 t Total 1 266 100 t Forecast total raw coal output 2000 1 440 000 t Longwall face raw coal output 98-99 1 009 900 t Total mine raw coal output 98-99 1 240 500 t Commenced longwall mining November, 1997 Longwall block dimensions 1999 Width 90 m Length 1975 m, 1975 m Shearer manufacturer Long -Airdox Type Electra 1000 DERDS Drum Diameter M/G 2.0m, TG 2.0 m Cutting height/method 3.1 m, uni-di Install power, kw 881 kw Roof support manufacturer DBT (Australia) Type, number of supports 2-leg chock shield,53 Yield Load, working range, control 860 t, 1.8 m-4.1 m, PM4 Face Conveyor Manufacturer DBT (Australia) Width/chain size 932 mm PF4/30 mm twin in-board Chain speed and Manufacturer 1.27 m/s kw 1 x 350 kw Beam stage Loader, kw DBT Australia/PF4-932, 125 kw Coal crusher, kw DBT (Australia), SB63UB, 125 kw Coal clearance (to surface) Type/Capacity Conveyor/ 1 600 tph Other face equipment 2 x Joy 12CM30 continuous miners 2 X Joy 15SC shuttle cars 2 x Long-Airdox uni-haulers 4 x Hydraulic 4000 series miners mounted rigs
82 Table 5.2
Physical damage caused by wind blast, Moonee Colliery, Longwall Panel no.l, first goaf fall on 10 January, 1998 (adapted from wind blast notes, Moonee Colliery)
Element Effect
Plaster board stoppings Seven destroyed Brattice ventilation “wing” (maingate) Displaced nearly 300 metres outbye Overcasts Superficial damage Tailgate regulator door Badly damaged Water barriers(first ten rows outbye the face) All tubs blown down Water barriers (greater than ten rows outbye) Occasional damage to tubs Computer keypads (maingate) Ripped from connections. One found 30 metres outbye, other not located Pick block inserts and picks Left on maingate drive. Found scattered to boot end Hydraulic pressure return pressure line Perforated through steel armouring and leaking (Butterfly plate found in vicinity) Plastic print canister from underside the Ripped off DCB Two found 10 metres on the return side of the Maingate lights main gate chock
floor. Its maximum thickness of more than 7 metres occurs towards the north- west with a decrease towards the south-east to less than 3 metres. It is mined by Moonee Colliery, amongst others, and the extracted seam height varies between 2.4 and 3.3 metres. Unfortunately, seam splitting, lensing and quality deterioration limit the extent of workings of the Great Northern Coal.
Overlying the Great Northern Coal seam is the Catharine Hill Bay Formation. This ranges in thickness from practically zero where the Great Northern Coal seam and Wallarah Coal seam merge, to over 60 metres. The major constituents members are the
83 Cowper Tuff Member j~ Wa I I or oh Coal Mannermg Pork Tuff Member Toukley Coal Member Buff Point Cool Member • Catherine Hill Bay F Teralba Conglomerate Member Moon I si an dl at i on Booragul Tuff Member Beach Great Northern Coal Awabo Tuff Member Sub Group Chain Volley Coal Member Eleebana Formation Bolton Point Conglomerate Member rFassifern Cool
Figure 5.1 Generalised stratigraphic sequence, Moon Island Beach Sub-Group and overlying Munmorah Conglomerate formation, Newcastle Coal Measures (adapted from Standing Committee on Coalfield Geology of New South Wales,1974)
EE1 10
Figure 5.2. Section through the Moon Island Beach Sub-Group passing near to Moonee Colliery (from Diessel. 1992;after Booking, Howes & Weber. 1988)
84 Booragul Tuff, The Teralba Conglomerate and Mannering Park Tuff. Overlying the Booragul Tuff Members over a limited area in the south of the coalfield are the Buff Point Coal and Toukley Coal. These coal members are separated from each other and from the Mannering Park Tuff by unnamed conglomerates similar in composition to the Teralba Conglomerate Member.
Over parts of the Newcastle Coalfield, the Booragul Tuff Members form the immediate roof of the Great Northern Coal seam, the so-called claystone roof. It comprises tuffaceous claystone, siltstone and sandstone, up to 8 metres in thickness, which have very low strength and deformation properties, and susceptible to alteration in the presence of water and hence form an unstable roof. Claystone roof is particularly prevalent in the mines south of Lake Macquaire where large reserves of Great Northern Coal seam are currently sterilised.
It will be seen from Figure 5.2 that, to the West, the Booragul Tuff ‘separates’ from the roof of the Great Northern Coal seam and rises to join the floor of the overlying coal member. Consequently, in this area the Booragul Tuff overlies the Teralba Conglomerate Member, which then forms the roof of the great Northern Coal seam. To the east, the Teralba Conglomerate Member (The ‘Marks Point Conglomerate’ as called by Bocking, Howes & Weber, 1988) overlies the Great Northern Coal seam in areas where the Booragul Tuff is absent owing to penecontemporaneous erosion. The transition in the colliery working between claystone roof and conglomerate roof is often quite rapid, occurring over a few metres.
The Teralba Conglomerate Member is up to 60 metres in thickness and is composed of conglomerate, sandstone and some mudstone. The conglomerate contains a variety of pebble types but chert and quartzite predominate. The pebbles are rounded to sub rounded and are generally less than 25 millimetres in diameter but range up to 150 millimetres. Generally the matrix is argillaceous and it is reported that bedding ranges from massive to well-bedded.
Variations in the lithology of the Teralba Conglomerate Member are evident from borehole logs and from the observations of the roof of workings in the great Northern Coal seam. Anderson (1991) reports that changes in the lithology of the immediate
85 roof from Conglomerate to sandstone or mudstone can be quite abrupt, occupying only a few metres. He also puts forward the view (p.l) that, “in general terms, the conglomerate at shallow cover (30-100m) is not as competent as that at greater depths (>150m)” and that “as a general rule conglomerates on the eastern and southern boundaries of Lake Macquaire tend to be more competent than those on the western boundary of the lake”.
A typical working section for development of the Great Northern seam at Moonee Colliery, together with the overlying stratigraphic sequence at Borehole JWA 30, is given in Figure 5.3. In order to leave sufficient roof coal to protect and support the overlying claystone, the standard working height in development is restricted to between 2.7 and 2.8 metres. Coal quality and safety considerations have resulted in variations in cutting height for the longwall panels with topcoal of 300 mm being left to the underside of the claystone, resulting in standard extraction height of up to 3.2 metres. The main roof comprises the Teralba Conglomerate member, which was of the order of 35 metres in thickness over LW1 to LW4. The Teralba Conglomerate thins in a south-easterly direction i.e. perpendicular to the long axis of the longwall blocks. Table 5.3 summarises the geo-technical properties of the Moonee Colliery lithological units.
5.2.2 Insitu Stress field at Moonee Colliery
Results of an underground stress measurement investigation by Enever and Walton (1988) at 90 meters depth are given in Table 6.1 in chapter 6. The major principal stress was determined to be 11.7 MPa, oriented to the north-northeast direction. It would be slightly anticlockwise of the axis of the present longwall panels. The minor horizontal stress was determined to be 8.3 MPa, acting in a perpendicular direction to the major principal stress. It would be more or less across the present longwall panels. The vertical stress was measured to be 2.4 MPa. The vertical stress calculated at the present depth of cover over the longwall panels would be around 4 MPa.
86 Colliery
Moonee
seam,
87 Table 5.3
Summary of Geotechnical Test Results, CWA 42 and 43, Moonee Colliery (after Edwards, 2000)
Test Conglomerate Sandstone Mudstone, Claystone Shale and Siltstone
Uniaxial Compressive Strength
Mean (MPa) 55.3 107.3 67.4 92.3
No. of Tests 27 36 11 28
Indirect Tensile Strength
Mean (MPa) 2.72 5.76 4.17 6.62
No. of Tests 14 26 8 15
Mean UCS/Mean ITS 20.5 18.5 16.0 14.0
Modulus of Elasticity
Mean (Gpa) 32.80 25.12 16.5 22.55
No. of Tests 16 26 9 18
Mean E/Mean UCS 593 234 245 244
Poisson’s Ratio
Mean 0.31 0.28 0.22 0.27
No. of Tests 16 26 9 18
Range 0.12-0.50 0.10-0.56 0.14-0.30 0.13-0.37
88 5.2.3 Panel Design Criteria
The Pacific Highway, a main road connecting Sydney and Newcastle, runs over the lease and the NSW Department of Mineral Resources had raised concerns regarding it being undermined. To allay the concerns, the mine adopted a design philosophy that would result in surface subsidence develop only as a function of pillar yield, rather than as a result of caving of the overburden. The objective was to minimise the sudden “stepping” of the highway pavement by producing a smoother subsidence that would result from the yielding of the chain pillars as they gradually became overloaded by the sequential extraction of the first three to four longwall blocks.
The mine also desired to ensure maximised tailgate stability and serviceability. The mine commissioned a parametric study of various goaf/chain pillar widths that would minimise tailgate instability resulting from the interaction with the previously extracted longwall block. The results of this study, combined with the results of the subsidence assessment, resulted in the eventual adoption of the current layout of 95 metres panel width separated by 35 metres wide chain pillars.
The issue of wind blasts was assessed empirically, based on experience with similar layouts, in similar geological conditions at the nearby Wallarah Colliery, where mining of the FCT Panels 1-3 had apparently not resulted in any significant wind blast events. At Newstan Colliery, wind blast events had been experienced but under different geological conditions. Hence the initial risk assessment for wind blasts was low for Moonee Colliery.
5.2.4 Wind Blast Management Plan
Following the major wind blast events in January 1998, the risk assessment for Moonee Colliery was changed from a low to high likelihood. The Wind Blast Management Plan was put in place so as to provide “ a planned and systematic approach of predicting wind blasts, with the objective of reducing the risk of injury to personnel at a level of As Low As Reasonably Achievable (ALARA).” (Wind Blast Management Plan, July 1998)
89 The plan is a “live document” and is subject to constant review using generated data and conclusions drawn from independent audits. It incorporates the following procedures for reducing the risk of personal injury:
1. Details for predicting wind blasts, by utilising micro-seismic monitoring and roof support leg pressure monitoring systems;
2. Details for determining the wind blast magnitude and intensity utilising the UNSW MK1 wind blast monitoring system;
3. Engineering improvements including safe havens, modifications and or re location of equipment and wind blast shields so as to reduce the intensity of wind blasts;
4. Details of “Longwall Zoning Rules” to control the location and number of personnel in specific areas;
5. Safe working procedures to ensure a planned, consistent and controlled work environment;
6. Details of personal safety equipment and protective clothing to minimise personal injury.
5.2.5 Monitoring
Success of the wind blast management is dependant on implementing an early warning initiation process. Micro-seismic monitoring and support leg pressure monitoring system has been installed in Moonee longwall panels. Observation of “roof talk” or signs of “weighting” and their correct interpretation is not always possible in longwall panels, where visual or audible clues are either absent or masked by the presence of equipment or the noise of mining. Two “wind blast flaps” are also positioned on longwall equipment in the maingate near to the face and adjacent to the velocity pod no. 2. The flaps are driven during the wind blast and provide a positive trip by physically pulling apart the three magnetic reed type switches which isolate the power supply to conveyor and the face 3.3 KV drives. One of the wind flaps detects a wind blast from the inbye and other from the outbye. Wind blast monitoring, the main 90 objective of the research, is discussed in detail later in section 5.3.
5.2.5.1 Microseismic monitoring
Microseismic monitoring involves the use of seismic detectors (geophones or accelerometers) to record the seismic events associated with rock fracturing, including those below the threshold of detection by the human ear. This technology was initially developed to assist in the prediction of rock bursts in deep underground gold mines in South Africa, and has since been applied to the prediction of goaf falls in coal mines in New South Wales.
The Microseismic monitoring system installed at Moonee Colliery is an integration of real-time micro-seismic monitoring with an on-line alarm system. It normally consists of four geophones installed in vertical holes drilled into the roof of the roadways on either side of the longwall face. Two geophones are usually installed behind the face and two installed ahead of the face. The geophones detect the passage of seismic energy radiating through the rock mass from the site of crack formation. The geophone signals are then transmitted to a centralised data recorder that process the signals for computer based interpretation using the ISS software. The interpreted response provides the details of the magnitude, frequency and location of the seismic event. The main parameters that appear to give the best predictions of an event are: Apparent Volume- The calculated volume of rock that inelastic deformation as a result of seismic event and which is a cumulative indicator of seismic activity, based on the frequency and magnitude of successive events; Energy Index, Seismic Viscosity, Seismic Softening and Log 10 Schmidt Number-all of which can be likened to indicators for the onset of the rockmass failure. A typical alarm warning would be given in response to an accelerated climb in apparent volume, accompanied by a rapid decrease in two or more of the failure indicators - particularly Energy Index and Log 10 Schmidt Number. Interpretation of the data may be totally computerised or may require input from the trained seismic operator. Three types of warnings are issued at Moonee: Alarm Warning - These are issued based on interpreted trends in seismic activity build up. These are long term warnings, usually of a few hours, and provide sufficient time
91 for the mine personnel to evacuate the face; “Scram” Warning - These are seismic warnings issued on the basis of a sudden flurry of seismic activity. These are short term warnings, usually of a few minutes and provide the crew with sufficient time to move to safe havens; Auto Warning - These are activated automatically based on 6 geophone triggers in 10 seconds. At Moonee, immediately upon the identification by the operator of a significant seismic event, a shut down procedure is implemented, isolating power to the face equipment and outbye conveyor and triggering a visual alarm.
5.2.5.2 Roof support leg pressure monitoring
A roof support leg pressure monitoring system has been operational in the longwalls at Moonee Colliery. Changes in the load exerted by the roof upon the supports may be a precursor to a goaf fall.
The Citect software used at Moonee Colliery records instantaneous leg pressure measurements on each longwall support at three minutes intervals. The monitoring system has the following minimum specifications:
All support legs are monitored and the information is transmitted to the surface, the data is analysed to determine its trend and plotted to indicate changes in pressures over a time period. A screen is set up to show the overall face average pressure. The time base for the information storage is one month, the information is then archived. An alarm is triggered if three adjacent supports exceed the trip pressure setting of 35 MPa and a shutdown procedure for the longwall is then automatically activated.
5.3 WIND BLAST MONITORING
5.3.1 Introduction
The Wind Blast Monitoring System (WBMS) is briefly described in Chapter 3 of this thesis. Initial monitoring at Moonee revealed that the differential pressure transducers dynamic pressure range of ± 1 kPa was inadequate as wind velocity range was limited to ± 40 m/s. In order to overcome this limitation, the sensor pods were refurbished
92 between the monitoring of panels LW1 and LW2 and differential transducers replaced by Honeywell Micro Switch model 143PC03D which afford a range of dynamic pressure and wind velocities of ±14 kPa and ±150 m/s.
The sensor pods were calibrated after refurbishment in the wind tunnel at Defence Science and Technology Organisation’s Aeronautical and Maritime Research Laboratory in Melbourne, which affords the highest velocity of any suitable facility in Australia. The deviation from theoretical was less than 2 % at air speeds upto of 95 m/s, the upper limit for the wind tunnel. This range is compatible with the range of wind velocities subsequently recorded.
The Wind Blast Monitoring System initially installed in panel LW1 at Moonee Colliery was activated on 11 February 1998. The monitoring has since continued for panels LW2, LW3, LW4A and LW4B. The monitored results from the mentioned panels are tabulated in Appendix C.
5.3.2 Location of Instrumentation
The location of the sensor pods in LW2 panel are shown in Figure 5.4 which illustrates the panel configuration at the time of the initial goaf fall.
It can be seen that the air overpressure induced by the goaf fall can be relieved through several openings. As it was impracticable to monitor all these openings, a decision was taken to concentrate the Wind Blast Monitoring System in two areas relative to the face.
Firstly it was considered important to position a wind blast sensor pod as close as practicable to the working face in order to be able to measure the overpressure and velocity of potential wind blast events in the “near field”, i.e. before they were much attenuated with distance. In addition, in order to measure the attenuation with distance and the celerity, or rate of propagation, of potential events, it was considered necessary to place a sensor pod as far outbye as possible relative from the near field one. Secondly, it was thought desirable to obtain a comparison of wind blast velocities and overpressures within pairs of headings and between maingate and tailgate.
93 Figure 5.4 Longwall panel no. 2 and sensor pods locations at time of initial goaf fall, Moonee Colliery (after Fowler & Sharma, 2000)
94 Consequently, monitoring locations were selected in the maingate (two locations), in the companion roadway (travelling road) and in the tailgate. The Wind Blast Data Logger was positioned in a cut through immediately adjacent to the longwall transformers which provided it with an external power source. The rapid retreat of the longwall face made it essential to mount the inbye maingate pod no. 2 (Pod2) on an item of longwall equipment at a distance of 11 metres from the face such that it moved outbye as the longwall face retreated. The cables were run in the monorail from which were hung all power and other services to the longwall equipment. This created the advantage of automatic reeling in of the cable between the sensor pod and the Wind Blast Data Logger. However, the other potential disadvantages remained, such as the possibility of the pod being affected by zones of water ingress and need to traverse intersections which might render the data unreliable.
The other measuring location in the maingate, that for pod no. 4 (Pod4) was selected to be as far outbye as practicable in order to obtain the best possible precision in the determination of the attenuation with distance and the celerity of potential events.
The location of the pod no. 3 (Pod3) in the tailgate was determined by safety considerations. Ideally the location would have been just outbye of the working face. The installation of the pod had of necessity, to be undertaken at a time when the potential failure of the standing goaf did not present a wind blast hazard. In addition, sufficient allowance had to be made in the distance between the face and the measuring location for the face to retreat to induce a goaf fall. In practice, this resulted in the distance between face and the pod at the time of the wind blast being of necessity, somewhat variable.
The location for the pod no. 1 (Podl) in the maingate companion (travelling road), was chosen so that during the retreat of the longwall face it remained at a distance from the extended face line which generally lay between those of Podl and Pod3. However on occasions, safety considerations precluded the repositioning of Podl and consequently at the time of the wind blast it was situated “behind” the extended face line, a less than ideal situation (as in Figure 5.4).
95 In practice, each of the actual locations for pods numbers 1, 3 and 4 were selected to be a nominal distance of 10 meters inbye of an intersection, so as to minimise turbulence. Location was also adjusted so as to avoid zones of poor roof or rib or of water ingress or for other operational reasons.
5.3.3 Installation of Instrumentation
The installation of the Wind Blast Monitoring System was done with the assistance of the Moonee Colliery personnel. It was considered desirable to rigidly mount the sensor pods as close to the centroid of the roadways as possible but, in order to effect this, differing arrangements had to be made in the above mentioned locations.
For the inbye maingate location, pod no. 2, a bracket and a single “leg” were fabricated and securely bolted to the DCB, an item of the maingate equipment just outbye of the BSL crusher. The longwall panel layout showing location of these equipment is given in Figure 5.5. Provision was made for adjusting the height, altitude and azimuth of the sensor pod so as to be able to accommodate areas of low roof and to align the major axis of the pod with that of the roadway. In practice, the position of the sensor pod had to be offset laterally from the roadway centroid in order to accommodate the belt structure and vertically upwards to avoid turbulence occasioned by the presence of bulky maingate equipment. At each of the other locations, an inverted tripod was secured to the roof from a single roof bolt, the legs being adjustable in length to allow for roof irregularities. Provision was again made for aligning the sensor pods with the roadway.
In the tailgate and maingate (outbye location), the sensor pods were generally able to be mounted as close to the centre line of the roadways as the roof bolting pattern would permit. However, in the maingate companion road (travelling road), an offset from the centroid up to 2.0 metres horizontally and up to 0.5 metres vertically upwards was necessary to allow safe passage to service vehicles.
Following standard practice, an orthogonal system of setting-out lines had been established by the colliery surveying department on the theoretical centrelines of
96 Colli'
Moonee
layout,
heading
A
maingate
and
panel
longwall
Typical
5.5
e
97 headings and cut-throughs, and intersections of the setting-out lines had been marked by semi-permanent survey pin in the roof. In addition, “chainage markers” denoting the distance from the face start line had been set out in the roof in the maingate and tailgate. In order to facilitate the determination of the “face distance” of each sensor pod after its installation or reinstallation, the orthogonal distance of each pod to the nearest chainage mark or survey pin was determined.
5.3.4 Wind Blast Monitoring Procedure
Table 5.4 gives the location and the trigger levels for the Wind Blast Monitoring System for the LW1 - LW4B panels. The delay time was set at 100 milliseconds and the operating mode was fixed.
The objective being to maximise the probability that all of the events associated with a wind blast would be recorded but the events associated with other environmental factors, such as turbulence created by moving vehicle and variation in ventilation air flow, would not. As experience was gained in operating under the particular conditions of Moonee Colliery, adjustments to settings were made to reduce the possibility of spurious triggering.
Table 5.4 Wind Blast Monitoring System Trigger Levels at Moonee Colliery
Pod# Location Trigger Trigger Level LW1 LW2 LW3 LW4A LW4B Velocity (m/s) 6 20 - - 20 Pod 1 Travelling Absolute 109.5 109.5 - - 106.4 Road pressure (kPa) Velocity(m/s) 6 off off 20 20 Pod 2 Maingate Absolute 109.5 109.5 107.4 107.4 106.4 pressure (kPa) Velocity (m/s) 6 off off - - Pod 3 Tailgate Absolute 109.5 109.5 107.4 - - pressure (kPa) Maingate Velocity (m/s) 6 20 20 20 20 * Pod 4 (outbye) Absolute 109.5 109.5 107.4 107.4 106.4 * pressure (kPa) * Located in Travelling Road
98 The task of downloading data from the Wind Blast Data Logger was carried out utilising a standardised procedure developed for transferring data to the hand held interface. This was done in accordance with the Wind Blast Management Plan and as soon after an event as was practicable. Statutory safety inspections of the equipment were carried out in accordance with the prescribed schedule by the Moonee Colliery personnel.
After each goaf fall, an assessment was made as to whether the face was likely to “overrun” the location of pod no. 3 in the tailgate before the expected subsequent fall. When this was the case, the pod was relocated outbye (generally by 100 metres), the work being carried out in accordance with the Wind Blast Management Plan.
A similar assessment was performed for Pod no. 1 in the travelling road in order to reduce the possibility of the pod being “behind the face “ at the time of a wind blast. Pod no. 4, situated at the maingate outbye location, had also to be relocated from time to time as the face retreated
At appropriate times a service retraction were carried out. As part of the procedure, those items of longwall that did not retreat with the face were moved 2 pillar lengths outbye. Included in the move were the longwall transformers and, of necessity, the Wind Blast Data Logger.
For the LW2 panel (Figure 5.4), the WBMS was reinstalled and powered on 31 July, 1998. Velocity trigger levels had been raised in order to obviate the recording of small events, which tended to fill up the memory of the WBDL during the mining of panel LW1 and so decrease the risk of its already being full at the time of occurrence of a larger event. In addition, velocity triggering had been turned off for the two sensor pods located in the gateroads near to the face ends as their nose ports had tended to become blocked by debris during some of the more intense wind blasts that had occurred in the course of the mining of LW1 panel.
With all four pods activated and both velocity and pressure data recorded when any one of the eight trigger levels is exceeded, the probability of failing to record any significant event as a result of some of the triggers being turned off, was minimal.
99 In longwall panel no. 3 the Wind Blast Monitoring System was activated on 18 March, 1999. Earlier in the first two panels, the inbye pods’ nose cones tended to get blocked by dust during the wind blast event, resulting in truncated or distorted record and subsequent spurious triggering. The nose cone design was improved resulting in a ‘sharper’ nose the included angle decreasing from 90 degrees to 60 degrees. The design change successfully eliminated the problem of nose blockage, as observed during subsequent wind blasts. Initially 2 pods were installed, one in the maingate and other in the tailgate. An additional pod was installed in outbye location in the maingate on 23 March, 1999. In the travelling road the pod was installed on 8 April, 1999. Velocity triggers were turned off for the inbye pods and the absolute pressure trigger levels were set at 107.4 kPa to avoid spurious triggering. The outbye pod no. 4 in the maingate was set to trigger at 20 metres per second velocity and 107.4 kPa overpressure.
In LW4a panel, the Wind Blast Monitoring System was activated on 4 February 2000. Initially only two pods were installed. One near the face in the maingate (Pod2) and another outbye (Pod4). Both the velocity trigger and the absolute pressure trigger were set at 20 metres per second velocity and 107.4 kPa overpressure respectively. The tailgate pod was subsequently installed but was not operational for many wind blast events due to a cable fault.
In LW4b panel, initially only two pods were installed on 8 August 2000, one near the face in the maingate (Pod2) and another outbye in the travelling road (Pod4). Trigger levels were set at 20 metres per second for velocity and 106.4 kPa for the absolute pressure. Delay time was set at 100 milliseconds. The settings were changed later and velocity triggering of the maingate pod turned off to avoid spurious triggering. An additional pod (Podl) was installed in travelling road inbye of Pod4 with same trigger levels. This was done to avoid the probability of failing to record any significant event as a result of some of the triggers being selectively turned off on the maingate pod or failing to trigger the outbye pod in the travelling roadway. No pod was installed in the tailgate.
100 5.3.5 Analysis of Field Data
Data which had been downloaded to the hand held interface were transcribed on to floppy disks using standard hardware and software. After each data file had been pre- processed, it was displayed (and printed, if required) in tabular mode, as shown in Appendix A for overpressure and velocity channel of one wind blast sensor pod. Files, which were not the result of spurious triggering were then further processed and graphical output produced using macros written in WaveMetrics Igor Pro version 3.13. The advantage of using this procedure was that the analysis was automated to a large extent and standardised.
Standard graphical output included, for each sensor pod location, the overpressure time history together with its integral and differential from which the impulse and rates of rise (and fall) of overpressure were obtained. Also included was the wind velocity time history together with its integral and differential from which the excursion and rates of rise (and fall) of velocity were obtained. Examples of some processed overpressure time histories of wind blast events at Newstan and Moonee Collieries are given in Appendix B.
Further data analysis utilised Microsoft Excel version 7.0 and SPSS DeltaGraph version 4.5.
The peak wind blast parameters recorded and generated from the time histories for each monitored event in the longwall panels at Moonee Colliery are tabled in Appendix C.
The raw time histories were smoothed by applying a “boxcar” smoother to obtain the half-second sliding mean. Figure 5.6 illustrates the effect of smoothing. The black line is the half-second sliding mean. The values in this record are spaced at one millisecond intervals. The smoothing operator acts consecutively on each value, computing the mean of the current value, the 250 values immediately preceding the current value and the 250 values immediately following. It gives equal weight to each value. Figure 5.7
101 200299# 1.pod2 Wind velocity time history
Elapsed Time (seconds)
Figure 5.6 Effect of‘smoothing’ the wind blast velocity time history
170499#2.pod2 Wind velocity time history
I'Suck back'[-
Elapsed Time (seconds)
Figure 5. 7 Wind velocity time history for goaf fall no. 1 in LW3,
Moonee Colliery
102 shows the smoothed wind blast time history for the first goaf fall in LW3 panel.The choice of a half-second base for the sliding mean is arbitrary. However it effectively eliminates from the time history all components with a frequency greater than 2 Hz (corresponding to a period of 500 ms). It is considered that “elements” of the mine such as personnel and stoppings are little affected by such high frequencies. The process of smoothing moderated the over pressure and wind velocity time histories and consequently, influenced the parameter values (Chapter 4, Section 4.3.1) including the following:
Peak wind blast velocity;
Maximum rate of rise of wind blast velocity;
Peak dynamic pressure;
Peak overpressure;
Maximum rate of rise of overpressure;
The values of maximum excursion (air flow distance) and impulse were unaffected as they are derived by integrating the respective “raw” time histories.
5.4 RESULTS OF MONITORING IN LW1 AND LW2 PANELS
During the mining of LW1 panel, both velocity and overpressure time histories were recorded in the maingate and travelling road and in the tailgate for a total of 24 events. Of these, 15 were considered significant wind blasts, i.e. of sufficient intensity to pose a risk of personal injury or of damage to the mine ventilation system.
During the mining of LW2 panel, trigger levels were increased to reduce the number of small events which were captured earlier. A total of 14 events were instrumentally recorded, of which 12 were considered to be significant. Goaf fall areas during the mining of LW2 were generally more extensive than those of LW1. The largest goaf fall in LW2 was in excess of 31,000 square metres and generated the most intense
103 wind blast that was instrumentally recorded. However from physical evidence, it is clear that this wind blast was by no means the most intense that has occurred in an underground coal mine. The maximum value of the key parameters recorded during significant wind blasts in above two panels are given in Table 5.5. The results of LW1 and LW2 panels are given in Table C.l and C.2 respectively in Appendix C. The parameters were defined in Chapter 4, Section 4.3.1
Table 5.5
Wind blast parameters, LW1 and LW2 panels, Moonee Colliery
Parameter Maximum value recorded during
the mining of LW1 and LW2
Peak wind blast velocity 130 m/s
Maximum rate of rise of wind blast velocity 174 m/s/s
Maximum excursion 181 m
Peak dynamic pressure 11 kPa
Peak overpressure 34 kPa
Maximum rate of rise of overpressure 38.5 kPa/s
Impulse 89 kPa.s
Comparing with Table 5.6, the Beaufort Scale of wind speeds, it will be observed that the peak wind blast velocities are well into the hurricane range, for which the lower bound is 33 m/s (118 km/hr).
5.4.1 Overpressure and Wind Velocity Time Histories
The overpressure time history recorded in the tailgate near to the face during goaf fall no. 6 of the LW2 panel is given in Figure 5.8. The figure illustrates the following salient features:
A wind blast comprises a rapid rise in the absolute pressure to a maximum, followed by a similarly rapid fall to below ambient atmospheric pressure. After decreasing to a minimum value, the absolute pressure gradually increases until it becomes equal to ambient atmospheric pressure.
104 At around the same time, although not necessarily in phase with the overpressure, the wind velocity also rises rapidly to a maximum and then exhibits a sudden reversal into the “suck back” phase.
Table 5.6
The Beaufort Scale of wind speeds (adapted from Linacre & Hobbs, 1977)
Beaufort Descriptio Nautical Land based Equivalent Scale n observation observation number wind speeds at 10 m
Knots km/h m/s
0 Calm Flat sea Smoke vertical 0 0 0 1 Light air Ripples form Smoke drifts 2 4 1 2 Light Wavelets Leaves rustle 6 11 3 3 Gentle Breaking wavelets Wind felt on face 9 17 4.5 4 Moderate White horses Thin branches 14 26 7 breeze move 5 Fresh Moderate waves Small, leafy trees 17 31 9 breeze sway 6 Strong White foam spray Large branches 23 43 12 breeze sway 7 Moderate Heaped sea Whole trees move 29 54 15 gale
8 Gale Long crests, blown Twigs break off, 37 68 19 foam progress impeded 9 Strong gale 10m waves, reduced Removes tiles 43 80 22 visibility 10 Storm Heavy rolling sea, Trees blown 50 93 26 over-hanging crests down
11 Spray impedes Widespread 58 107 30 visibility damage 12 Hurricane Sea white with foam Extreme damage 64 118 33
105 The peak overpressure is always greater than the peak underpressure.
The duration of the negative phase is often longer than that of the positive phase, i.e. the underpressure often lasts for longer than the overpressure
There is no acoustic precursor to the event. Consequently people in the working place will receive no warning of the wind blast before it strikes them unless they hear “roof talk” or a wind blast warning system is in place.
The onset of the event is very sudden, both overpressure and air velocity exhibiting a rapid rise. Very little time is available for the personnel to take cover, drop to the floor or even “brace” themselves.
The intensity of the wind blast phase can be very severe, comprising a very large overpressure and corresponding high wind velocity.
The intensity of the “suck back “ phase can also be as severe comprising a large underpressure and often a high suck back velocity.
Figure 5.8 illustrates the overpressure time history recorded in tailgate and Figure 5.9 illustrates the overpressure histories recorded at all four measuring locations, namely the gateroads near the face ends, the travelling road and outbye in the maingate itself during goaf fall no. 6 of the LW2 panel. The following salient points are observed:
Overpressure time histories are similar, but not identical, for equivalent locations in both gateroads in the near field close to the face ends.
It will be seen by comparing the time histories that the two overpressure pulses which are propagating in the maingate and travelling road are out of phase and that the maingate pulse is of a greater intensity than that in the travelling road. The two pulses are believed to “equalise” as they propagate outbye by mechanisms which include flow through cut-through(s) immediately outbye the face and frictional losses.
The overpressure pulse in the maingate will be seen to be attenuated with distance as it propagates outbye. Consequently, its peak overpressure and underpressure 106 (and its peak wind velocity and “suck back” velocity) all decrease as it moves away from the goaf. The attenuation is believed to be predominantly as a result of air viscosity, although other factors, such as that mentioned above, may contribute.
It will be noticed that the duration of the positive overpressure phase increases slightly as the pressure pulse propagates outbye. This is believed to be due to dispersion, i.e. the tendency for different frequencies to travel at slightly different velocities. The successive high-pressure portions of the compression pressure pulse travel at higher sonic speeds and overtake the preceding low-pressure portions.
A lapse in time between the arrival of the overpressure pulse at the measuring location in the maingate near the face and at the outbye location is also evident. From the time difference and the distance between the measuring locations, the celerity or the rate of propagation of the event may be calculated. For weak shocks or pressure pulses the celerity equals the speed of sound.
While most of the goaf falls have resulted in overpressure time histories similar to that shown in Figure 5.9, some compound events have also been recorded as shown in Figure 5.10. A triple fall is believed to have given risen to this event, an inference that is independently supported by microseismic evidence (Edwards, 2000).
Wind velocity time histories also follow the same overall shape as the overpressure time histories, with rapid rise to a maximum and then a sudden reversal into the suck back phase. As peak overpressure is always greater than the peak underpressure, consequently it is believed that the peak air velocity in the direction away from the fall is always greater than the peak “suck back” velocity as shown in Figure 5.7. The raw time histories exhibit “spikiness”, believed to be due to intense turbulence in the air flow, which tend to obscure the underlying characteristics of the record. This was particularly pronounced in the case of wind velocity time histories.
Figure 5.7 illustrates an event recorded in the maingate near to the face. However, not all velocity time histories clearly exhibit flow reversal, as shown in Figure 5.11 for goaf fall no. 11 in longwall panel no. 2. It will be seen that the peak wind velocity is 80 metres per second but that the suck back is almost absent. The latter may be due to the
107 Peak Overpressure 101198#1.oppod3 32 kPa Overpressure time history
■Ambient Atmospheric Pressure-
Peak Underpressure 10 kPa
Elapsed Time (seconds)
Figure 5.8 Overpressure time history in tailgate, goaf fall no. 6, LW2,
Moonee Colliery.
Pod 3 (tailgate, Event No. 101198# 1 near face end) Pod 2 (maingate, Overpressure time histories near face end)
Pod 1 (travelling road) Pod 4 (maingate, 552 metres outbye Pod 2)
Elapsed Time (seconds)
Figure 5.9 Overpressure time histories around face and outbye in maingate,
goaf fall no. 6, LW2, Moonee Colliery
108 | First goaf fall | 190199#1 +2.oppod2 Overpressure time history
I Second goat fall]
|Third goaf tall|
Elapsed Time (seconds)
Figure 5.10 Overpressure time history of goaf fall no. 10, maingate LW2,
near to the face, Moonee Colliery
Peak Wind Blast 220199*1 pod3 Velocity = 80 m/s Wind velocity time history
Elapsed Time (seconds)
Figure 5.11 Wind velocity time history of goaf fall no. 11,tailgate LW2,
near to the face Moonee Colliery
109 time history being prematurely truncated as a result of the recording time being restricted to a nominal eight seconds. The same has also been observed by the failure of the inbye “wind flap” switch located near to pod 2 in the maingate of the longwall panels, to operate during many of the wind blast events monitored. However the outbye wind flap switch has always been triggered by significant wind blast events, near to the face Moonee Colliery It must also be noted that the velocity pods are pointed inbye towards the face and such orientation is not suitable to accurately measure the suck back velocities.
5.4.2 Analysis of Overpressure and Wind Velocity Data
This section compares some of the interrelationships between the parameters listed below for the wind blast events recorded at Moonee longwall panels LW1 and LW2.
The data set is separately analysed because in LW3 panel, attempts to hydro-fracture the roof began. The effects of hydro-fracturing on wind blast parameters are analysed in Section 5.5. The parameters are:
Peak overpressure;
Impulse;
Maximum rate of rise of overpressure;
Peak wind blast velocity;
Maximum excursion;
Maximum rate of rise of wind blast velocity;
Distance from the goaf fall; and
Roof fall area.
5.4.2.1 Characteristics of the overpressure time history
Figure 5.12 illustrates the relationship between impulse and peak overpressure. It utilises data from the subset obtained from pod no. 2, which was mounted in the maingate at a fixed distance of 11 metres from the face.
110 The relationship is non-linear and a second-degree polynomial curve has been fitted to the data. The coefficient of correlation is high, as given by the R square value in Figure 5.12.
The implication of the non linearity is that as wind blasts become more intense, the impulse increases more than the overpressure, i.e. the duration of the positive (compression) phase becomes longer. This is also suggested by visual inspections of the overpressure time histories.
Figures 5.13 and 5.14 illustrate the relationship between the maximum rate of rise of overpressure and peak overpressure at Pod 2 and Pod 4 locations respectively. The relationship is again non-linear a second-degree polynomial curve has been fitted to the data. The coefficients of correlation are reasonably high as given by the R square values in the respective figures. The trends illustrate that as wind blasts become more intense, the maximum rate of rise of overpressure increases non-linearly and a decreasing rate compared to peak overpressure.
5.4.2.2 Characteristics of wind velocity time history
Figure 5.15 illustrates the relationship between the maximum excursion and peak wind blast velocity. It utilises data from the subset obtained for Pod 4 which has been mounted on an inverted tripod secured to the roof of the maingate at an outbye location The relationship is non-linear and a second-degree polynomial curve has been fitted to the data. The implication of non linearity is that as wind blasts become more intense, the maximum excursion increases more than does the peak wind blast velocity, i.e. the duration of the positive (wind blast) phase becomes longer. This is also suggested by the visual inspection of the wind velocity time histories.
The discussed relationship is also well defined at Pod 2 location, near the face as illustrated in Figure 5.16.
Figure 5.17 illustrates the relationship between maximum rate of rise of wind blast velocity with peak wind blast velocity. It utilises data from the subset obtained from Pod 3. In this case, the relationship is linear and a straight line has been fitted to the
111 100 n Pod 2 on DCB in maingate
O Moonee LW1 & LW2 (all data)
—Poly. ( Moonee LW1 & LW2 (all data))
y = 0.042X2 + 0.79x R2 = 0.95
J= 40 -
Peak Overpressure (kPa)
Figure 5.12 Relationship between impulse and peak overpressure
Pod 2 on DCB in maingate
O Moonee LW1 & LW2 (all data)
------Poly (Moonee LW1 & LW2 (all data)) i 35 -
2 30 -
6 25 -
ra 15 -
i 10- y = -0.021 x2 + 1.53x R2 = 0.85
Peak Overpressure (kPa)
Figure 5.13 Relationship between max. rate of rise of overpressure and peak overpressure 112 Pod 4 in maingate (outbye)
© Moonee LW1 & LW2 (all data)
------Poly (Moonee LW1 & LW2 (all data))
o 10 -
<3> O
-0.030X2 + 1,58x R2 = 0.79
Peak Overpressure (kPa)
Figure 5.14 Relationship between maximum rate of rise of overpressure and peak overpressure
Pod 4 in maingate (outbye)
© Moonee LW1 & LW2 (all data)
------Poly. (Moonee LW1 & LW2 (all data))
Peak Velocity (m/s)
Figure 5.15 Relationship between maximum excursion and peak velocity
113 data. The coefficient of correlation is reasonably high as shown by the R square value in Figure 5.17. The implication of linearity is analogous to that for overpressure, i.e. as wind blasts become more intense, the maximum rate of rise of wind blast velocity increases in direct proportion to the peak wind blast velocity. This suggests, although it is not yet “proven”, that the wind blast velocity rise time may be independent of the peak wind blast velocity. Consequently it is suspected that an increase in the wind blast velocity fall time, rather than the rise time, which is responsible for the increase in the duration of the positive (wind blast) phase discussed in Chapter 4, section 4.3.1
5.4.2.3 Relationships between overpressure and wind velocity time histories
The Figure 5.18 illustrates the relationship between peak wind blast velocity and peak overpressure. It utilises the data set from Pod 4. It will be observed from the figure that though the parameters are evidently positively correlated their relationship is far from clear. The parameters shows a tentative linear relationship with the following equation:
Peak wind velocity (in m/s) = 3.75 x peak overpressure (in kPa) (5.1)
In case of an ideal explosive shock front, the relationship between peak particle velocity “blast wind” and peak overpressure can be derived from the Rankine - Hugoniot equations as discussed in Chapter 4 section 4.1.3.1. The relationship for low overpressures is effectively linear and can be stated by the following equation for
o explosive shock in dry air at 21 C and at sea level (Kinney, 1962)
Peak particle velocity (in m/s) = 2.17 x peak overpressure (in kPa) (5.2)
The Rankine-Hugonoit relationship is close to the lower bound of the data. The extreme case in Figure 5.18 is given by the following equation:
Peak wind velocity (in m/s) = 7.0 x peak overpressure (in kPa) (5.3)
Figure 5.19 illustrates the above relationship at Pod2 location, near the face, given by the following equation:
Peak wind velocity (in m/s) = 4.1 lx peak overpressure (in kPa) (5.4)
114 200 -| Pod 2 on DCB in maingate
O Moonee LW1 & LW2 (all data) 160 ------Poly. (Moonee LW1 & LW2 (all data))
120 - y = 0.0104X2 + 0.74x R2 = 0.85
Peak Velocity (m/s)
Figure 5.16 Relationship between maximum excursion and peak velocity
180 Pod 3 in tailgate
160 - O Moonee LW1 & LW2 (all data)
------Linear (Moonee LW1 & LW2 (all data)) 140 -
> 100 -
£ 80 -
y= 1.31x R2 = 0.92
2 40 -
Peak Velocity (m/s)
Figure 5.17 Relationship between maximum rate of rise of velocity and peak velocity 115 Pod 4 in maingate (outbye)
O Moonee LW1 & LW2 (all data)
100 ------Linear (Moonee LW1 & LW2 (all data))
-----Linear (Rankine-Hugoniot)
y = 3.75x g 60 ■ R2 = 0.54
Peak Overpressure (kPa)
Figure 5.18 Relationship between peak velocity and peak overpressure
160 Pod 2 on DCB in maingate 140 ■ O Moonee LW1 & LW2 (all data) -----Linear (Rankine-Hugoniot) y = 4.11x
------Linear (Moonee LW1 & LW2 (all data)) R2 = 0.71 120 -
« 100 -
£ 80 •
Peak Overpressure (kPa)
Figure 5.19 Relationship between peak velocity and peak overpressure near to the longwall face
116 At Pod2 location, the roadway cross-section is constricted due to the mining equipment including the DCB over which the Pod was mounted.
5.4.2.4 Attenuation of wind blast intensity with distance from goaf fall
Attenuation of peak overpressure with distance along the maingates of longwall panels is shown in Figure 5.20. It utilises data from Pod 2, which has been mounted in the maingate at a fixed distance of 11 metres from the face and that from Pod 4 mounted on an inverted tripod secured to the roof of the maingate at an outbye location. The ordinate of the graph is the quotient of the peak overpressures while the abscissa is the distance between the measuring locations.
The parameters are evidently negatively correlated. An exponential decay curve has been (tentatively) fitted to the data from which it is observed that the peak overpressure is approximately halved every 500 metres as the wind blast travels outbye.
Less data is available for peak wind velocity attenuation with distance along the maingates of LW1 and LW2, insufficient for an analysis to be performed. The attenuation of both the overpressure and wind blast velocity along the double entry system on the maingate side of the longwall block is considered to be mainly due to the viscosity of the air and frictional resistance, another factor, geometric spreading, may result in a further reduction in wind blast intensity as a wind blast propagates through a “network” of interconnected roadways and cut-throughs.
5.4.2.5 Relationship between wind blast intensity and overall roof fall area
Figure 5.21 illustrates the relationship between peak overpressure and roof fall area. It utilises the data subset obtained from Pod 2. The fall area was determined by inspection some time after the fall or falls had occurred. However, wind blast time histories as well as microseismic evidence suggests that on many occasions the roof failed in a sequential or progressive mode rather than the assumed monolithic mode, leading to multiple events being registered in the Wind Blast Data Logger. The overall area has of necessity, been assigned to the event with the highest recorded intensity.
117 Although the data in Figure 5.21 are positively related, it tends to ‘plateau’ for larger falls. It is considered that there are three potential arguments as to why there might be an upperbound to peak overpressure, generated by roof falls of a particular thickness, regardless of area (Fowler & Sharma 2000):
1. The “geological argument”. As the plan area of the potential fall increases, the probability of its falling as a “monolithic piston” decreases and potential “piston efficiency” drops. This is evidenced by some of the larger goaf falls that appear to be have been compound or multiple events, as illustrated in Figures 5.10
2. The “acoustic argument”. It takes a finite length of time for a pressure transient to travel the full length of a falling goaf, for example of the order of one second, to cross a 300 metre long goaf. Consequently, the duration of the pressure pulse increases for falls of larger area, moderating the increase in peak overpressure that would otherwise occur;
3. The “fluid mechanics argument”. The greater increases in air pressure beneath the falling roof, associated with larger areas, begin to significantly affect the duration of the fall, i.e. the roof no longer falls at an acceleration which is effectively 1 g.
An upper bound has been appended to the graph in Figure 5.21. However this is very tentative. Later in the chapter in Figure 5.39 the relationship is defined with data available from all the panels monitored.
Figure 5.22 illustrates the relationship between impulse and goaf fall area. The data reflects the trend for linearity. However, the positive correlations are quite poor as given by the R square values in Figure 5.22. The implication of a linear relationship is that unlike peak overpressure, impulse does not tend to “plateau” for larger falls and that, consequently the duration of the positive (compression) phase increases as the fall area increases. The same is corroborated by the visual inspections of the overpressure time histories.
This appears to support the postulated “acoustic argument” or “fluid mechanics” argument above rather than the “geological argument”. The relatively high overpressure values recorded for the falls associated with triangle shaped areas of fall,
118 Pod 2 & 4 in maingate
0.90 -
0 80 - -0.0013x ■g 0.70 -
q. 0.50
0.40 -
o 0.30
§ 0.20 -
0.10 ♦ Moonee Lw1 & Lw2 (all data) ----- Expon. (Moonee Lw1 & Lw2 (all data))
200.0 400.0 600.0 800.0 1000.0 1200.0 Distance between pods (m)
Figure 5.20 Attenuation of peak overpressure with distance along maingates of longwall panels
Pod 2 on DCB in maingate Upper bound
Upper bound
i 25 - -4E-08X2 + 0.0023X R2 = 0.66
O Moonee LW1 & LW2 (all data)
------Poly. ( Moonee LW1 & LW2 (all data))
10000 15000 20000 25000 30000 35000 Overall Area of Roof Fall (m2)
Figure 5.21 Relationship between peak overpressure and goaf fall area
119 formed due to geological structures near the junctions of the maingate and tailgates of the panels, and which fell after the main goaf falls, also support the above arguments. The high value anomalies associated which such falls become apparent when the peak overpressure data is normalised. However, definite conclusions cannot be reached. The wind blast monitoring has also revealed some falls to be multiple events (sequential falls) and this has been confirmed by microseismic monitoring results. In a mining environment, with structural discontinuities present, the “geological argument” is also equally valid.
Similarly, Figure 5.23 illustrates the relationship between peak wind velocity and overall goaf fall area for data set from pod no. 2 located on the DCB in the maingate. The trends are similar to that for overpressure and the peak wind velocity also tends to “plateau” for larger falls. An upper bound has been tentatively indicated to the figure.
5.4.3 Roof Failure and Precusors to Wind Blast Events
Tables 5.7 and 5.8, present the results of microseismic and support leg pressure monitoring for LW1 and LW2 panels, respectively. It will be seen that rarely has audible warning (roof talk) been the alarm criteria. Similarly, Leg Pressure Strata Alarm has also been quite rare at Moonee. The microseismic warning times have been short and in a few occasions no warnings preceded the wind blast events. On many occasions false alarms have also been triggered.
Chocks hitting the roof during support setting after the support movements also give rise to false signals.
Evidence of periodic weighting was not observed, nor any long term trends (72 hours observation) in leg pressures leading up to the goaf falls (Macdonald, 2000). The Citect software utilises data obtained by recording leg pressure measurements at three minutes intervals. Figures 5.24 & 5.25 illustrate the leg pressures recorded for fall no. 6 at LW2 panel for a two hour period preceding the fall that occurred at 14:13:08 on 10-10-98. A seismic strata alarm from micro-seismic monitoring preceded this fall by 26 minutes. From Figure 5.25 it is clear that, leg pressures of the supports towards the tailgate were lower than those towards the maingate at the time of the goaf fall.
120 Pod 2 on DCB in maingate
y = 0.0029X R2 = 0.74
O Moonee LW1 & LW2 (all data)
------Linear ( Moonee LW1 & LW2 (all data))
10000 15000 20000 25000 30000 35000 Overall Area of Roof Fall (m2)
Figure 5.22 Relationship between impulse and goaf fall area
Pod 2 on DCB In maingate Upper bound
Upper bound
y = -3E-07X2 + 0.0125x R2 = 0.64
O Moonee LW1 & LW2 (all data) ------Poly. ( Moonee LW1 & LW2 (all data))
10000 15000 20000 25000 30000 35000
Overall Area of Roof Fall (m2)
Figure 5.23 Relationship between peak velocity and goaf fall area
121 Colliery
Moonee
maingate,
to
tailgate
LW2,
6,
no.
fall
goaf
preceding
pressures
leg
Support
5.24
Figure
122 Coll
Moonee
LW2,
tailgate,
to
maingate
6,
no.
fall
goaf
preceding
pressures
leg
Support
5.25
Figure
123 Table 5.7
Monitoring of Goaf Falls at Longwall Panel no.l, Moonee Colliery (adapted from wind blast notes, Moonee Collieiy)
Date Goaf Fall Face Goaf Fall Alarm Audible Seismic Actual
Fall Area Location Hanging Time Type Warning Alarm Warning No. sqm m m hrtmin time Criteria time
11-Mar 14a 5 770 1.401 67 22:40 seismic freq 6 min
12-Mar 14b 460 1,401 0 1:30 none freq no
13-Mar 15 390 1.399 2 5:00 none frea/mag. no
14-Mar 16 1.630 1.384 15 1:02 n/a freq n/a
20-Mar 17 4.470 1,336 48 5:58 seismic 1 min freq/trend 8 min
24-Mar 18 4.230 1.295 41 19:31 audible 1.5 min ffeq/trend 1.5 min
26-Mar 19 1,460 1.275 20 2:57 seismic freq 9 min
26-Mar 20 140 1,267 8 12:06 none no
3-Apr 21 9.830 1.170 97 2:07 seismic freq/mag./trend 2 min
3-Apr 22 380 1.170 0 6:30 none no
9-Apr 23 7.610 1.075 95 12:34 audible 2-3 min freq/mag./trend 2 min
15-Apr 24 1.450 1.065 10 6:20 seismic ffeq/mag./trend 46 min
23-Apr 25 9.800 938 127 13:38 seismic 10 min freq/mag./trend 13 min
27-Apr 26 2,290 910 28 23:42 seismic freq/trend 1 min
28-Apr 27 320 906 4 17:05 seismic freq/mag./trend 10 min
29-Apr 28 3.220 900 6 20:41 audible 20-30 s freq/trend 20-30 s
7-Mav 29 600 900 0 22:10 seismic freq/mag. 1 hr 37 min
14-Mav 30 762 138 2:28 none freq/mag. no
19-Mav 31 3.380 724 38 20:50 seismic freq/trend 24 min
21-Mav 32 2.600 694 30 22:04 seismic freq/mag./trend 3 hrs 29 min
26-May 33 3.920 635 59 18:34 leg pr. freq/mag./trend 2 hrs 19 min
3-Jun 34 8,600 529 106 2:54 seismic freq/mag./trend 5 min
4-Jun 35 5.390 502 27 6:13 seismic freq/mag./trend 1 hr 59 min
12-Jun 36 11.200 393 109 6:27 seismic freq/mag./trend 1 hr 30 min
21-Jun 37 13.000 262 131 10:30 n/a n/a
23-Jun 38 undefined 255 7 20:41 none no
26-Jun 39 5,300 193 69 13:24 leg pr. freq/mag. 6 s
3-Jul 40 9,040 111 82 4:01 Seismic freq/mag. 30 s
124 Table 5.8 Monitoring of Goaf Falls at Longwall Panel no. 2, Moonee Colliery (adapted from wind blast notes, Moonee Colliery)
Date Goaf Fall Face Goaf Fall Alarm Audible Seismic Actual
Fall Area Locat Hanging Time Type Warning Alarm Warning
No. sqm m m hr:min:s time Criteria
19-Sep-98 1 24,308 1,735 248 13:03:28 Seismic freq/mag./ 8 min trend
6-Oct-98 2 12.900 1.594 130 15:16:12 Seismic frea/maa. < 7 s
13-Oct-98 3 6,480 1,503 76 17:11:40 Seismic freq/mag./ 10s rend
22-Oct-98 4 13,940 1,366 159 18:10:40 Seismic freq/mag./ 9 min 23 s trend
30-Oct-98 5 6.160 1.292 76 10:36:29 Seismic 30 sec frea/maa. 8 min 51 s
IO-Nov-98 6 12,010 1,141 139 14:12:16 Seismic 26 min freq/mag./ 26 min 14 trend s
19-NOV-98 7 12.540 1.006 125 9:20:56 Seismic frea/maa. 25 s
4-Dec-98 8 undefined 894 - 13:30:30 Seismic/ 15 min freq/mag./ 15 min trend leg pr.
18-Dec-98 9 3.230 824 45 3:35:30 Seismic 1 min frea/maa. 1 min 8 s
19-Jan-99 10 31.560 508 303 0:29:23 Seismic 30 sec frea/maa. 2 min 15 s
22-Jan-99 11 3.910 473 48 12:46:25 Seismic frea < 6 s
7-Feb-99 12 11.450 314 134 n/a n/a n/a
20-Feb-99 13 15.878 153 162 8:52:45 Seismic frea/maa. 2 min 1 s
A trough in leg pressures immediately before the fail was observed. This is caused by a sudden decrease in leg pressure within an hour before the fall. The leg pressures remain constant until immediately before the fall, when the leg pressure rapidly increases to the previous level or greater forming a trough. Some of the other falls studied resulted in similar observations. However this feature cannot be used as a precursor to the fall as similar troughs were observed over the 72 hours leading to the fall.
Caving is influenced both by geological features and by the layout of the longwall panel and does not follow a predictable pattern. It will be seen that the frequency of the events reduced for LW2 panel in comparison with LW1 panel.
125 The reasons could be attributed to changes in local geology related to thinning of the Teralba conglomerate, changes in pore pressure regime and also the effect of adjacent, previously caved LW1 panel. A brief description of the first falls in the first 3 longwall panels is given below. Induced caving of roof was introduced at Moonee after fall no. 2 in LW3 panel and the same has been discussed in detail in section 5.5 of this chapter.
5.4.3.1 LW1 panel first fall
The first fall in LW1 panel occurred on 10 January, 1998 at 12.50 am, after the face had retreated 200 metres. It was perhaps the most intense fall as indicated from the damage report outlined in Table no. 5.2.
The goaf fall area was approximately 18,510 square. Fortunately, no one was injured as it was a Saturday morning with no production in the panel. However, 10 mins prior to the fall a crew leader & 2 technicians were in the tailgate and noticed no unusual noises or signs. At the time of the fall this crew were travelling in a mine vehicle outbye of the panel when they noticed their ears ‘pop’ followed shortly after by a blast of wind travelling outbye, a suck back inbye then a small secondary blast back travelling outbye.
5.4.3.2 LW2 panel first fall
The first fall in LW2 panel occurred on 19 September, 1998 at 13.03 hrs. No one was present in the longwall panel as it was a weekend. An eight minutes seismic warning was given. The face had retreated 248 metres and the goaf fall area was approximately 24,308 square metres. Figure 5.4 in section 5.3.2, gives the plan of LW2 panel at time of first fall. Both outbye and inbye Wind Flap trip were triggered suggesting that the wind blast in the outbye direction was followed by a suck back in the inbye direction. Additionally, there was evidence of wind blast from both directions in maingate (B heading) due to mud on both sides of machinery. Water barriers were damaged between the boot end and pump station. The brattice stopping in travelling roadway (A heading) was blown 130 metres outbye. All the goaf had fallen except for a triangular area in the tailgate approximately 15 metres x 25 metres in dimension. There was very
126 little evidence of wind blast in cut throughs no. 20, 21, 22 and 23 in maingate 2. Also there was very little evidence of wind blast along the face. There was only one person underground at the time and he was near pit bottom when the event occurred. Table 5.9 gives the damage caused by this fall.
5.4.3.3 LW3 panel first fall The first fall in LW3 panel occurred on 17 April, 1999 at 2:09:34 hrs. The face had retreated 283.4 metres prior to the fall and the goaf fall area was approximately 26,980 square metres. The plan area of the fall is shown in Figure 5.26, which shows all the goaf falls in longwall LW1 - LW4B panels. A seismic strata alarm warning time of 3 minutes 4 sec preceded the event. No persons were present in the longwall panel at that time. However a person was working in Longwall 4 maingate and reported a 3 minutes
Table 5.9
Physical damage caused by wind blast, Moonee Colliery, LW2 panel, first goaf fall (adapted from Wind Blast Notes, Moonee Colliery)
Damage Location in panel Water barriers damaged Between bootend and pump station 4 fluorescent lights dislodged Tensar mesh dislodged and blown 20metres outbye Maingate A hdg Face ventilation blown over Relief stopping blown over Slight damage to LW regulator Tailgate (Outbye) Brattice Stopping blown 130 metres outbye Maingate A hdg audible warning despite 140 metres of solid strata between the person and the goaf. A triangular area of roof extending for 30 metres along tailgate A heading and across to no.43 roof support remained hanging at the tailgate end of the goaf. There was a small triangular area hanging in the maingate. Roof supports no.l and no.2 were peppered with mud giving evidence of a wind blast or suck back from outbye to inbye. This was supported by the outbye wind flap being triggered. A high fall was evident from observations at 20, 21, and 22 cut-throughs of maingate 3. The estimated height above the conglomerate was 6 metres. The overall intensity of the fall appears to
127 Coll
Moonee
panels,
longwall
in
falls
Goaf
5.26
Figure
128 have been much less compared to LW2 first fall. The reason may be that the roof failure was sequential and that the event was the largest of the series and also due to assisting influence of the previously caved adjoining goafs. Table 5.10 gives the damage caused by this fall.
Table 5.10
Physical damage caused by wind blast, Moonee Colliery, LW3 panel, first goaf fall (adapted from Wind Blast Notes, Moonee Colliery) Damage Location in panel
Water Barriers (6) Two on the barrier just outbye the boot end and four inbye and outbye the pump station
Stopping door pushed out. Dyke stopping Acquacrete stopping shattered Cut-through 9 Damage to door hinges caused by MG4, cut-through 7, 8, 11 and 12 acquacrete stopping doors being blown open. stoppings Plaster stopping blown over. Cut-through 13 The relief door blown open and MG3, cut through 8 damaged slightly.
5.5 INDUCED CAVING BY HYDRAULIC FRACTURING
Induced caving of the goaf using hydraulic fracturing technique was first tried at Moonee Colliery LW3 panel with collaboration of SCT Operations and CSIRO Petroleum to produce controlled caving events “on demand”.
The successful use of this technique has reduced the exposure of men to wind blast by taking control of the timing of caving events that earlier were mostly unpredictable. However, even induced caving of roof has resulted in wind blasts of severe intensity. Changes in local geology have also resulted in the failure of hydraulic fracturing treatment to bring down the roof on
129 many occasions in LW4A panel. The roof subsequently fails after the longwall is retreated a further distance after the hydraulic fracturing had been attempted.
For the purposes of analysis of the wind blast data, the roof falls at Moonee can be broadly characterised as: falls due to mining (Class M), hydraulic fracture induced falls (Class F) and falls induced by hydraulic fracturing and subsequent mining (Class C). This section analyses the impact of hydraulic fracturing on wind blasts at Moonee Colliery and the differentiation of the stated three classes of roof falls.
5.5.1 Hydraulic Fracturing Procedure
The first underground trial of hydraulic fracturing involved drilling an injection hole from the maingate side of the panel out to a point approximately 12 metres above the roof in the centre of the Longwall 3 Panel. An injection pipe was grouted into this hole so that the inner last few metres were left open. The 55 metres of standing goaf measuring approximately 6335 square metres was induced to cave following a hydraulic fracture treatment lasting approximately 2 hours on June 30, 1999. Approximately 40,500 litres of water were injected, mainly at a supply pressure of about 1.8 MPa.
Subsequent hydraulic fracturing involved vertically drilled holes from the centre of the face connected to a high pressure hose trailed out into the goaf as the face advanced. Figures 5.27 and 5.28 illustrate the hydraulic fracturing sequence and layout for LW3 and LW4 respectively. A brief outline of the procedure as implemented in LW3 is as follows:
1. Mine for 20 metres retreat;
2. Drill two vertical holes (or three holes) on the face through the immediate roof to a depth of 7 to 9 metres into the conglomerate roof;
3. Retreat a further 26 metres;
4. Commence hydraulic fracturing, from a “safe haven’, by pressurising the upper 1.4 metres of both holes using the shearer water pump and ordinary mine water;
130 Figure 5.27 Hydraulic fracturing sequence in LW3 panel, Moonee Colliery
131 CENTRE HOLE NIG 8- TG HOLES
1 0 8m 1 0 8m
Jtc PRESSURE ;* anO
installation arrangement FULL UPHOLE SECTION
Figure 5.28 Hydraulic fracturing sequence in LW4 panel, Moonee Colliery
132 5. Maintain pressure (at several hundred kPa) until a fracture starts to propagate as evidenced by a fall in pressure;
6. Continue pumping (at several hundred L/s) until the goaf falls.
In the event of a roof fall not occurring after pumping for a specified time period (usually 2 hrs), drill two further vertical holes on the face, retreat a further 26 metres and recommence hydraulic fracturing.
The mine has trialled with 2 to 3 sets of vertical holes located at roof supports RS 16- 17, RS 28-29 and RS 35-36 respectively as shown in Figures 5.27 and 5.28. In some instances, a couple of injection holes were drilled only at the face centre but to varying roof heights. The holes are pressurised simultaneously. In Appendix E, a sample hydraulic fracturing report for hydraulic fracture no. 3 in LW3 panel of Moonee Colliery is given along with figures showing the hydraulic fracturing pressures for the first two underground hydraulic fracturing trials.
5.5.2 Impact of Hydraulic Fracturing on the Roof Failure Mechanism
Roof failure mechanism for narrow width longwall panels under massive roof has been discussed in chapter 2 of this thesis. Earlier, in section 5.4, Class M falls has been discussed. The basic mechanism of rock failure in the goaf at Moonee Colliery involves the horizontal stresses failing the conglomerate strata in horizontal shear when the vertical confinement is removed by mining. Pore water pressures within the strata, due to ground water, also help to further reduce the shear strength of the conglomerate. To understand the roof failure mechanism, it is essential to know the relative magnitude of stresses acting on the roof strata.
5.5.2.1 Relative magnitude of stresses acting on the roof strata
At Moonee Colliery the relative magnitude of stresses acting on the failing roof strata are:
1. The horizontal stresses of the order of 8-12 Mpa (Section 5.2.2).
2. The vertical stresses of the order of 3-4 MPa prior to mining but which are
133 significantly reduced over the goaf area by mining even becoming tensile in some parts of the rock mass due to destressing.
3. The weight of the unsupported conglomerate strata that eventually fails is of the order of 1600-1900 tonnes for each linear metre of longwall retreat. The force that the rock exerts as a stress on the failure surface is only of the order of 0.15 - 0.23 MPa, assuming thickness of the failing strata to be around 6-9 metres and density, 2.70 tonnes per cubic metre.
4. The pore water pressures in the strata are capable of exerting pressures up to 1- 1.5 MPa, based on the strata depth 100-150 metres below the phreatic surface. The pore pressures are very high compared to the weight of failing roof element.
5. Hydraulic fracturing is able to generate pressures of more than 10 MPa if required, but much lower pressures (0.5-5 MPa) have been sufficient to initiate caving.
5.5.2.2 Hydraulic fracturing and roof failure process
At Moonee, the insitu stress measurements and computer based modelling conducted by the mine reveal that minimum stress in the conglomerate above the goaf follows a trajectory that is approximately the same shape as the stable geometry of the fallen goaf. The boreholes are drilled to a height in the conglomerate roof so that water is injected at a point close to an arch on the final stable geometry. This produces two effects. First, a more or less horizontal fracture plane is induced along a trajectory that approximates the final stable geometry. Second, fluid pressure in the fracture exerts a downward force on the uncaved strata. The pore pressure is sufficiently higher in magnitude compared to weight of the strata which falls.
The fracture, which orients perpendicular to the least principal stress, is thought to be essentially circular in plan, expanding radially outward from the injection point during the hydraulic fracture treatment. With the pump pressures and flow rates used so far at Moonee, indications are that the fracture has grown to be about 60-70 metres in diameter by the time the cave is initiated. The radially expanding fracture has the effect
134 of artificially removing the tensile strength of the rock over a larger and larger area while at the same time pressing down on top of the uncaved strata. Continued growth of fracture inevitably leads to strata instability and subsequent caving. Once failure of the roof strata occurs and a caving event is initiated, the weight of rock becomes significant as it falls to the ground, for generating a wind blast as per the analytical model (McPherson, 1995b) discussed in section 4.4.2 in Chapter 4.
5.5.3 Wind blast events associated with Hydraulic fracturing
Hydraulic fracturing of roof commenced after goaf fall number 2 in Longwall Panel 3. Out of a total of 27 goaf falls in Longwall Panel 3, wind blast events were monitored for up to goaf fall number 21. Wind blasts of sufficient intensity were associated with each goaf fall monitored. Hydraulic fracturing injection numbers 3 and 8 resulted in sequential goaf falls with 2 wind blast events associated with each.
The Longwall Panel 4A witnessed more Class C falls i.e. falls induced by hydraulic fracturing and subsequent mining. Hydraulic fracture number 6 resulted in sequential falls and 3 wind blast events associated with the goaf falls. Hydraulic fracturing no. 9 also resulted in two falls 9a and 9 associated with wind blasts. Out of the 21 goaf falls which resulted in a total of 24 significant wind blast events including the sequential ones, 10 were class F falls and 1 was Class M fall. Fall number 3, 5, 11 and 13 were not monitored due to the Wind Blast Logger memory being full or power to the Wind Blast Logger being ‘off.’
Longwall Panel 4B witnessed 15 roof falls associated with wind blasts. Hydraulic fracturing treatment no 4. resulted in 2 falls, fall no. 4 on 13 September, 2000 and one additional unnumbered fall recorded on 14 September, 2000. Towards the end of the panel on 7 November, 2000 there were two Class M falls numbered 15A and 15B respectively. The falls were reported to extend high.
Tables 5.11, 5.12 and 5.13 present the results of microseismic monitoring and details of roof falls for LW3, LW4a and LW4b panels, respectively. The seismic warning time has been calculated from difference of the fall time and the seismic alarm time as reported in wind blast notes for each roof fall. The table column ‘advance after hydro-
135 Table 5.11
Monitoring of Goaf Falls at Longwall Panel no. 3, Moonee Colliery (adapted from wind blast notes, Moonee Colliery)
Date Goaf Fall Fall Face Goaf Fall Audible Seismic Advance
Fall Class Area Location Hanging Time Warning Warning After Hydrofrac No. m2 m m hr:min:s time Time m
17-Apr-99 0A M - 1676 02:08:10 na 17-Apr-99 1 M 26640 1676 283.4 02:08:32 3 min 3 min 45 s na 30-Apr-99 2 M 10360 1548 08:56:45 na 30-Jun-99 3 F 6335 1493 55 15:33:46 - 36 min (trend) - 15-Jul-99 4 F 7080 1408 75 10:38:58 1 min 43 s - 5-Aug-99 5 C 6950 1301 68 03:48:47 16 min 2 s - 5-Aug-99 5A F - 03:49:16 - 11 -Aug-99 6 M 4770 1254 48.5 05:38:13 15s na 17-Aug-99 7 F 3940 1214 40 11:05:05 3 s - 21-Aug-99 8 C 4310 1172 42 09:44:25 15 min <2 1-Sep-99 9 F 2780 1073 46 10:50:38 17s - 1 -Sep-99 9A F - 10:52:57 - 4-Sep-99 10 F 6230 1027.5 46(MG-36 RS) 09:36:24 18 * 97(36 RS-TG)
13-Sep-99 11 F 6270 960.4 67 19:42:01 23 s -
20-Sep-99 12 F 4370 916 17:54 - -
6-Oct-99 13 F 5250 834 44 09:29:19 1min 27s -
12-Oct-99 14 F 4900 788 46 11:39:54 3 min 57 s -
19-Oct-99 15 C 2700 731 9 + large 04:51:01 50 s 9 Triangle
21-Oct-99 16 F 4460 685 46 09:26:24 53 s -
26-Oct-99 17 F 4630 639 46 22:11:51 7 min 15 s -
29-Oct-99 18 F 4552 593 46 23:05:24 9 s -
5-Nov-99 19 F 3310 544 39 03:55:20 - 24 s -
5-NOV-99 20 C 1820 542 10 + MG 102031 - - 2 Triangle 12-Nov-99 21 F 8510 453.8 84.6 21:23:43 - 12 min 25 s - 20-NOV-99 22 F 7803 360 88 02:50:39 - 56 s - 24-Nov-99 23 F 5410 306 54 16:23:40 - 18s - 8-Dec-99 24 F 4653 168 48 11:31:19 - 40s -
13-Dec-99 25 F 4773 121 49 05:02:03 - 6 min 9 s - 23-Dec-99 26 C 5040 72 48 14:43:10 30 s 1 min 19 s <2
5-Jan-OO 27 F 5579 14 56 02:42:49 - 1 min 41 s -
136 Table 5.12
Monitoring of Goaf Falls at Longwall Panel no. 4a, Moonee Colliery (adapted from wind blast notes, Moonee Colliery)
Date Goaf Fall Fall Face Goaf Fall Audible Seismic Advance
Fall Class Area Location Hanging Time Warning Warning After
Hydro-frac
No. m2 in m hr:min:s time time m
15-Feb-00 1 C 12600 1318.5 127 19:52:19 <1 min 9 s 31.5
23-Feb-OO 2 c 5200 1262 56 00:30:30 1 hr 38 s 1 hr 10 min 10
1-Mar-00 3 c 5141 1198.5 - 11:40:40 - 1 min 1 s 11.5
6-Mar-OO 4 c 4710 1144.0 48.5 13:32:27 1 hr 2 min 1 hr 2 min 4 43s
14-Mar-00 5 c 6165 1066.6 64 18:25:14 1 hr 33 min 1 hr 33 min 23 22-Mar-00 6a c 7150 993 73 9:10:43 1 min 9 s 1 min 9 s 13
22-Mar-OO 6b 993 9:13:19
25-Mar-00 7 F 5140 943 50 3:22:29 9 min 4 s 6 min 51 s -
29-Mar-OO 8 F 4428 893 50 21:22:41 1 min
3-Aor-OO 9a F 450 841.5 47 11:26:52 27 s 27 s
5-Aor-OO 9 C 6730 816 2 74 3:18:33 6 min 33 s 4 min 22 s 25
11-A Dr-00 10 C/M 2570 750 13:45:42
21-25ADr- 11 C 7859 631.4 107 11.5
11-Mav-00 12 M 11265 531.3 144 8:13:14 2 min 49 s 2 min 48 s
20-Mav-00 13 F 4989 480 49.8 02:48:20 2 min 3 s
26-Mav-00 14 F 3440 406 61 19:31:31 3 min 3 s
31-Mav-00 15 F 6600 364.5 61.2 12:53:58
9-Jun-00 16 F 5400 303 63 11:34:59 1 min 2 s
30-Jun-00 17 F 5323 174 64.5 11:22:47 3 min 45 s
5-Jul-00 18 F 5022 125 49 23:35:47 14 min
11-Jul-00 19 C 5962 66 14:42:15 12min 16sec 5
21-Jul-00 20 F 2780 14.2 50 01:24:49
21-Jul-00 21 C 3123 6 _ 12:03:35 . 8
137 Table 5.13 Monitoring of Goaf Falls at Longwall Panel no. 4b, Moonee Colliery (adapted from wind blast notes, Moonee Colliery)
Date Goaf Fall Fall Face Goaf Fall Audible Seismic Advance
Fall Class Area Location Hanging Time Warning Warning After Hydro-frac
No. sqm m m hr:min:s time time m
20-Aug-00 1 C 9620 874 102 23:21:37 hrs 4 min 31 s 7
29-Aug-00 2 M 3875 804 60 03:36:12 25 s 25 s
3-Sep-00 3 F 4493 771 52 12:52:00 4 min 15 s
13-Sep-OO 4 F 6860 665.6 101 19:03:18 3 min 24 s
15-Sep-00 5 C 2350 649 48 04:16 12s 12s -
21-Sep-00 6 F 5410 550 100 12:20:31 9s
29-Sep-00 7 F 5275 497.7 96 03:53:36 33 s
2-Oct-OO 8 C 1691 476 40+ 12:20 0s 1
4-Oct-OO 9 C 6183 436 86 21:00:25 hrs 9s -
11 -Oct-OO 10 F 5560 376 58 05:46:00 33 s
15-Oct-OO 11 F 4690 308 54.6 13:44:00 19 min 8 s
18-Oct-OO 12 F 5900 240.2 68 4 19:29:27 12s
25-Oct-OO 13 F 5640 185 55 02:11 9s
30-0ct-00 14 C 5704 125 58 14:59:34 9s 9s -
7-Nov-00 15a M 8920 38.6 75.4 13:52:09 32 s 32 s
7-Nov-OO 15b M 8920 38.6 75.4 13:52:26 -
138 ffac’ identifies the Class C falls and states the distance the face advanced prior to the fall subsequent to the hydraulic fracturing treatment which failed to bring down the roof at that time.
Longwall Panel 4A witnessed wind blasts of maximum overpressure intensity of 24.3 kPa for a Class C fall and 15.2 kPa for a Class F fall. Longwall 3 had maximum overpressure intensity of 29.5 kPa for a Class F fall. Longwall 4B witnessed a maximum overpressure intensity of 36.5 kPa for the Class M fall (Fall no. 15B), 27.7 kPa for a Class F fall and 14 kPa for a Class C fall. Based on visual inspections of goaf, the mine personnel also recorded height of goaf falls as given in wind blast notes for one event in Appendix D. Goaf falls to greater heights have resulted in more intense wind blasts.
5.5.3.1 LW4a panel first fall
The first fall in LW4a panel, occurred on 15 February, 2000 at 7:52 pm, after the face had retreated 127 metres. The goaf fall area was approximately 12,600 square metres as shown in Figure 5.26. This followed an underground hydraulic fracturing treatment no. 1, which was successful in generating a significant fracture within the goaf For the hydraulic fracture, one vertical hole was drilled at mid face after 50 metres retreat, nine metres into the conglomerate. The hole was initially pressurised using the solsenic pump and followed by the shearer water pump. The hydraulic fracturing treatment was started 3 metres short of the planned panel chainage due to high seismic activity.
The hydraulic fracture treatment proceeded for 4 hours and was successful in generating a significant fracture within the goaf but failed to bring down the roof. Production was recommenced after a prolonged period of low seismic activity. Production continued for a further 31.5 metres before the goaf fell naturally.
The observed caving of the goaf, as reported in wind blast notes for this fall, was lenticular in shape starting from the back of the supports with a triangular shaped part of goaf from 23 Roof Support (RS) back 20 metres left hanging towards the maingate.
139 5.5.3.2 LW4b panel first fall
The first fall in LW4b panel, occurred on 20 August, 2000 at 11:21 pm, after the face had retreated 102 metres. The goaf fall area was approximately 9,620 square metres as shown in Figure 5.26. This was preceded by underground hydraulic fracturing treatment no. 1 that was unsuccessful in caving the goaf. The hydraulic fracturing treatment took 6 hours and 31 minutes and used 4 vertical holes. Hole 1 at 27 -28 Roof Support (946 metres location), Holes 2 and 3 at 13 -14 Roof Support and 41-42 Roof Support respectively (931 metres location). Hole 4 was drilled at 27-28 Roof Support (916 metres location). The holes were drilled into the conglomerate to 10.8 metres.
The face advanced 7 metres after the hydraulic fracturing treatment, before the goaf fell. A small triangular area was standing at the back of the goaf along the maingate side of LW4b face
5.5.4 Comparative Analysis of Overpressure and Wind Velocity Data
This section compares some of the interrelationships between the parameters listed below for the three classes of fall witnessed in the longwall panels at Moonee Colliery.
Peak overpressure;
Impulse;
Maximum rate of rise of overpressure;
Peak wind blast velocity;
Maximum excursion;
Maximum rate of rise of wind blast velocity;
Distance from fall; and
Roof fall area.
5.5.4.1 Characteristics of the overpressure time history
Figure 5.29 illustrates the relationship between impulse and peak overpressure. It utilises data from the subset obtained from pod no. 2, which was mounted on the DCB,
140 in the maingate at a fixed distance of 11 metres from the face.
The relationship is non-linear and a second-degree polynomial curve has been fitted to the data. As indicated by the R square values in the diagram, the coefficients of correlation are high for all three classes of falls.
The implication of the non linearity is that for all the three classes of falls, as wind blasts become more intense, the impulse increases more than the overpressure, i.e. the duration of the positive (compression) phase becomes longer.
Figure 5.30 illustrates the relationship between impulse and peak overpressure. It utilises data from the subset obtained from Pod 4 which was mounted on an inverted tripod secured to the roof of the maingate at and outbye location. Once again the relationship is non-linear and a second-degree polynomial curve has been fitted to the data. The coefficients of correlation are high for Class M and C falls but somewhat less for Class F. The figure illustrates similar trends for Class M and Class F falls over the range of values.
Figure 5.31 and 5.32 illustrates the relationship between the maximum rate of rise of overpressure and peak overpressure at Pod 2 and Pod 4 locations respectively. The relationship is again non-linear with close similarity for Class M and Class F falls. The coefficients of correlation are reasonably high for the above fall classes, however there is insufficient data and more scatter for Class C falls near to the face, resulting in poor correlation (Figure 5.31). This could also be due to the pod being located adjacent or near to the cut-throughs at the time of the falls. The trends illustrate that as wind blasts become more intense, the maximum rate of rise of overpressure increases non-linearly and a decreasing rate compared to peak overpressure.
S.5.4.2 Characteristics of wind velocity time history
Figure 5.33 illustrates the relationship between maximum excursion and peak wind blast velocity. It utilises data from the subset obtained for Pod 4 which has been mounted on an inverted tripod secured to the roof of the maingate at an outbye location. A second-degree polynomial curve has been fitted to the data. The coefficient of correlation for Class M is high while that for Class F and Class C
141 Pod 2 on DCB in maingate O Class M (LW1 - LW4b panels) □ Class F A Class C ------Poly (Class M) (all data) - —Poly (Class F) (all data) - - Poly. (Class C) (all data) 0.030X2 + 1,03x R2 = 0.91
y = 0.0046X2 + 1 28x R2 = 0.83
Peak Overpressure (kPa)
Figure 5.29 Relationship between impulse and peak overpressure for all fall classes, Moonee Colliery
Pod 4 O Class M in maingate (outbye) □ Class F (LW1 - LW4b panels) A Class C ------Poly. (Class M) (all data) — —Poly (Class F) (all data) - - - Poly. (Class C) (all data)
y = 0.0048X2 + 1,41x R2 = 0.78 ;‘A'" □
Peak Overpressure (kPa)
Figure 5.30 Relationship between impulse and peak overpressure for all fall classes, Moonee Colliery 142 Pod 2 O Class M on DCB in maingate □ Class F (LW1 - LW4b panels) A Class C ------Poly. (Class M) (all data) - —Poly. (Class F) (all data) - - Poly (Class C) (all data)
y = -0.017X2 + 1 46x R2 = 0.84
-0.037X2 + 2.00x □1 O R2 = 0.47
> $■'
Peak Overpressure (kPa)
Figure 5.31 Relationship between max. rate of rise of overpressure and peak overpressure for all fall classes, Moonee Colliery
Pod 4 O Class M in maingate (outbye) (LW1 - LW4b panels)
------Poly. (Class M) (all data)
- —Poly (Class F) (all data) - - Poly (Class C) (all data)
y = -0.024X2 + 1,49x R2 = 0.84
y = -0.056x2 + 1 86x R2 = 0.85
y = -0.083X2 + 2.23x R2 = 0.73
Peak Overpressure (kPa)
Figure 5.32 Relationship between max. rate of rise of overpressure and peak overpressure for all fall classes, Moonee Colliery 143 C falls shows more scatter and poor correlation. The trends are however similar for all class of falls. The poor correlation for Class F and Class C falls could be attributed to variation in the hydraulic fracturing parameters related to height of fracture induction due to variation in drilling heights, changes in the cutting height and also the effect of some falls being sequential.
The trends of non linearity indicate that as wind blasts become more intense, the maximum excursion increases more than does the peak wind blast velocity, i.e. the duration of the positive (wind blast) phase becomes longer. This is also suggested by visual inspection of the wind velocity time histories.
The above relationship is similarly defined at Pod 2 location as illustrated in Figure 5.34. The poor correlation for Class C and F could also be due to the location of the pod being fixed in relation to the face and occasionally the pod being adjacent to a cut through. However, the trends show close similarity for Class M and Class F falls.
Figure 5.35 illustrated the relationship between maximum rate of rise of wind blast velocity with peak wind blast velocity. It utilises data from the subset obtained from Pod 2. In this case a linear relationship has been fitted to the data. The coefficient of correlation is reasonably high for Class M. Class C and Class C falls data exhibits poor correlation. The reasons are Class C and Class F falls have been more of sequential roof falls.
The implication of linearity is analogous to that for overpressure, i.e. as wind blasts become more intense, the maximum rate of rise of wind blast velocity increases in direct proportion to the peak wind blast velocity.
5.5.4.3 Relationships between overpressure and wind velocity time histories
Figure 5.36 illustrates the relationship between peak wind blast velocity and peak overpressure. It utilises the data set from Pod 2. Pod 2 location has been constant for all fall classes. It will be observed that though the parameters are evidently positively correlated their relationship is far from clear. The Class M Falls for the range selected
144 200 -i Pod 4 in maingate (outbye) O Class M 180 - (LW1 - LW4b panels) □ Class F A Class C 160 ------Poly. (Class M) (all data) — —Poly (Class F) (all data) - - - Poly. (Class C) (all data) 140 -
*100 ■
y = 0.018X2 + 0.019x R2 = 0.66
o.oio^ + o.osx R2 = 0.68
I 50 6 Peak Velocity (m/s)
Figure 5.33 Relationship between maximum excursion and peak wind velocity for all fall classes, Moonee Colliery
250 Pod 2 on DCB in maingate O Class M (LW1 - LW4b panels) □ Class F A Class C ------Poly. (Class M) (all data) ------Poly (Class F) (all data) - - Poly (Class C) (all data)
O □ 0.0048X2 + 0.64x R2 = 0.83
y = 0.0027X2 + 0.88x R2 = 0.65
y = -0.0023X2 + 0.99x R2 = 0.65
80 Peak Velocity (m/s)
Figure 5.34 Relationship between maximum excursion and peak wind velocity for all fad classes, Moonee Colliery 145 Pod 2 on DCB in maingate O Class M (LW1 - LW4b panels) □ Class F A Class C 500 • ■■ - "Linear (Class M) (all data) — — Linear (Class F) (all data) - — Linear (Class C) (all data)
~ 400 -
y = 3.05x R2 = 0.50 3) 300 •
200 -
y = 1 89x R2 = 0.32
100 -
y = 1,27x R2 = 0.56
Peak Velocity (m/s)
Figure 5.35 Relationship between maximum rate of rise of velocity and peak velocity, for all fall classes, Moonee Colliery
160 -1 Pod 2 O Class M on DCB in maingate □ Class F (LW1 - LW4b panels) 140 - A Class C ------Linear (Class M) (all data) - - Linear (Class C) (all data) 120 - - —Linear (Class F) (all data)
y = 3.98x R2 = 0.59
y = 4.91x R2 = 0.33
Peak Overpressure (kPa)
Figure 5.36 Relationship between peak velocity and peak overpressure for all fall classes, Moonee Colliery 146 shows a tentative linear relationship with the following equation:
Peak wind velocity (in m/s) = 3.98 x peak overpressure (in kPa) (5.5)
Class F falls show even more scatter and the tentative equation is given by:
Peak wind velocity (in m/s) = 4.91 x peak overpressure (in kPa) (5.6)
The equation for Class C falls is:
Peak wind velocity (in m/s) = 4.92 x peak overpressure (in kPa) (5.7)
The values show marked deviation from the Rankine-Hugonoit relationship given by equation 5.2, relevant for the lower range of overpressures. This shows the impact of the physical displacement of the goaf air due to the falling roof and its expulsion to the adjoining roadways. The geometry of the roadway at the pod location also has an impact on the peak wind velocity with higher values for lower cross-sectional area and vice versa. Consequently, when the pods are sometimes located near or adjacent to cut- throughs, as the face retreats, lower values are recorded. The mining equipment in the roadways also constrict the cross-sectional area.
5.5.4.4 Relationship between wind blast intensity and overall roof fall area
It is evident from the inspection of Figure 5.22 that the relationship between impulse and roof fall is linear. Consequently, straight lines have been fitted to the data in Figure 5.37, which illustrates the relationship between impulse and roof fall area. This means that a single number description can be used to quantify the relationship for each class of roof fall as indicated in Table 5.14 in which impulse is normalised with respect to roof fall area.
A similar analysis is undertaken for maximum excursion in Figure 5.38. The normalised values for each class of the roof fall are also given in Table 5.14. Figure 5.38 utilises the complete data subset obtained from Pod 2. A linear relationship has been fitted to the data. However, the positive correlations are again quite poor.
From the data in Figure 5.37, it is apparent that the parameters are positively related
147 Pod 2 on DCB in maingate O Class M (LW1 - LW4b panels) □ Class F A Class C ------Linear (Class M) (all data) — —Linear (Class F) (all data) - - - Linear (Class C) (all data)
y = 0.0025X E 40- R2 = 0.54
10000 20000 30000 40000 Goaf Fall Area (m2)
Figure 5.37 Relationship between impulse and goaf fall area for all fall classes, Moonee Colliery
10000 20000 30000 40000 Goaf Fall Area (mz)
Figure 5.38 Relationship between maximum excursion and goaf fall area for all fall classes, Moonee Colliery 148 for Class M falls within the selected range. Data from Class F and Class C falls show scatter. The data reflects the trend for linearity within the lower range of goaf fall area. However the positive correlations are quite poor.
A particular problem arises in performing a similar analysis for peak overpressure and peak velocity as the relationship will be seen from Figures 5.20 and 5.23 to be non linear. However by restricting the analysis to fall areas of less than 10,000 square metres, linearity may be assumed. The normalised values of peak overpressure and peak velocity, over this restricted range of fall areas, are also indicated in Table 5.14.
Figure 5.39 illustrates the complete data set for all fall classes monitored at Moonee Colliery, longwall panels. It is apparent that there is an upper bound for the overpressure values. This trend is apparent with falls areas exceeding 10,000 square meters for Class M falls. For Class F and Class C falls this is apparent for even smaller areas. This suggests that sequential and progressive falls, which resulted in lower overpressure values, were more frequent after hydraulic fracturing was introduced, for larger spans. This trend is more obvious for goaf fall areas exceeding 5,000 square metres.
Figure 5.40 illustrates the relationship between peak overpressure and goaf fall area for falls less than 10,000 square metres in area. As discussed above, the linear relationship is now better defined for Class M falls. For Class F and Class C Falls, the data reflects the trend for linearity within the lower range of goaf fall area. A linear relationship is fitted for both these classes of falls. However, the positive correlations are quite poor.
Figure 5.41 illustrates the relationship between peak wind velocity and goaf fall area for falls less than 10,000 square metres in area. The trends are similar to that observed for overpressure.
5.5.5 Conclusions Regarding the Impact of Hydraulic Fracturing
Tables 5.14, illustrates the normalised wind blast parameters for the 3 different classes of falls. The figures in bold represent values with higher positive correlation. Other values are based on tentative relationships with weak correlation. As discussed in
149 Figure 5.39 Relationship between peak overpressure and goaf fall area for all falls, Moonee Colliery
Figure 5.40 Relationship between peak overpressure and goaf fall area for all fall classes with area less than 10,000 square metres, Moonee Colliery 150 Table 5.14
Comparison of wind blast parameters with fall class, Moonee Colliery, LW1 - LW 4b panels
Pod 2 (on DCB, near face, maingate B Hdg) Parameter Class of Fall MFC
Normalised Peak Overpressure 2.4 2.5 1.8 (kPa/1000 m2) Utilises data for all falls < 10,000 m2
Normalised Impulse (kPa.s/1000 m2) 2.5 3.5 1.7 Utilises all data
Normalised Peak Velocity (m/s/1000 m2) 11 13 9.3 Utilises data for all falls < 10,000 m2
Normalised Maximum Excursion (in/1000 m2) 6.2 15 7.3 Utilises all data
section 5.4.2.5, the wind blast time histories as well as microseismic evidence suggests that on many occasions the roof failed in a sequential or progressive mode rather than the assumed monolithic mode, leading to multiple events being registered in the Wind Blast Data Logger. However, the overall area has of necessity, been assigned to the event with the highest recorded intensity and the subsequent events not considered in the analysis. This also contributes to the weak correlations obtained for the normalised Figure 5.39 wind blast parameters.
Based on the trends observed between the wind blast parameters and from wind blast time histories, some conclusions can be derived for the three classes of falls observed at Moonee Colliery:
Failure process is always sudden for the conglomerate roof, irrespective of the class of
151 roof fall i.e. Class M, hydraulic fracture induced (Class F) or falls induced by hydraulic fracturing and subsequent mining (Class C).
Wind blasts are invariably associated with roof falls of all the above classes at Moonee Colliery.
In longwall panel numbers 3, 4a and 4b, Class F falls dominated and showed more deviations from the Rankine - Hugoniot relationship between peak velocity versus peak over pressure when compared to longwall panel no. 1 and no. 2 falls. Also at the outbye Pod 4 location, the deviation was less compared to Pod 2 location near the face. This illustrates the impact of expulsion of the goaf air, by the falling roof, on the peak velocity parameter and the effect of panel geometry at pod locations.
Hydraulic fracturing induced caving achieved caving “on demand” by initiating fracture near to the final stable goaf geometry.
Success of the hydraulic fracturing treatment depends on the following:
1. State of stresses acting in the roof strata;
2. Fracturing achieving sufficient dimensions with proper pump pressures and flow rates maintained throughout the treatment;
3. Sufficient roof overhang so that instability caused by hydraulic fracturing initiates a failure process that is inherently unstable.
Compared to the mining induced falls, hydraulic fracturing introduces the high fluid pressures, which exert a downward force on the uncaved strata along with artificially removing the tensile strength of the rock over an increasing area.
Monitoring results suggest that sequential and progressive falls, which resulted in lower peak overpressure values, were more frequent for the larger spans after hydraulic fracturing was introduced. This trend is more obvious for goaf fall areas exceeding 5,000 square metres as illustrated in Figure 5.39.
Normalised wind blast parameters of overpressure, impulse and peak velocity for Class C falls are apparently lower than Class F or Class M falls. Class F falls have
152 slightly higher values of normalised wind blast parameters compared to Type M falls and consequently, are more intense.
The immediate impact of hydraulic fracturing on wind blasts is that the intensity of the event can be reduced based on the area of roof induced to fail. Moreover, the situation is not worsened in case of hydraulic fracturing failing to induce a fall immediately, as Class C Falls are less intense.
5.6 EMPIRICAL RELATIONSHIP BETWEEN WIND BLAST PARAMETER AND SCALED DISTANCE
The data obtained from wind blast monitoring in all the longwall panels at Moonee Colliery was combined and analysed irrespective of the class and mode of roof failure. This was done so as to identify any empirical wind blast scaling laws applicable to the particular geological and geometric environment.
The scaled distance, based on cube root scaling, is given by the ratio of distance (R) and cube root of energy (E). In case of wind blasts the energy is proportional to the goaf volume for a given thickness of roof element and discussed in section 2.4.4, Chapter 2 and consequently with the goaf area for a given extraction thickness. The distance has accordingly been scaled by cube root of the goaf area. The distance is calculated from the difference between face location and the outbye pod location.
A tentative relationship for wind blasts has been derived between peak overpressure and scaled distance based on the data from all the goaf falls below 10,000 square metres in area. Also, falls involving maingate or tailgate triangles have been excluded. The relationship is inverse and Figures 5.42 and 5.42 plotted in a logarithmic scale, illustrate the relationship at Pod 4 location, outbye of face, in maingate or travelling road and at Pod 2 location, a fixed distance of 11 metres from the face, respectively.
The empirical equation for Pod 4 location, outbye of face is:
Peak overpressure (in kPa) = 12.23 x (Scaled distance) '028 (5.8)
The empirical equation for Pod 2 location, near to face is:
153 Figure 5.41 Relationship between peak velocity and goaf fall area for all fall classes with area less than 10,000 square metres, Moonee Colliery
Pod 4 in maingate (outbye) x panels LW1 - LW4b (selected data)
----- Power ( panels LW1 - LW4b (selected data))
y= 12.23x
£ 10.0 -
Scaled Distance (m/A1/3)
Figure 5.42 Relationship between peak overpressure and scaled distance at Pod4 location in maingates, Moonee Colliery 154 Peak overpressure (in kPa) = 4.91 x (Scaled distance) -2.04 (5.9)
The analysis assumes a constant thickness of roof element and constant height of fall or constant extraction thickness. Cube root scaling of distance gave a comparatively better fit for the data compared to square root scaling. The correlation coefficients are poor as data shows deviations from scaling laws. The deviations are apparently due to variation in thickness of the falling roof element, variation in thickness of the immediate readily cavable roof and consequently in goaf volume, variation in extraction thickness and also due to mechanism of the roof fall and its interaction with air below it. For larger fall areas the peak overpressures tend to plateau as discussed in section 5.4.2.5. The goaf falls have often been sequential falls and associated with multiple events as revealed by wind blast time histories and microseismic monitoring. However the goaf areas involved with such sequential falls cannot be determined and consequently the area of fall is attributed to the largest event. The expulsion of air resulting from small goaf falls is much more efficient and this has also been observed especially from the falls of the triangular areas adjacent to maingate and tailgate
Figure 5.44 illustrates the attenuation of peak overpressure with distance along the maingates of all five longwall panels. It utilises data from the subset obtained from Pod 2, and also from Pod 4 at an outbye location. For panel 4b, the data is obtained from Pod 1 and Pod 4 located outbye to face in the travelling road, maingate B heading.
Although these parameters are evidently negatively related, except for a few occasions, when higher values were recorded by the outbye pods, the relationship is not well defined. The scatter in the data is due to the fact that the pods are often situated in non ideal locations, where the air flow is not parallel to the pods and there are obstacles to the air flow, which increase turbulence, also occasionally the inbye pod being located in an intersection.
An exponential decay curve has been tentatively fitted to the data with the following relationship:
Overpressure quotient = e -0.0013 x distance (5.10)
155 x panels LW1 - LW4b (selected Pod 2 data) on DCB in maingate (11 metres from face)
----- Power (panels LW1 - LW4b (selected data))
y = 4.91 x‘2 04 R2 = 0.36
Scaled Distance (m/A1'3)
Figure 5.43 Relationship between peak overpressure and scaled distance at Pod2 location in maingates, Moonee Colliery
Pod 2(near face)& Pod 4(outbye) in maingates
x LW1 - LW4b (all data)
| 0.9------Expon. (LW1 - LW4b (all data)) — 0.8 -
S. 0.7 - -0.0013x
3 0.6- R2 = .01 o. 0.5 -
O 0.4 -
Distance (m)
Figure 5.44 Attenuation of peak overpressure with distance along maingates, longwall panels, Moonee Colliery 156 As discussed in section 4.1.5 and 4.1.6 of Chapter 4, peak overpressures for air blasts due to unconfined explosive charges also show deviations from scaling laws in an underground mine.
157 CHAPTER 6
COMPARATIVE STUDY OF WIND BLASTS IN NEWSTAN AND MOONEE COLLIERIES
6.1 INTRODUCTION
Newstan Colliery faced wind blast hazards during the extraction of the 95 metres wide longwall 6 panel (LW6) under massive conglomerate strata, in August 1995. Subsequently, the Wind Blast Monitoring System was deployed and a total of 23 wind blast events recorded during the mining of Longwall panels 7, 8 and 9 between 1995 and 1997. Of these 8 were significant i.e. of sufficient intensity to pose a risk of personal injury or of damage to the mine ventilation system (Simpson, Hebblewhite & Fowler, 1997).
Wind blast activity at Newstan was found to occur exclusively where the conglomerate channel was thick enough to bridge the goaf, and where it occurred within 7 metres of the working roof or approximately twice the extraction height (Creech, 1996) as discussed briefly in Section 2.5.4. Face lengths for Longwall panels 8 and 9, were increased from 95 metres to 127 metres between gateroad centrelines. However, for part of the extraction of above panels, during which the stated geological conditions persisted, wind blasts remained a hazard.
The wind blast data set for Newstan Colliery has been reprocessed, using the WaveMetrics Igor Pro version 3.13 macros developed for the analysis of Moonee Colliery data, using standardised smoothing algorithms for the overpressure and wind velocity time histories and necessary corrections. The previous work done at Newstan Colliery by Fowler and Torabi (1997) had not analysed the overpressure time histories for each event. The relationship between the wind blast parameters defined in this section extends the previous work. This chapter also compares the two data sets of Moonee and Newstan Colliery and analyses the significance of the geological and operational differences in the two mines on wind blast time histories and parameters.
158 6.2 GEOLOGICAL AND TECTONIC SETTING AT NEWSTAN COLLIERY
Newstan Colliery is located approximately 20 km south west of Newcastle, New South Wales, within the Lake Macquaire district of the Newcastle Coalfield. It currently mines the Young Wallsend / West Borehole seams. The four longwall panels, where the series of wind blast events occurred between 1995 and 1997, mined the West Borehole seam towards the south-western side of the working as shown in Figure 6.1, with the panel centrelines oriented approximately south-east. Panel widths were 95 metres for LW6 and LW7 panels and 127 metres for LW8 and LW9 panels while chain pillars had widths of 30 metres (between riblines) and the working height was 3.3 metres. The depth of overburden varied from 117 to 240 metres and consequently the width to depth ratio for the panels ranged from 0.4:1.0 to 0.75: 1.0. The panel lengths were 1100 metres.
6.2.1 Regional and longwall panel geology
A generalised stratigraphic sequence of the basal sub-group of the Newcastle Coal Measures, the Lambton, together with the immediately overlying Kotrara Formation of the Adamstown Sub-Group, is given in Figure 6.2. The constituents are coal, conglomerate, sandstone, shale and rocks of pyroclastic origin called “tuff”. The West Borehole coal splits within the Newstan Colliery leasehold into the Young Wallsend, Yard and Borehole coal as shown in Figure 6.3. The split is one of a set believed to have been caused by tectonically induced differential settlement on the flank of the Lochinvar Anticline (Warbrooke & Roach, 1986). The latter is located on an old basement high and forms the western limit of the Newcastle Coalfield. The Coal seams comprise of high volatile bituminous coal with medium to high ash and low sulphur contents. They contain minor, occasional shale bands up to 150 millimetres in thickness, and some deterioration in coal quality in those “plies” which comprise the uppermost element of the constituent coal seams. The maximum thickness of the West Borehole Coal within Newstan Colliery is approximately six metres at the line of convergence.
The Shepherds Hill Formation overlies the Young Wallsend/ West Borehole coals. The basal member of which comprises the Nobbys Tuff, which is the most persistent marker
159 Figure 6.1 Newstan Colliery longwall panels, (after Fowler & Torabi,1997)
160 Kotaro Formation Merewether Conglomerate Member
'Victoria Tunnel Coal Shepherds Hill Formation Nobby s Tuff Member Nobbys Coal ^ Bor Beach Formation o Signal Hill Conglomerate Member L Young Wallsend -5 L om bton Dudley Coal J Coal .5 ~o Su b- Grou p< CQ O Bogey Hole Formation Cockle Creek Conglomerate Member Y ard Co al Tighes Hill Formation Ferndale Conglomerate Member - Borehole Coal __ Waratah Sandstone
Figure 6.2 Generalised stratigraphic sequence, Lambton Sub-Group and overlying Kotara Formation, Newcastle Coal Measures (adapted from Standing Committee on Coalfield Geology of New South Wales, 1974)
Nobbys S. Young Wallsend S. West
Borehol Dudley S. Seam Yard S. Borehole
Figure 6.3 Section through Young Wallsend and The West Borehole coals passing near to Newstan Colliery. (after Warbrooke & Roach, 1986)
161 horizon through out the Newcastle Coalfield. The Nobbys Tuff Member is overlain by the unnamed sandstones and shales which constitute the remainder of the Shepherds Hill Formation. In the generalised stratigraphic sequence, the Shepherds Hill Formation is immediately overlain by the Victoria Tunnel Coal, the roof of which forms the upper limit of the Lambton Sub- Group. However the Victoria Tunnel Coal deteriorates to the west and is difficult to identify on the western side of the Lake Macquaire.
Overlying the Lampton Sub-Group is the Adamston Sub-Group at the base of which is the Kotara Formation. This formation shows considerable variation in its lithology but the Merewether Conglomerate Member is particularly well developed and forms the basal member in an area to the north-west of Lake Macquaire which includes part of Newstan Holding. The main body of the Merewether Conglomerate exhibits a wide variation in particle size ranging from pebbly gravel to silt. Large cross-beds alternate with massive conglomerate sheets in combinations which are up to 50 metres in thickness. The various particle sizes are concentrated in thin, discontinuous lenses with abrupt transitions. Both clast-supported and matrix-supported beds occur. The latter are thicker and contain the largest pebbles, up to 60 mm in diameter. In the longwall panels, the Nobbys Tuff Member forms the immediate roof to the Young Wallsend /West Borehole coals. In some localities, however, the Victoria Tunnel Coal and some other elements of the Shepherds Hill Formation were absent with the result that the main roof comprised the Merewether Conglomerate member. Figure 6.4 gives the longitudinal section through the set of up-holes drilled in roof strata of maingate 7 at Newstan Colliery and illustrates the variation in thickness of the conglomerate member along with the interburden. The thickness of the immediate roof between the coal seam and the conglomerate varies from 1.5 metres to up to 25 metres. The immediate roof comprises mainly siltstone and sandstone and is readily cavable. The unconfined compressive strength (UCS) of siltstone is 115 MPa while that of the overlying conglomerate only 68 to 83 MPa (Creech, 1996). Within the panels shown in Figure 6.1, a set of closed, widely spaced roof joints trends in a north-north westerly direction, sub parallel to the predominant cleat. Faulting is commonly of a normal type with sub
162 Newstan 7, maingate 1997) strata, 30 Torabi, roof & the Fowler through (after section Longitudinal 6.4 e *3 00 • t-H Uh 163 vertical fault planes and vertical displacements of less than 0.75 metres although, just outside the project panels, a throw of 7 metres has been recorded. Most faults strike approximately north-north-west. Dolerite dykes are sub-vertical to vertical and usually strike in a similar direction to that of the faults (Fowler & Torabi, 1997). 6.2.2 Stress field at Newstan Measurements of the insitu stresses had been carried out at the nearby West Wallsend Colliery (Walton, Enever, Litterbach & Crawford, 1987). The results have been shown in Table 6.1 along with that for Moonee (Enever & Walton, 1988). The directions of the major (ai) and intermediate (o2) principal stresses are horizontal or sub horizontal and consequently there is little difference between their magnitudes and azimuths and those of secondary (aH, c*h) principal stresses in the horizontal plane. The minor principal stress (o3) is nearly vertical. The minor secondary principal stress (ah) is much higher than the vertical stress (ctv). 6.3 WIND BLAST EVENTS AT NEWSTAN COLLIERY The first incidence of a significant wind blast occurred on 8 August, 1995 in longwall panel no. 6 which resulted in injuries to three persons, overcast damage and stoppings being blown over. The wind blast monitoring system was subsequently deployed and during the mining of LW7 to LW9 between 1995 and 1997, a total of 23 events were recorded. Eight of these events were classified as significant (Fowler & Torabi 1997). The Wind blast Data Logger velocity trigger levels (all pods) at Newstan Colliery longwall panels were as follows (Fowler & Torabi, 1997): LW 7: 5 metres per second; LW8: 7 metres per second; LW9: 5 metres per second. The overpressure trigger level (all pods) was 109.5 kPa in all panels. Table 6.2 lists the significant events recorded at Newstan Colliery. The events were widely spaced in the panels as seen from the face chainage/locations. Some of the events were apparently due to sequential falls of the massive conglomerate roof elements. 164 Table 6.1 Measurements of Insitu stress , Lake Macquaire District (after Walton et al, 1987; Enever & Walton, 1988) Colliery West West West Moonee Wallsend Wallsend Wallsend Method CSIRO HI. CSIRO H I. CSIRO H I. CSIRO H I. Cell & Cell Cell Cell Hydrofrac Rock Type Sandstone Sandstone Siltstone, Variegated Sandstone, Sandstone Siltstone, laminated Depth (m) 190 190 190 90 CTi (MPa) 22.4 25.3 32.8 - Dip/Dip Dir.(deg) 02/191 04/015 02/196 ct2 (MPa) 17.6 15.1 25.2 - Dip/Dip Dir.(deg) 10/281 17/281 00/286 a3 (MPa) 8.4 9.2 3.3 - Dip/Dip Dir.(deg) 80/087 73/117 89/017 av (MPa) 8.7 9.8 3.3 2.43 cth (MPa) 22.3 25.2 32.8 11.7 Dip/Dip Dir.(deg) 00/009 00/015 00/015 00/010 ah (MPa) 17.4 14.6 25.1 8.3 Dip/Dip Dir.(deg) 00/094 00/105 00/105 00/100 1.70 X 2.45 5.3 3.85 6.3.1 Caving Mechanism at Newstan Colliery During the retreat mining of the 225 metres wide Longwall Panel no. 5 (LW5) under a depth of cover of nearly 200 metres, Newstan Colliery had faced caving problems with heavy periodic weighting, mid face roof falls and face spalling. This panel was under a massive sandstone/conglomerate filled channel up to 50 metres thick and 5 to 25 metres above the worked seam. The above problems caused production delays and cost overruns. To avoid the LW5 scenario the subsequent panel was split longitudinally into two sub-panels (LW6 and LW7) as shown in Figure 6.1. During the mining of these panels, wind blasts were experienced. 165 Table 6.2 Significant wind blast events at Newstan (adapted from Fowler & Torabi ,1997; Creech ,1996) Longwall Date of Time of Face Peak wind velocity panel event event Chainage at no. maingate/tailgate* /Location (uncorrected) (m) (m/s) 7 9 Nov 95 5:17:05 675 22 7 29 Nov 95 11:41:45 255 40 8 17 April 96 18:47:50 465 39 8 28 April 96 15:55:21 360 40 8 28 April 96 16:07:57 360 39 8 30 May 96 20:37:33 200 32 8 12 June 96 13:29:58 128 23 * 9 1 Feb 97 9:22:09 - 33 * tailgate Figure 6.5 and Figure 6.6 illustrate the location of the wind blast events in the longwall panel in relation to the conglomerate channel height above seam and in relation to the subsidence profile respectively. The narrow panels did not face periodic v/eighting. Good mining conditions prevailed as the massive strata bridged the goaf and followed shortwall caving mechanism, with the breakage line behind the face. While the immediate roof was friable and readily cavable, the caving was insufficient to fill the goaf and consequently resulted in a 1.5 to 2 metres gap above and behind the chock. Wind blasts were experienced where the massive strata bridged the goaf and the immediate roof septum was less than twice the extraction height. These conditions created a significant gap above and behind the chock and exposed the face to air displacements following any major roof fall. However, provided that the base of the 166 350500E 35T000E 0 9/ / Conglol 'o 28.5 Conglomerate splits 352000E 3505O0E 3S1000E 351SOOE Figure 6.5 Location of the wind blast events in the longwall panels in relation to the height of channel above seam, Newstan Colliery. Barbed lines indicate areas where channel is less than 7 m above seam (adapted from CreechJ1996) 167 channel was sufficiently distant (i.e. vertically separated) from the seam horizon, then the immediate roof bulked up sufficiently to fill the goaf void and cushioned any failure of the overlying channel strata thereby obviating wind blasts. Face lengths for LW8 panel and LW9 panel, were increased to 127 m between gateroad centrelines. However, for part of the extraction of above panels, during which the geological conditions described above persisted, wind blasts remained a hazard (Simpson, Hebblewhite & Fowler, 1997). Microseismic monitoring at Newstan revealed that large magnitude events (Richter magnitude >1.0) preceded each wind blast with average source location at 324 metres lateral distance away from the geophone. The geophone was located 200 metres ahead of the face or sometimes behind it. Numerous smaller seismic events followed the face position and were considered to be caused by the failure of the septum material immediately above and behind the face (Creech, 1996). The above observations confirmed that the wind blasts were caused by the failure of the basal block of the conglomerate unit and localised to where the thickness of the channel is such that it bridges the goaf. In contrast, at Moonee, incidence of large goaf falls and associated wind blasts continued for virtually the whole length of the longwall panels other than few localised faulted zones where “regular caving” took place. Also, unlike Moonee, the goaf at Newstan was not accessible due to the caving of the immediate roof and the areal dimensions of the roof element contributing to the wind blast remain unknown. In the light of the Moonee experience and also based on face locations for Newstan panels during significant wind blast events, as shown in Table 6.2, it is evident that large spans of conglomerate roof were involved. For instance, at Newstan, the wind blast events for LW8 panel on 17 April, 1996 and 28 April, 1996 occurred after a retreat of 105 metres, and the incident of 30 May, 1996 after a retreat of 160 metres and that of 12 June, 1996 after a retreat of 78 metres respectively. The two events on 28 April, 1996 were apparently due to sequential falls of the roof within a short time interval. Variations in local geology, especially the thickness of the conglomerate member, apparently had a marked influence on the caving mechanism and the 168 dimensions of the roof elements involved and consequently on peak over pressures recorded during the wind blasts in the different panels. 6.4 REANALYSIS OF NEWSTAN WIND BLAST DATA The wind blast data set for Newstan Colliery has been reanalysed using the WaveMetrics Igor Pro version 3.13 software and using macros developed for the analysis of the Moonee Colliery data set. The data set includes all the significant wind blast events recorded and also some of the less significant events and is given in Appendix G. The procedure for processing the raw data and analysis is similar to that used for Moonee and discussed in sections 5.3.5 and 5.4.2 of Chapter 5. Earlier analysis (Folwer & Torabi, 1997) did not process the overpressure time histories of the wind blast events and only the peak wind velocities (uncorrected) associated with the events were reported as given in Table 6.2. Comparing Table 6.2 and Table G.l, Appendix G, it is seen that the smoothed and corrected values for the peak wind velocities are slightly lower compared to the previous analysis for the same events at the inbye sensor Pod 1 location. Some values (during the reanalysis) have been discarded as the wind velocity time histories were distorted due to blockage of the inbye sensor pod’s nose cone. The advantage of reanalysing the ‘raw” data is that same standardised procedures are followed for both the collieries and comparisons can be made. At Newstan, two wind blast sensor pods were located in the maingate, the first (Pod 1) mounted on the Pantech outbye to the face and the second (Pod 3 or Pod 4) at a fixed outbye distance of 116 metres from the former, on the turning circle as shown in Figure 6.7. The third sensor pod (Pod 2) was installed in the travelling roadway outbye to the face. 6.4.1 Overpressure and Wind Velocity Time Histories The overpressure time history and wind velocity time history of a significant wind blast recorded in the maingate near to the face during goaf fall dated 30 May, 1996, in LW8 panel, are given in Figures 6.8 and 6.9 respectively. The Newstan wind blast time histories exhibit the following salient features: 169 STARTLINE-LS06-1100m FINISHLINE-LS49 WINDBLASTS Figure 6.6 Location of the wind blast events in relation to the subsidence profile and channel thickness, Newstan Colliery. (after Creech, 1996) Figure 6.7 Location plan of wind blast sensor pods, Newstan Colliery. (after Torabi-1997) 170 1. The overpressure time histories of two significant events dated 29 November, 1995 and 30 May, 1996 (Figure 6.8) in LW7 and LW8 panels respectively, exhibit a large positive phase duration compared to other wind blast events. This may be due to larger spans of roof involved in the goaf fall and also increase in the wind blast overpressure fall time; 2. In contrast, the less significant events exhibit a short positive duration and peak values as shown in Figures 6.10 and 6.11 for overpressure and wind velocity time histories of the wind blast event dated 29 November, 1995 at LW7 panel; 3. All overpressure time histories exhibit similar but not identical characteristics for equivalent locations in both gate roads in the near field close to the face ends; 4. The “suck back” or underpressure phase is well defined in the time histories though truncated for longer duration events by the set recording interval and occasionally affected by secondary falls; 5. Wind velocity time histories (Figure 6.9) also follow the same shape as the overpressure time histories except for events during which the nose of the sensor pods became blocked by air-entrained debris. 6.4.2 Analysis of Overpressure and Wind Velocity Data This section examines some of the interrelationships between the parameters listed below for the wind blast events witnessed at Newstan longwall panels: Peak overpressure; Impulse; Maximum rate of rise of overpressure; Peak wind blast velocity; Maximum excursion; Maximum rate of rise of wind blast velocity. 171 300596#! oppod2 Overpressure time history Peak Smoothed Overpressure - tme (seconds) Figure 6.8 Overpressure time history of a significant wind blast, panel 8, Newstan Colliery. 300596# t pod2 Wind velocity tin Velocity = 24 nVs 3 4 5 Elapsed Time (seconds) Figure 6.9 Wind velocity time history of a significant wind blast, panel 8, Newstan Colliery. 172 051295#1 oppodl Overpressure time history Peak Smoothed Overpressure = 2 kPa 3 4 5 Elapsed Time (seconds) Figure 6.10 Overpressure time history of a less significant wind blast, panel 7, Newstan Colliery. Peak Smoothed 05129541 pod 1 Wind Blast Wind velocity time history Velocity = 4 m/s o 1.5 — ime (seconds) Figure 6.11 Wind velocity time history of a less significant wind blast, panel 7, Newstan Colliery. 173 6.4.2.1 Characteristics of the Overpressure time history Figures 6.12 and 6.13 illustrate the relationship between impulse and peak overpressure. It utilises data from the subset obtained from Pod 1, which was mounted on the Pantech in the maingate, outbye of the face and from Pod 2 on the turning circle outbye of Pod 1. The relationship is non-linear and a second-degree polynomial curve has been fitted to the data. The coefficients of correlation are high as shown by the R square values in the figures. This trend is similar to that observed at Moonee Colliery and discussed in section 5.4.2.1, Chapter 5. Figure 6.15 illustrates the relationship between the maximum rate of rise of overpressure and peak overpressure at Pod 1 location. A non-linear relationship is plotted. The relationship is not well defined but shows a positive correlation. The trend is again similar to that at Moonee Colliery where the data set was much larger and the relationship better defined (Figure 5.13). 6.4.2.2 Characteristics of wind velocity time history Figures 6.16 and 6.17 illustrate the relationship between maximum excursion and peak wind blast velocity for Pod 1 and Pod 2 respectively. The relationship is non-linear and a second-degree polynomial curve has been fitted to the data. The implication of non linearity is that as wind blasts become more intense, the maximum excursion increases more than does the peak wind blast velocity, i.e. the duration of the positive (wind blast) phase becomes longer. This trend is again similar to that observed at Moonee Colliery and discussed in section 5 4.2.2, Chapter 5. Figure 6.14 illustrates the relationship between maximum rate of rise of wind blast velocity with peak wind blast velocity at Pod 1 location. A linear relationship is plotted. The coefficient of correlation is reasonably high as shown by the R square values in the figure. 6.4.2.3 Relationships between overpressure and wind velocity time histories Figures 6.18 and 6.19 illustrate the relationship between peak wind blast velocity and peak overpressure at Pod 1 and Pod 2 locations respectively. It will be observed from the R square values in the figures that the parameters are positively correlated with good correlation. 174 Pod 1 O Newstan —Poly (Newstan) 0.12X2 + 0.52x R2 = 0.90 Peak Overpressure (kPa) Figure 6.12 Relationship between impulse and peak overpressure Pod 2 O Newstan ----- Poly. (Newstan) 0.064X2 + 0.93x R2 = 0.89 O O Peak Overpressure (kPa) Figure 6.13 Relationship between impulse and peak overpressure 175 O Newstan Pod 1 —Linear (Newstan) 50 - o 30 - y = 1 36x R2 = 0.69 Peak Velocity (m/s) Figure 6.14 Relationship between maximum rate of rise of velocity and peak velocity 0 J------'------'------'------'------1------1 0 2 4 6 8 10 12 Peak Overpressure (kPa) Figure 6.15 Relationship between maximum rate of rise of overpressure and peak overpressure 176 O Newstan Pod 1 ------Poly (Newstan) ■=• 60 - £ 50 - E 40 - S 30 - y = 0.04X2 + 0.41 x R2 = 0.91 Peak Velocity (m/s) Figure 6.16 Relationship between maximum excursion and peak velocity Pod 2 O Newstan ------Poly. (Newstan) y = 0.10X2 - 0.081x R2 = 0.90 Peak Velocity (m/s) Figure 6.17 Relationship between maximum excursion and peak velocity 177 Pod 1 O Newstan ------Linear (Newstan) g 25 - y = 4.4x R2 = 0.66 S 20 ■ Peak Overpressure (kPa) Figure 6.18 Relationship between peak velocity and peak overpressure Pod 2 O Newstan ------Linear (Newstan) o 15 - y = 2.62x R2 = 0.86 Peak Overpressure (kPa) Figure 6.19 Relationship between peak velocity and peak overpressure 178 The parameters shows a tentative linear relationship at Pod 1 location in maingate near to the face with the following equation: Peak wind velocity (in m/s) = 4.4 x peak overpressure (in kPa) (6.1) The relationship is better defined for Pod 2 location in the travelling road outbye of the face and given by the relation: Peak wind velocity (in m/s) = 2.62 x peak overpressure (in kPa) (6.2) The Rankine-Hugonoit relationship is close to the lower bound of the data for Pod 1 location and is closer to data for Pod 2, outbye to the face location. A similar trend was observed at Moonee Colliery. 6.4.3 Attenuation of wind blast intensity with distance from goaf fall Attenuation of peak overpressure and peak velocity with distance along the maingates of longwall panels is shown in Figures 6.20 and 6.21 respectively. It utilises data from Pod 1, which has been mounted in the maingate on the Pantech and that from Pod 3 or 4 mounted on turning circle at a fixed distance of 116 meters outbye. Some of the values recorded have been discarded as they exhibited higher values with attenuation ratio greater than one. The possible reason for lower values of overpressures at Pod 1 location compared to the outbye Pod 2 location (for the discarded values) could be due to the Pod 1 being near to a cut-through or in an intersection. The parameters are evidently negatively correlated. An exponential decay curve has been (tentatively) fitted to the data. The data is too scanty and scattered to derive any meaningful conclusion. 6.5 COMPARISON OF RESULTS The wind blast parameters at Newstan and Moonee collieries are compared in Table 6.3. The wind blast parameters maximum values associated with significant wind blast events at Newstan are nearly a third in magnitude when compared to those at Moonee. The comparison is based on the values recorded near the face in the maingate for natural falls (Class M) in the first two longwall panels at Moonee and the influence of attenuation due to location of the sensor pods would be marginal. The relationships between the wind blast parameters exhibit similar trends at both collieries. 179 Overpressure attenuation in main gate § 07- -0.0018x o 0.5 R2 = 0.40 Distance between Pods (m) Figure 6.20 Peak overpressure attenuation with distance Velocity attenuation in main gate o 0.6 « 0.5 - I0031X R2 = 0.27 Distance between Pods (m) Figure 6.21 Peak velocity attenuation with distance 180 Table 6.3 Wind blast parameters at Newstan and Moonee Collieries Parameter Maximum value Newstan Moonee Peak air velocity 39 m/s 130 m/s Max. rate of rise of velocity 52 m/s/s 174 m/s/s Peak overpressure lOkPa 34 kPa Max. rate of rise of pressure 25 kPa/s 38.5 kPa/s Impulse 21 kPa.s 89 kPa.s Excursion 79 m 181 m (air flow distance) 6.6 CONCLUSIONS The wind blast events at Newstan were localised and confined to where the massive strata was within twice the extraction thickness and bridged the panel. In contrast, at Moonee, the incidence of large goaf falls and associated wind blasts continued for virtually the whole length of the longwall panels except in a few localised faulted zones where “regular caving” took place. The wind blast parameters maximum values associated with significant wind blast events at Newstan are nearly a third in magnitude when compared to those at Moonee. The relationships between the wind blast parameters exhibit similar trends at both collieries. Although the goaf at Newstan was not accessible due to the caving of the immediate roof, the spacing of the events, especially at LW8 panel, suggests that alike Moonee, large spans of roof were involved. The lower magnitudes of the wind blast parameters at Newstan are consequently due to the following facts: the fall of the roof element was cushioned by the caved immediate roof, lesser volume of air being displaced from the void and also the higher resistance to flow due to partial packing of the goaf by the prior caving of immediate roof. 181 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS REGARDING WIND BLASTS Wind blast monitoring at Moonee Colliery and prior to that, at Newstan Colliery, in conjunction with the micro-seismic and support leg pressure monitoring by the mine management has lead to some understanding of the geo-technical and the fluid-dynamic aspects related to the occurrence of wind blasts and their characteristics in longwall panels of underground coal mines. The main conclusions and recommendations based on the field investigations related to the above mentioned aspects are discussed in the following sections. 7.1 GEO-TECHNICAL ASPECTS RELATING TO WIND BLASTS The geo-technical factors affecting wind blasts in longwall panels under massive strata in underground coal mines are summarised as follow: 1. Similar to the geo-mechanical conditions in shortwall panels (Frith, 1993), massive strata is expected to bridge the goaf or fail well behind the face line in longwall panels of narrow widths; 2. Incomplete caving, delayed caving and caving arrested by the presence of competent strata in close proximity to the seam leads to gap formation at the base of the goaf and the possibility of wind blasts during the caving of the overhanging roof/competent strata; 3. Geological factors affect the way in which the roof falls. The geological structure of the roof, the insitu stresses, the mechanical properties of both the rock fabric and of the discontinuities have significant impact on roof cavability with respect to the panel layout; 182 4. Geometrical factors such as thickness of the falling roof element and the distance through which it falls, together with both its plan area and the total plan area of the standing goaf and the roadways connected to it, affect the magnitude of the wind blast; 5. At Newstan Colliery, Wind blast activity was exclusively confined to where the conglomerate channel was thick enough to bridge the goaf or where it occurred within 7 m of the working roof or approximately twice the extraction thickness (Simpson, Hebblewhite & Fowler, 1997). 7.1.1 Caving at Moonee Colliery At Moonee Colliery, caving behaviour and the incidence of wind blasts were studied in 5 longwall panels. Induced caving of roof using hydraulic fracturing technique was introduced at Moonee Colliery longwall panel no. 3 for caving “on demand”. However, sometimes the roof failed to cave immediately after hydraulic fracturing and caved after further retreat of the longwall face. The observed roof falls at Moonee can be broadly classified as: i. Class M: ‘Natural’ falls due to mining ; ii. Class F: Hydraulic fracture induced falls; and iii. Class C: Falls induced by hydraulic fracturing and subsequent mining. It has been observed that: 1. Roof falls occurred only after the development of large roof spans and generated significant wind blast events for all the classes of falls. Goaf falls from greater heights resulted in more intense wind blasts. (Sections, 5.4, 5.5.3, 5.5.5, Appendix D); 2. Visual inspections of roof falls revealed that the falls ranged from 6 to 10 metres height into the conglomerate (section 5.5); 3. Rarely did any rock noise (roof talk) acted as the alarm criterion for the roof falls (Tables 5.7, 5.8); 183 4. The microseismic warning times were small and in a few occasions no warnings preceded the roof falls (Tables 5.7, 5.8, 5.11-5.13); 5. Evidence of periodic weighting was not indicated by the measurements of leg pressures of the supports, nor any long term trends (72 hours observation) in the period leading up to the goaf falls. Face conditions had been relatively good (section 5.4.3); 6. Support Leg Pressure Strata Alarm had rarely suggested the approach of roof falls (section 5.4.3). 7.2 FLUID-DYNAMIC ASPECTS OF WIND BLASTS A wind blast comprises a rapid rise in the absolute pressure of air in the goaf to a maximum (positive compression phase), followed by a similarly rapid fall to below the ambient atmospheric pressure (expansion phase “suck back”). After decreasing to a minimum value, the absolute pressure gradually increases until it again equals the ambient atmospheric pressure. At around the same time, although not necessarily in phase with the overpressure, the wind velocity also rises rapidly to a maximum and then exhibits a sudden reversal in to the “suck back” phase. The events usually last for a few seconds. From wind blast monitoring and the recorded overpressure and wind velocity time histories, it has been observed that: 7.2.1 Acoustic Effect 1. There is no acoustic precursor to the event. Consequently people in the working place will receive no warning of the wind blast before it strikes them unless they hear “roof talk” or a wind blast warning system is in place (Section 5.2.5); 2. The onset of the event is very sudden, both overpressure and air velocity exhibiting a rapid rise (Section 4.3); 184 3. The intensity of the wind blast phase can be very severe, comprising a very large overpressure and corresponding high wind velocity (Section 5.4.1; Appendices B & C). 7.2.2 Overpressure Characteristics 1. The peak overpressure is always greater than the peak underpressure. Overpressure time histories are similar, but not identical, for equivalent locations in both gateroads in the near field close to the face ends (Figures 5.8, 5.9, Section 5.4.1); 2. The overpressure pulse in the maingate attenuates with distance as it propagates outbye. The attenuation is predominantly as a result of air viscosity, frictional resistance of the roadways although other factors, such as flow through cut- throughs (spreading) contribute (Figure 5.9, Section 5.4.1; Section 5.4.2.4; Section 5.6); 3. The duration of the positive overpressure phase increases slightly as the pressure pulse propagates outbye due to dispersion, i.e. the tendency for different frequencies to travel at slightly different velocities (Figure 5.9, Section 5.4.1). It is seen from the typical overpressure time histories for wind blasts (Figure 4.8, Section 4.3.1) that the end of the positive pressure phase is marked by the zero overpressure. As zero overpressure is characteristic of a sound wave, the zero overpressure condition moves forward with the local speed of sound, slower than any pressure pulse with finite overpressure, which is necessarily at supersonic speed (Section 4.2.4). The duration also increases with intensity of the event and could be related to the caving mechanism, interaction of the roof element with the air below, the panel and goaf geometry (Section 5.5.4.1; Section 4.4.4); 4. The celerity or the rate of propagation of the event for pressure pulses far outbye of the longwall goaf, equals the speed of sound (Figure 5.9); 5. Some wind blast events consist of sequential fall of roof elements (Figure 5.10, Section 5.4.1; Figures B.7 - B. 10, Appendix B); 185 6. There appears to be an upperbound to peak overpressure with increasing goaf/fall area (Figure 5.21, Section 5.4.2.5; Figure 5.39, Section 5.5.4.4); 7. Empirical scaling laws for wind blasts based on explosive equivalency, could not be reliably defined due to the unpredictability and the variability associated with the roof falls, the caving mechanism in the goaf and the geo-technical properties of the coal measures involved in the caving process (Section 5.6). 7.2.3 Wind Velocity Characteristics 1. Wind velocity time histories also follow the same overall shape as the overpressure time histories, with rapid rise to a maximum and then a sudden reversal into the suck back phase (Figure 4.9, Section 4.3.1; Figure 5.7, Section 5.4.1); 2. Peak air velocity in the direction away from the fall is always greater than the peak “suck back” velocity; 3. Wind blasts with peak wind velocities greater than 20 m/s are considered significant as they can result in injuries to mine personnel. Hurricane level wind speeds have been monitored in mine roadways (Appendix C) but not exceeding Mach one values which exclude the possibility of normal shock conditions developing in roadways (Section 4.2.3); 4. While the overpressure time histories for wind blast events show resemblance with characteristic of weak shock waves, the wind velocities monitored are much higher than predicted values based on Rankine-Hugoniot relationship between peak particle velocity and overpressures. This reveals the impact of air expulsion from goaf and the panel geometry at the measuring locations. Departure from theoretical values is higher for locations nearer to the goaf. However, the possibility of shock conditions developing in the goaf cannot be ruled out (Figures 5.18 & 5.19; Section 5.5.4.3); 5. Occasionally, the suck back phase with negative velocities (towards goaf) has not been recorded (Figures 5.11, B.10 & B.14). This could be due to blockage of air return paths to the goaf, truncation of the velocity record prior to the event being 186 over because of the set recording time, flow being affected by sequential falls and associated pressure pulses and also the orientation of the velocity pods not being appropriate for reverse flow monitoring. 7.2.4 Impact of Cushioning The maximum values of wind blast parameters associated with significant wind blast events at Newstan Colliery are nearly a third in magnitude compared to those at Moonee Colliery. The relationships between the wind blast parameters however, exhibit similar trends at both the collieries. Although the goaf at Newstan Colliery was not accessible to observation due to the caving of the immediate roof, the spacing of the events, especially at LW8 panel, suggests that alike Moonee Colliery, large spans of roof were involved. The lower magnitudes of the wind blast parameters are considered to be due to the fall of the roof element being cushioned by the caved immediate roof, lesser volume of air being displaced from the void and also, the higher resistance to flow due to partial packing of the goaf by the prior caving of immediate roof (Sections 6.5 & 6.6). 7.2.5 Wind Blast Events Associated with Hydraulic Fracturing of Roof The failure process was always sudden for the conglomerate roof, irrespective of the class of roof fall i.e. Class M, hydraulic fracture induced (Class F) or falls induced by hydraulic fracturing and subsequent mining (Class C) (Tables 5.7, 5.8, 5.11-5.13). Wind blasts were invariably associated with the above classes of roof falls at Moonee Colliery (Appendix C). For comparable goaf falls areas, wind blasts resulting from Class F falls are apparently more intense compared to those from Class M and Class C falls (Table 5.14, Section 5.5.4.4). Hydraulic fracturing induced caving provides caving “on demand” by initiating fracture near to the determined final stable goaf geometry. Success of the hydraulic fracturing treatment depends on the following: 1. State of the stresses acting on the roof strata; 187 2. Fracture achieving sufficient dimensions with proper pump pressures and water flow rates maintained throughout the treatment; 3. Sufficient roof overhang so that instability caused by the hydraulic fracturing initiates a failure process that is inherently unstable. Compared to the natural mining induced falls (Class M), hydraulic fracturing introduces the high fluid pressures, which exert a downward force on the uncaved strata along with artificially removing the tensile strength of the rock over an increasing area. The immediate impact of hydraulic fracturing on wind blasts is that the intensity of the event can be reduced based on the area of roof induced to fail. The trend of upperbound for peak overpressures (section 5.4.2.5) is obvious for the goaf fall areas exceeding 5000 square metres in case of hydraulic fracturing induced falls (Class F) or falls induced by hydraulic fracturing and subsequent mining (Class C). However, this trend for Class M falls was apparent for falls areas exceeding 10,000 square metres (Section 5.5.5). Moreover, as the mine personnel are evacuated from the face/workings prior to the hydraulic fracturing, the risk of injury due to the wind blasts is eliminated. 7.3 RECOMMENDATIONS Wind blast will continue to be a problem in some parts of the Newcastle Coalfield, especially in narrow width longwall operations which have near seam massive conglomerate roofs and the immediate roof is of insufficient thickness so as to fill the void. Microseismic monitoring and support leg pressure monitoring for roof failure prediction as well as induced roof caving techniques may help in reducing the associated risks. Wind blast monitoring results during the current investigations will help in establishing safe layout and design norms for mining under massive conglomerate roof and in risk assessment for operational and safety considerations. The quantification of wind blast intensity and of the attenuation of intensity with distance from the face have helped the Moonee Colliery in the following: 188 1. Determination of the relationship, between the area of roof fall and the magnitude and intensity of the associated wind blast with implications for operational risk assessment based on the extent of roof overhang; 2. Defining the extent of areas within which full ‘wind blast working conditions’ must be implemented for the protection of mine personnel. And developing safe working procedures under a wind blast management plan including longwall zoning rules so as to reduce the risk of personal injury to a level as low as reasonably achievable. A “Green Zone” rule with 0-20 meters of goaf standing since last fall, with low risk of injury in event of roof fall and a “Red Zone” rule with greater than 20 meters of goaf standing since last fall also in the event of triangular roof elements greater than 500 square meters, was initially implemented. The Red Zone has eventually been extended to coincide with start of “cutting” in the face irrespective of the goaf overhang; 3. Determining safe distances or exclusion zones for periods of high wind blast risk and providing guidance regarding ‘safe havens’; 4 Providing guidance for overall panel design; 5. Quantification of the results of wind blast reduction measures which include the following : i. Providing wind blast safety shields to close potential pathways by which wind blast may directly impinge on mine personnel at the coal face; ii. Erecting wind blast proof stoppings or regulators in mine openings; iii. Determining the effect of induced caving by hydraulic fracturing. 6. Correlation of measured wind blast intensity with physical damage. The extent of damage to elements of the mine such as water barriers and mine stoppings has been shown to be related to the intensity of the event as determined by measured overpressures and wind velocities. This has proved valuable for design purposes. 189 7.3.1 Recommendations for Further Work 7.3.1.1 Monitoring of wind blasts It is recommended to continue the field monitoring program in order to collect more data to clearly establish a set of empirical wind blast ‘laws’. This would however be dependent upon the availability of monitoring sites in geological and geometric environments that differ from those of the project panels at Newstan and Moonee Collieries. The present wind blast monitoring system is capable of recording events for a maximum duration of 8 seconds. It is recommended that longer duration monitoring may help in complete characterisation of larger events and employment of the Mk2 Wind Blast Monitoring System, presently under construction, will facilitate such monitoring with higher resolution. Also, greater number of sensor pods are recommended in order to better define the distribution of overpressures and wind velocities around the working panel during wind blasts. Better characterisation of the wind blast negative “suck back” phase needs sensor pods orientated in direction outbye from the face. Risk of methane explosion due to expulsion of methane from goaf needs to be addressed by further research using the new Mk2 Wind Blast Monitoring System with up to four infra-red methane monitors. 7.3.1.2 Physical modelling Physical modelling (Fowler, Torabi & Daly, 1995; Fowler & Torabi, 1997) has given some insight to the wind blast phenomenon. However, the study was confined to fall of thin roof element simulating thickness less than 4 to 5 metres. The model layout was not suited for simulating the wind blast phenomenon in longwall panels. Moreover, the height of fall for the roof element/ piston was such that the wind blast exhibited a secondary phase characterised by a residual airflow corresponding to the period during which the roof element is falling at its terminal velocity (Sections 3.3 & 4.3). Attempts to simulate a wind blast corresponding to that observed in the field failed due to the 190 limitations of the physical model. It is recommended to further modify the model to make it suitable for simulating wind blasts in longwall panels 7.3.1.3 Numerical modelling The development of the numerical wind blast model is still continuing and it is recommended that early completion of the numerical model will help simulate complex roof falls and provide insight into the dynamics of the interaction between the roof elements and the air during a roof fall and help in confirming the results obtained from field investigations. Commercially available Computational Fluid Dynamics (CFD) packages presently lack this capability but new versions with additional features are under development and in distant future such commercially available packages may be suitable for numerical modelling of the wind blast phenomenon. 7.3.1.4 Research in caving mechanism The geo-technical aspects related to the goaf caving mechanism strongly influence the wind blast phenomenon. More research efforts need to be directed to caving mechanism in massive and competent strata. It is envisaged that with advancements in micro- seismic monitoring techniques and close collaboration with geo-technical engineers, a better understanding of the complex and presently unpredictable caving mechanism may result. Numerical modelling of the caving aspects for Moonee Colliery had failed to assess the risk of wind blasts accurately at the design stage. Research efforts need to be directed towards better numerical models calibrated for massive roof strata failure mechanism. 191 REFERENCES Anderson, I, 1991, ‘NSW Inspectorate views on pillar extraction: a case study from the Great Northern Seam’, Proc. Colliery Safety Symposium, North Ryde, NSW, Australia, 15 Oct. 1991, Publ. Sydney: New South Wales Department of Mineral Resources, pp. 4-21. Anderson, I, 1997, ‘Endeavour Colliery explosion investigation’, Minfo, No. 54, pp. 12-14. Anon, 1989, News Briefs - Mine Collapse, Coal, Volume 26, No. 12, pp. 29. Armstrong, M & Mishe, C, 1998, 1998 New South Wales Coal Industry profile, Publ. Sydney: New South Wales Department of Mineral Resources, ISSN 0814-3285. Baker, EW & Tang, MJ, 1991, ‘Gas dust and hybrid explosions’, Fundamental Studies in Engineering, Volume 13, Elsevier, Amsterdam, ISBN 0-444-88150-6 Baker, EW, 1973, Explosions in air, University of Texas press, Austin and London, ISBN 0-292-7200-3. Baker, EW, Cox, PA, Westine, PS, Kulesz, J.J & Strehlow, RA, 1983, Explosion hazards and evaluations, Elsevier, Amsterdam Baker, EW, Esparza, ED & Kulesz, JJ, 1977, ‘Venting of chemical explosions and reactions’ Proc. 2nd International Symposium on loss prevention and safety promotion in the process industries, Heidelberg, Germany. Barczak, TM & Goode, CA, 1982, ‘Considerations in the design of Longwall Mining Systems’, Proc. State of the art of Ground Control in Longwall Mining and Mining Subsidence, SME, June 1982, (eds.) Chugh, YP & Karmis, M, pp. 39-50. Bartknecht, W, 1981, Explosion - course, prevention, control, Springer-Verlag, Heidelberg, Germany, 1981, pp. 251. 192 Beer, G & Meek, JL, 1982, ‘Design curves for roofs and hanging walls in bedded rock based on “voussoir” beam and plate solutions’, Trans. InstnMin. Metall (Sect. A: Min. industry) 91, pp. A18-A22. Booking, M, Howes, M &Weber, C, 1988, ‘Palaeochannel development in the Moon Island Beach Sub-Group of the Newcastle Coal Measures’, Advances in the study of the Sydney basin, Proc. 22nd Newcastle Symposium, Newcastle, NSW, 15-17 April, 1988, Publ. Newcastle: University of Newcastle, pp. 37-45. Bofinger, CM, Cliff, D & Bell, S, 1984, ‘Characterisation of audible indicators of coal mine roof fall’, Australian Coal Journal, R.N. no. 44, pp. 3-8. Bofinger, CM, Cliff, DI & Bell, SL, 1993, ‘A study of audible coal mine roof fall warning signals’, Recent Developments in Underground Coal Mining in the Bowen Basin, Mini-Symposium, Emerald, Qld, 23 July, 1993, Extended abstracts, Brisbane: Australasian Instn Min. Metall, Central Queensland Branch, pp. 16.1-16.6. Brannagran, DF & Johnson, MW, 1970, ‘Permian Sedimentation in Newcastle Coalfield, NSW’, Proc Australian Inst Min Metall, no. 235, pp. 1 -36. Brode, HL, 1956, ‘Point source explosion in air’, RM-1824 -AEC, The Rand Corp., Santa Monica, California. Brown, ET & Windsor, C R, 1990, ‘Near surface insitu stresses in Australia and their influence on Underground construction’, Under ground domain, Proc. 7th Australian Tunnelling Conf. , Sydney 11-13 Sept, 1990, Publ, Sydney: The Institution of Engineers, Australia, pp. 18-48. Brown, M, 1999, ‘Four miners killed in underground air piston’, Sydney Morning Herald, 25 Nov., 1999. Caffery, M, 1990, ‘Air blast effects - summary of recent experiences at Southern Colliery’, Capricon Coal management Pty Ltd (unpublished report,) Carroll, B, 1977, An illustrated history of Australian mining, Macmillan Company of Australia, Pty Ltd, ISBN 0333 22978 9. Coal Mines Regulation Act 1982, No.67, New South Wales. 193 Coal Mining Inspectorate and Engineering Branch, 1996, ‘Explosion at Endeavour Colliery -28 June 1995-Summary of Investigation’, MDG no. 1007, New South Wales Department of Mineral resources. Coles, RRA, Garinther, GR, Hodhe, DC & Rice, CG, 1968, ‘Hazardous exposure to impulse noise’, JAcoust Soc, Am 43, 336. Creech M, 1995, ‘A review of subsidence data from Newcastle Coalfield and a new subsidence prediction method’, Mine Subsidence Technology Society, 3rd Triennnial Conference Proc., Febuary 1995. Creech, M, 1996, ‘Microseismic monitoring, wind blast and longwall mining under massive strata at Newstan Colliery’, Geology in Longwall mining, Proc. Symposium, The Univ. of New South Wales, Australia, 12-13 Nov., Publ. Sydney, pp. 239-249, Diessel, CFK, 1992, Coal-bearing depositional systems, Publ. New York, Springer - Verlag, ISBN 0 387 5216 5. Edwards, JL, 2000, Personal Communication. Enver, JR, Walton, RJ, 1988, ‘New instruments for strength and stress measurements in coal mines’, Field measurements in geomechanics, Proc. 2nd Int. Symposium, Kobe, 6- 9 Apr. 1987, Volume 1, AA Balkema, Rotterdam, pp. 59-74. Ethridge, NH, 1965, ‘A procedure for reading and smoothing pressure-time data from HE. and nuclear explosions’, BRL Memo Report No. 1961, Aberdeen Proving Ground, Md. Falkner, A, 1993, ‘Sedimentological studies in the German Creek Coal Measures and their relevance to longwall mining’, New developments in coal geology, Proc. Symposium, Brisbane, November, 1993, (ed.) Beeston, JW, Geological Society of Australia, Coal Geology Group, pp. 143-148. Flynn, PD, 1950, ‘Elastic response of simple structures to pulse loading’, BRL, Memo Report No. 525, Aberdeen Proving Ground, Md. Foster, M, 1979, ‘Development of roof control technique for massive roof problem in longwall operations’, US, Dept of Commerce, NTIS, 10F2, PB61-125270 194 Fowler, JCW & Torabi, SR, 1997, The dynamics of wind blasts in underground coal mines, Project Report No. 3, Publ. Sydney: The University of New South Wales School of Mines, Dept, of Mining Engg., ISBN 07334 1440 0. Fowler, JCW & Sharma P, 2000, The dynamics of wind blasts in underground mines, Project Report no. 4, Publ. Sydney: The University of New South Wales, School of Mining Engg., ISBN 07334 0700 5 Fowler, JCW, 1997, The dynamics of wind blasts in underground coal mine, Interim Project Report, Publ. Sydney: The University of New South Wales, Dept, of Mining Engg., ISBN 07334 1550 4. Fowler, JCW, Torabi, S R & Daly, CR, 1996, ‘Field investigations into wind blasts resulting from large falls of roof in Australian underground coal mine’, Proc. 96 Inti Symposium on Mining Science and Technology, Xuzhou, Jiangsu, China, 16-18 October 1996, (eds.) Yuguang, G& Golosinski, TS, A. A. Balkema, pp. 145-149. Fowler, JCW, Torabi, SR & Daly, CR, 1995, ‘Wind blasts in underground coal mines: a study of the effect of the rock mass properties of the falling roof using a physical model’, Proc. 6th Underground Operators' Conference, 13-14 November, 1995, (ed.) Golosinski, TS, AusIMM Publication Series, RN: 7/95 pp. 325-330. Frith, RC, Stewart, A & Price, D, 1992, ‘Australian longwall geomechanics - a recent study’, Proc. of 11th Int. Conf. on Ground Control in Mining, (eds ) Aziz, NI &. Peng SS, The University of Wollongong, NSW, July 1992, pp. 131-139. Fritzsche, CH, 1962, Bergbaukunde, Volume 2, 10th Edition, Springer Verlag. Gale WJ & Nemick JA, 1998, ‘Prediction of strata caving characteristics and its impact on longwall operation’, Proc. of the 1st Australasian Coal Operators Conference, Coal98, Wollongong 18-20 February, pp. 156-166. Gale, WJ & Stone, L, 1987, ‘Effects of virgin stress field orientation on coal mine roof condition in the Bulli Seam Southern Coalfield, New South Wales, Australia’, Coal Journal, RN: no. 16, 1987, pp. 21-28. Galvin JM, 1983, ‘The development of the theory of dolerite sill behaviour’, Proc. Fifth International Congress of the International Society of Rock Mechanics, El 27-El 32. 195 Galvin, JM, 1996, Strata control for coal mine design, Publ. Sydney, The University of New South Wales, school of Mines, ISBN 0 7334 14440 0. Gierke, Von HE, 1966, ‘Effects of sonic boom on people: review and outlook’, J. Acoust. Soc. Am., 39, S43. Gravistrom, SA, 1956, Loading characteristics of air blasts for detonating charges, Handlingarur 100,The Royal Institute of Technology, Stockholm, Sweden. Gupta RN & Ghose AK, 1992, ‘Strata -support interaction of a powered support longwall face under a massive dolerite sill - a study’, Proc. of the 11th International Conference on Ground Control in Mining, Wollongong, 7-10 July, 1992 Hanna NE & Zabetakis MG, 1968, Pressure pulses produced by underground blast, Report of investigation 7147, USBM, 16 pp. Hatherly, P, Xun L, McKavanagh, B, Dixon, R & Devey, D, 1995, ‘Seismic monitoring for mapping of longwall caving at Gordonstone mine, Bowen Basin’, Proc. Symposium 1995: 150 Years On, organised by the Bowen Basin Geologists Group and the Geological Society of Australia, Coal Geology Group, Mackay, Queensland, 1-3 October, 1995, (eds.) Follington, IL, Beeston, JW & Hamilton, LH, Sydney, Geological Society of Australia, pp. 117-126. Holla, L & Hughson, R, 1987, ‘Strata movement associated with longwall mining’, Proc. 6th Australian Tunnelling Conference, Melbourne, March, 1987, Volume 2, Parkville, Vic.: Australasian Institute of Mining and Metallurgy for the Australian Underground Construction and Tunnelling Association, pp. 211-216. Holla, L, 1986, ‘Evaluation of the surface subsidence characteristics in the Newcastle Coalfield of NSW’, The Coal Journal, Volume 11, pp. 11-22 Holt, GE, Caffery, M & Draheim, P, 1988, Water infusion and hydraulic fracturing to control massive roof caving. Stage l.Moura No. 2 Mine, Qld., Australi. National Energy Research, Development and Demonstration Program End of Grant Report, RN: EG 89/796, April, 1988 Hopkinson B, 1915, British Ordnance Board Minutes 13565. 196 Jin, SY & LinSheng, X, 1992, ‘Study of the impact of mining under massive roof at Datong coal mines, China’, Proc. 11th International Conference on Ground Control in Joint Coal Board, 1990, The wind blast phenomena (unpublished report.) Joint Coal Board, 1995, ‘Australia's longwalls: exclusive 12-page statistical review’, Mining Monthly, March 1995, pp. 23-38. Joy, J, 1990, ‘Applications of preliminary hazard analysis and job steps and condition charting to the introduction of longwall mining systems’, Case study no.l, System Safety and Risk Assessment for the Coal Mining Industry, Seminar/ Workshop, Brighton Beach, New South Wales, 25-26 July, 1990, Sydney: New South Wales Coal Association, Safety and Training Committee. Kidybinski, A, 1962, ‘Effect of cleavage of rocks on manifestation of rock pressure’, Przegl. Gorn., 18(1), pp. 16-20. Kidybinski, A, 1991, ‘Reliability of predictive methods for control of roof strata in longwall faces’, Proc. International Conference on reliability, production and control in coal mines, Wollongong, New South Wales, 2-6 September, 1991, Papers, AusIMM Publication Series, RN: 6/91, pp. 235-241. Kinney, GL, 1962, Explosive shocks in air, The Macmillan Company, New York, 1962, pp. 198. Liepmann, WH & Puckett, AE, 1950, Introduction to aerodynamics of a compressible fluid, John Wiley & Sons, Inc., New York, pp. 262. Macdonald, C, 2000, Moonee Colliery, Internal Report (unpublished). Mallett, LG, Vaught, C & Peters, RH, 1992, ‘Training that encourages miners to avoid unsupported roof, United States Bureau of Mines, Information Circular, RN: IC 9332, pp. 32-45. Mark, C, Chase, FE & Iannacchione, A, 1991, ‘Longwall mining: geologic considerations for better performance’, Engineering and Mining Journal, RN: 192(5) May, pp. 16C-16K. 197 McCowan, B & Mikula, PA, 1983, ‘Mathematical modelling for design of longwall layouts’, Proc. Longwall Mining Design - development and Extraction seminar, ACRIL, 11-12 Nov. 1983, Sydney, NSW, Australia, pp. 1-19. McPherson, MJ, 1980, ‘Air pressures developed in collapsing mine workings’, Trans. Second International Mine Ventilation Congress, Reno, 4 -8 Nov. 1979, Publ. New York Society of Mining Engineers of AIME, 1980, pp. 317 -325. McPherson, MJ, 1993, Subsurface ventilation and environmental engineering, Chapman & Hall, London, 905 pp. McPherson, MJ, 1995b, ‘The adiabatic compression of air by large falls of roofs’, Proc. 7th Symposium on Mine Ventilation, Lexington, Kentucky, 1995, Publ. Virginia Polytechnic Institute and State University, 1995. McPherson, MJ, Wu, XL & Karfakis, MG, 1995a, ‘The compression of air under large falls of roof ‘, Proc. 26th International Conference of Safety in Mines Research Institutes, Katowice, Poland, Sept 1995, pp 1455-159, Publ. Katowice: Central Mining Institute. Mills, KW & Grady, PO, 1998, ‘Impact of longwall width on overburden behaviour’, Proc. 1st Australasian Coal Operators Conference, Coal98, Wollongong 18-20 Feb, 1998, pp. 147-155. Mining, 7-10 July, 1992, (eds.) Aziz, NI &. Peng SS, The University of Wollongong, NSW, July 1992, RN: 7/92, pp. 540-547. Mumford, CJ & Lihou, DA, 1979, ‘New projects -hazard assessment and control: part 2-assessment and engineering controls’, Design for safety, Proc. Symposium, Institution of Chemical Engineers, Midland Branch, Univ. of Aston, Birmingham, 12-14 Sept, 1979, pp. 6/3:1 -23. Nguyen, VV, Enever, JR & Mallett, CW, 1983, A study of site conditions affecting roof caving potential in an area of massive sandstone, CSIRO. Division of Applied Geomechanics, Coal Mining Report, RN: no. 45, 1983. Niu, XZ & Gu, TG, 1982, ‘Control of the very hard roof by softening method using water injection, in: Strata Control: Co-operation between the Mining Engineer and the 198 Researcher’, Proc. 7th International Strata Control Conference, Liege, 20 -24 Sept. 1982, publ. Liege: Institut national des Industries Extractives, 1983, pp. 452 -461 Peng, SS & Chiang, HS, 1984, Longwall mining, John Wiley & Sons, Inc., New York, pp. 708. Peng, SS & Hsiung, SM, 1982, ‘Development of roof control criteria for longwall mining - parametric modelling’, Proc. of Developments in Geotechnical Engineering, Volume.32, Strata Mechanics, (ed.) Farmer, IW, 5-7 April, 1982, Newcastle upon Tyne, U.K., pp. 51-58. Persson, PA , Holmberg, R & Lee, J, 1996, Rock blasting and explosives engineering, CRC Press, Boca Raton, Florida, ISBN -0-8493-8788-X. Peter, J,1968, Annals of the New York Academy of Sciences, Volume 152, 283. Saad, MA, 1985, Compressible fluid flow, Prentice Hall Inc, New Jersey, USA. ISBN:0-13-163486-0 Sachs, RG, 1944, The dependence of blast on ambient pressure and temperature, Ballistic Research Laboratory Report 466, pp. 8. Salamon, MDG, Oravecz, KI & Hardman, DR, 1972, ‘Rock mechanics problems associated with longwall trials in South Africa ’, Proc. of 5th Int. Strata Control Conf., London ,NCB, paper 19, pp. 8. Salamon, MDG, Oravecz, KI, 1976, Rock mechanics in coal mining, Chamber of Mines of South Africa, PRD Series No. 198, ISBN 0 620 00232 8. Schaller, S & Hebblewhite, BK, 1981, Rock mechanics design criteria for longwall mining at Angus Place Colliery, Australian Coal Industry Research Laboratories, Published Report, RN: PR 81-3, May 1981 Schrier, S, 1982, Compressible flow, John Wiley & Sons, Toronto, Canada, ISBN 0- 471- 05691. Shepherd, J, 1996, ‘Geotechnical mapping methods for longwall layout’, Geology in Longwall Mining, Proc. Symposium, The Univ. of New South Wales, Australia, 12-13 Nov., pp 127-135, Publ. Sydney. 199 SIMTARS, 1990, Report - coal mine explosion research project, Safety in Mines Testing and research Station, May 1990. Singh, TN & Singh, B, 1982, ‘Design criteria of longwall supports’, Proc., State of the Art of Ground Control in Longwall Mining and Mining Subsidence, SME, June, 1982, (eds.) Chugh, YP & Karmis, M, pp. 39-50. Siskind, DE, Stagg, MS, Kopp, JW, & Dowding, CM, 1980, Structure response and damage produced by ground vibration from surface mine blasting, USBM, Report of Investigation 8507. Smart, BGD, 1989, ‘Recent advances in the science of strata control’, The Mining Engineer, Feb., 1989. Song, YJ & Xu, LS, 1992, ‘Study of the impact of mining under massive roof at Datong coal mines, China’, Proc. 11th International Conference, on ground control in Mining, Wollongong, NSW, 7-10 July, 1992, pp. 540 -547. Standing Committee on Coalfield Geology of New South Wales 1974, ‘Report of subcommittee for the Northern Coalfield - stratigraphic nomenclature’ Rec Geol Surv NSW, Volume 16, no. 1, pp. 9-105. Standing Committee on Coalfield Geology of New South Wales, 1986, ‘Redefinition of coalfields in the Sydney and Gunnedah Basins’, Rec Geol Surv NSW, Volume 22, no. 1, pp. 159-161. Styles P, Bishop I & Toon S, 1992, ‘Surface and borehole microseismic monitoring of longwall faces; their potential for three- dimensional fracture imaging and the geomechanical implications’, Proc., 11th International Conference on Ground Control in Mining, The Univ. of Wollongong, NSW, July 1992, pp. 177-192. Suorivemi, FT, Dusseault, MB & Brummer RK, 1995, ‘Probabilistic risk and reliability evaluation of ground control practices in underground mining, the safety and economic implications’, Proc. 35th U.S. Symposium on Rock Mechanics, 1995, Mackay School of Mines, Nevada. Torabi, SR, 1997, ‘Wind blasts due to roof collapse in underground coal mines’, PhD Thesis, University of New South Wales (unpublished), 1997. 200 Ulrich, S, 1992, ‘Planning coal mine layout with improved safety’, Gluckauf, v53, April 1992, pp. 53-59. Wagner, H, 1983, ‘The design and support of longwall faces in South African collieries’, Proc. of Longwall Mining Design - development and Extraction seminar, ACRIL,11-12 Nov. 1983, Sydney, NSW, Australia, pp.20-40. Walton, RJ, Enver, JR, Litterbach, N & Crawford, G, 1987, Stress measurements at West Wallsend No. 2 Colliery, CSIRO Aust., Division of Geomechanics, Site investigation report no. 36. Warbrooke, PR & Roach, MJ, 1986, ‘The relationship between seam splitting, coal forming environment and coal properties in the Newcastle coal seam’, Advances in the study of the Sydney basin, Proc. 20th Newcastle Symposium, Newcastle, NSW, 16-18 May, 1986, Publ. Newcastle, University of Newcastle, pp. 23-27. Wardel, LJ, 1993, ‘The use of numerical modelling for underground coal mine design’, Comprehensive Rock Engineering - Principle, Practice and Projects, (eds.) Hudson, JA, Brown, ET, Fairhurst, C.& Hoek, E, Pergamon Press, Oxford, pp. 733-748. Wavemetrics, 1991, ‘Igor graphing and data analysis manual -version 1.2’ Wavemetric Inc., PO Box 2088, Lake Oswego, OR 97035, 397 p. Weibull W, 1947, Detonation of spherical charges in air: travel time, front velocity and wave length of emitted wave, Tidskrift for Kustarilleteriet, 5(4), (In Swedish) Whittaker, BN & Singh, RN, 1979, ‘Evaluation of the design requirements and performance of gate roadways’, The Mining Engineer, Feb., 1979, pp. 538-548. Whittaker, BN, 1982, ‘A review of progress with longwall mine design and layout’, Proc. Int. Sym. on State - of - the - Art of Ground Control in Longwall Mining and Mining Subsidence (eds.) Chugh, YP & Karmis, M, SME-AIMMPEI, New York, Chapter.8, pp. 77-84. Wiederman AH, 1986, Air blast and fragment environments produced by the bursting of vessels filled with very high pressure gases, PVP Volume 106, ASME, New York. 201 Wilson, AH, 1982, ‘Pillar stability in longwall mining’, Int. Sym. on State - of - the - Art of Ground Control in Longwall Mining and Mining Subsidence (eds.) Chugh, YP & Karmis, M, SME-AIMMPEI, New York, Chapter.9, pp. 85-95. Wilson, AH, 1986, Keynote address: ‘The problems of strong roof beds and water bearing strata in the control of longwall faces’, Proc. of Sym. on Ground Movement and Control Related to Coal Mining, (ed.) Aziz, NI, AusIMM Illawarra Branch, Australia, August, 1986, pp. 1-8. Wilson, J, Ucciardi, F & Armstrong, W, 1976, Multilift longwall mining methods for thick coal, design and feasibility, report prepared by Consolidated Coal Co. for U.S.B.M, U.S. Dept, of Commerce, NTIS, PB 272-460. Wind Blast Committee, 1991, Report of investigation, Department of Mineral Resouirces, NSW Government, Australia, 16 p. Wind Blast Management Plan, 1998, Moonee Colliery, Coal Operations Australia Limited, unpublished report. Wold, MB & Pala, J, 1986, Aspects of support and strata performance on longwall no. 1 at Ellalong Colliery, GCM Report no. 62, CSIRO, Mount Waverley, Vic., Australia Wu, X & Karfakis, MG, 1993, ‘Mathematical modelling of strong roof beds in longwall mining’, Proc. of 12th Int. Conf. on Ground Control in Mining, (eds.) Peng SS & Holland, CT, Aug. 3-5, 1993. Yeates, RA, Enever JR & Hebblewhite BK, 1983, ‘Investigations prior to the introduction of longwall mining’. Proc. 5th International Congress of the International Society for Rock Mechanics, Melbourne, April, 1983, Volume2, AA Balkema, Rotterdam, pp. E71-E77. 202 APPENDICES APPENDIX A WIND BLAST DATA LOGGER, SAMPLE FILE FORMAT FOR OVERPRESSURE AND VELOCITY CHANNEL, INTERPRETATION AND CORRECTIONS 203 rue 4 1 24 □ATE 5 3 7 98 pan BP 156 ~ PGD2 BP 158 " PGD3 BP 156 “ PGD4 BP 155 “ TEMPERATURE n PCDl VELOCITY 231 Z30 233 233 233 233 233 233 233 233 232 231 233233 231 233 230 233 233 233 237 231 233 233 233 233 232 231 233 231 237 231 233 233 233 233 233 233 233 233 233 233 233 231 233 233 233 233 233 233 233 233 ilSIiiiiiiil 233 231 233 231 231 230 231 234 231 233 230 233 231 233 231 230 233 230 233 231 2§2 231 231 233 231 233 231 233 233 233 237 230 233 231 233 231 233 233 231 233 233 230 233 233 233 233 231 233 233 233 233 233 231 9T3 iiiliiiliiii 233 237 233 231 233 230 233 237 233 233 230 233 231 233 230 233 233 233 234 231 230 231 231 233 230 233 233 111 233 230 233 233 233 233 233 233 230 233 233 233 233 233 231 231 232 237 233 231 230 231 7-3 233 230 233 233 229 233 231 232 231 233 231 233 230 233 233 231 233 233 233 231 ?33 235 ?33 233 233 233 233 233 233 230 231 233 237 233 231 233 233 233 231 233 230 230 233 231 233 230 230 237 231 233 233 231 233 231 231 233 230 233 233 235 230 233 231 230 233 233 233 231 230 237 233 233 229 231 234 233 233 233 233 231 230 233 233 230 231 231 230 233 233 232 230 233 233 233 233 231 233 233 233 237 231 230 233 233 233 233 230 233 230 231 233 231 230 231 933 233 233 233 231 233 230 231 233 230 233 230 230 233 230 233 231 233 233 233 233 233 233 237 233 233 231 233 230 231 233 230 233 233 233 233 233 237 231 231 233 233 232 233 230 233 231 233 2^ gi §2 m §* IS 3* IS Hi % li HI 233 230 233 233 233 237 230 233 233 233 233 230 233 233 233 233 230 233 233 237 233 233 231 233 233 233 233 230 231 231 233 233 233 233 231 233 233 231 232 230 233 233 233 230 230 231 230 233 231 233 233 231 233 233 233 231 233 230 233 233 230 230 233 233 233 231 235 233 233 233 230 233 231 230 233 233 231 231 231 233 233 233 231 231 233 231 233 230 233 Hi 3 £ iii £ £ S £ H? 3 Hi 3 233 233 231 233 231 231 233 231 232 233 230 233 233 233 230 233 233 230 233 230 231 231 233 233 230 233 231 230 233 230 230 230 231 233 233 233 231 233 233 yv* ?33 231 9T3 933 233 999 231 231 ?33 9^1 233 233 233 230 933 233 237 231 231 233 231 233 231 233 231 231 233 230 233 233 230 233 233 233 233 230 233 230 231 233 230 231 231 233 231 230 233 233 233 237 231 233 233 230 230 231 233 231 230 232 233 231 233 230 233 233 233 233 230 231 233 230 233 233 230 233 231 233 231 230 231 230 233 233 233 230 237 231 232 231 230 231 230 233 231 231 231 231 233 232 233 233 233 233 231 233 230 230 233 233 233 233 231 233 233 233 233 233 231 231 231 230 233 233 231 233 231 232 233 231 233 231 233 233 231 233 231 233 233 233 233 233 233 233 233 233 230 231 233 233 233 231 231 237 234 233 233 230 233 233 231 233 231 233 233 233 229 231 234 233 233 233 233 233 230 233 233 233 233 231 231 233 233 233 231 230 231 233 233 231 230 233 233 233 231 233 233 230 233 237 230 233 233 237 233 233 231 233 233 233 233 233 231 230 231 233 237 233 231 233 233 233 233 231 230 233 233 233 233 233 233 233 233 233 231 231 230 233 233 233 230 233 233 231 233 233 231 233 231 233 233 230 231 231 233 233 230 233 233 233 231 233 233 230 233 233 233 233 230 233 233 233 231 232 231 231 233 230 231 233 230 233 237 237 233 230 233 233 230 233 233 231 233 230 233 233 231 233 233 235 233 233 233 233 233 233 230 231 231 233 231 233 231 233 237 237 233 231 233 231 233 233 231 233 233 233 233 233 231 233 231 237 237 233 233 231 233 237 233 233 233 237 238 235 233 233 233 233 237 233 233 231 233 233 231 229 233 231 233 233 233 233 231 233 233 237 233 233 233 229 233 235 233 233 230 233 233 233 233 233 233 233 237 233 233 233 233 233 233 233 233 237 237 233 233 233 237 233 237 233 233 933 233 233 233 233 237 233 233 237 237 237 233 233 237 237 237 233 233 233 233 233 237 231 233 233 233 233 233 233 233 233 237 237 233 237 233 237 233 233 233 231 233 233 233 233 231 233 232 237 233 233 233 233 233 237 233 237 233 237 237 233 235 233 234 237 233 237 233 233 237 233 233 233 237 238 234 237 237 233 233 237 237 237 233 237 233 237 232 232 233 233 237 233 233 233 233 237 237 233 233 233 237 233 233 233 233 233 237 233 237 237 233 237 237 237 232 233 233 233 237 233 233 233 233 233 237 233 237 237 234 237 237 237 233 233 233 237 233 233 233 237 237 233 233 233 233 237 237 237 233 237 237 233 234 233 233 237 233 237 233 233 237 233 237 1 237 231 235 232 233 237 233 233 237 233 237 233 233 237 237 237 237 233 234 237 233 237 233 233 233 233 233 233 233 235 233 237 237 237 237 231 233 233 233 233 237 233 233 237 233 233 233 237 233 233 233 233 233 233 237 233 233 233 233 233 233 233 233 233 233 237 237 233 233 231 233 237 230 233 233 233 233 231 233 237 233 237 229 237 233 237 233 233 233 230 233 230 233 233 233 237 231 234 233 233 237 230 233 233 233 237 233 233 233 232 231 233 233 233 237 233 233 233 233 233 233 231 237 233 233 233 233 233 233 231 230 237 237 233 233 w 229 238 233 231 231 233 233 233 233 231 233 233 237 235 233 237 237 233 233 233 233 233 233 233 237 237 233 233 237 233 233 233 233 237 233 237 233 233 237 233 237 237 232 237 233 237 237 237 233 233 237 237 237 233 237 237 233 237 237 237 237 233 237 233 233 233 233 237 233 234 237 231 237 237 237 237 233 233 237 233 237 233 233 237 233 233 233 233 233 237 233 237 237 237 237 237 237 233 233 237 233 233 237 233 233 237 233 237 237 237 237 233 237 233 233 237 233 237 237 233 233 233 237 237 233 237 233 233 237 237 233 233 233 237 233 233 233 233 237 237 237 233 237 233 233 233 233 237 233 233 233 229 237 233 237 233 237 237 233 233 237 237 237 233 231 237 237 237 233 233 233 233 233 233 237 233 237 238 233 237 237 233 233 233 233 237 231 233 233 233 237 233 233 237 233 237 233 233 233 233 237 233 233 233 237 237 237 237 233 233 237 237 237 233 237 237 233 237 237 233 233 233 233 237 233 237 237 240 237 234 232 235 237 233 237 233 233 237 233 237 237 233 233 233 237 237 233 233 233 233 233 233 231 233 233 233 234 237 237 233 233 237 237 233 233 237 237 237 237 233 233 237 237 233 ?33 233 233 237 237 237 233 234 232 237 233 233 233 233 233 237 237 237 237 237 237 233 233 237 233 233 233 233 231 233 233 237 237 233 233 237 234 237 233 233 233 231 233 233 237 237 233 237 233 233 237 233 233 237 233 233 233 233 237 237 237 233 233 232 234 233 237 233 237 237 237 233 237 237 237 237 233 237 231 233 233 233 233 237 237 233 237 237 233 234 232 237 237 233 237 233 233 237 237 237 237 237 237 236 235 237 231 233 233 233 237 233 233 237 237 237 237 233 235 237 237 237 233 237 237 237 237 237 237 237 234 237 237 237 237 237 237 237 237 237 237 237 237 235 238 233 237 237 237 237 237 237 237 237 237 233 237 237 237 237 233 233 233 237 233 237 237 233 237 233 237 237 237 237 233 237 237 237 233 237 233 233 237 237 237 237 237 237 233 233 237 237 237 237 237 237 237 237 237 237 237 237 237 237 237 233 237 233 237 237 235 233 233 233 233 237 233 237 233 237 237 233 233 237 233 237 233 233 233 233 237 233 237 237 233 237 235 237 233 237 233 233 233 233 233 233 233 237 233 233 233 231 237 237 233 233 230 233 233 233 233 230 233 233 233 237 233 233 231 234 233 233 233 233 237 233 233 237 237 237 233 237 233 233 233 233 237 233 233 233 233 233 233 237 231 233 233 233 237 233 237 233 237 237 233 232 235 233 233 231 233 231 231 233 231 233 237 237 233 237 233 233 233 233 233 233 237 233 233 233 231 237 237 233 204 237 233 237 233 233 233 231 233 233 233 233 233I 233 233 232 235 233 233 233 233 233 233 233 233 233 233 233 233 237 233 233 233 233 233 233 233 237 237 237 233 233 233 230 233 231 233 233 231 233 233 237 237 231 233 233 233 233 233 233 233 233 237 233 231 231 231 233 233 233 233 233 233 232 237 233 237 231 233 234 233 233 233 233 233 233 233 237 233 233 233 233 237 233 233 233 232 237 233 237 237 233 233 237 237 237 235 233 233 237 233 237 233 237 237 237 240 237 237 237 232 233 237 233 233 233 233 237 237 237 237 233 237 237 237 237 237 233 233 237 237 237 238 237 233 237 234 233 234 235 237 237 237 237 237 237 233 237 237 237 237 237 236 235 237 237 233 233 233 237 237 237 237 234 237 237 233 237 233 233 237 237 233 237 237 237 237 237 237 237 237 237 233 235 240 237 237 233 237 237 237 237 233 237 240 237 237 238 237 233 237 237 237 237 233 233 240 237 237 240 237 237 237 237 237 237 237 237 233 238 237 237 237 233 237 237 237 237 235 234 237 233 237 237 237 237 237 233 237 233 232 238 233 237 233 237 237 237 234 233 237 237 233 237 237 237 237 237 237 237 237 240 238 237 234 237 237 237 237 240 237 235 237 237 240 237 237 237 237 237 237 237 237 237 237 237 237 237 237 233 237 237 237 237 237 237 237 237 237 237 237 237 237 240 240 237 235 237 240 237 237 237 237 237 237 237 237 237 235 237 238 237 237 237 237 237 237 237 240 238 235 237 237 237 237 237 237 237 240 237 240 237 237 236 243 240 240 237 237 237 240 237 243 243 241 237 237 240 240 237 237 237 237 237 237 237 235 237 240 237 237 240 237 237 ’237 237 240 240 237 237 240 240 240 240 240 237 240 243 243 240 237 237 237 240 240 243 237 240 243 243 240 241 240 237 237 240 240 240 237 243 243 243 243 243 243 240 243 243 243 240 243 243 247 243 241 247 247 243 247 250 243 243 2BaCKBBKKggBBBKeCBggg -gKgC§eKgSBgg^gKgKggg 240 243 243 247 243 247 247 250 247 2 1 1 247 250 250 1 245 245 247 1 247 £S^BgBgBK§g^BBBgBggg 3 1 243 247 247 3 250 247 250 1 1 1 1 1 3 3 5 6 2 250 3 u, JO,H JM 5 5 5 5 4 6 5 5 6 6 5 5 5 7 7 5 5 7 7 7 7 ' 7 7 7 7 6 6 7 7 7 7 7 7 7 7 8 8 7 7 a 9 7 J 9 9 7 8 11 11 11 9 9 11 11 9 7 8 9 11 8 9 11 9 U 9 11 11 9 9 11 11 11 11 11 11 11 11 11 11 11 n 11 11 11 U U 11 11 11 13 11 10 11 11 11 11 11 11 12 11 U n 11 11 11 12 11 u 12 11 11 11 11 12 12 12 12 13 13 13 12 13 13 13 14 13 13 13 14 14 14 14 14 15 14 14 14 15 14 14 14 14 14 14 14 15 14 14 14 15 14 14 15 15 15 15 15 15 14 15 15 IS 15 15 15 15 15 15 15 16 15 15 16 16 16 16 16 17 17 17 16 17 17 17 17 16 16 16 16 17 16 16 16 17 16 17 17 17 17 17 18 18 17 17 17 17 17 17 17 17 17 17 17 17 17 17 18 17 18 18 18 18 18 18 19 19 19 20 19 19 19 19 19 20 20 20 20 20 19 20 20 20 21 22 22 22 22 22 22 22 22 22 21 21 21 21 22 22 22 22 22 22 23 22 22 23 23 23 23 22 22 23 23 23 23 23 23 23 23 23 23 23 22 23 23 25 25 24 24 24 24 25 25 25 25 24 24 25 25 25 25 25 25 25 25 25 25 25 26 26 26 26 26 26 26 27 27 27 26 26 27 27 28 28 28 27 27 27 28 28 28 28 28 28 28 27 27 28 27 27 28 27 27 27 28 28 28 28 28 28 28 28 28 28 28 28 29 28 29 29 29 29 29 29 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 31 31 31 31 31 31 31 30 30 31 31 31 32 32 32 32 33 32 33 33 33 33 33 33 33 33 32 32 32 33 33 33 34 33 34 34 34 33 33 33 33 33 33 33 33 34 33 34 34 34 34 34 34 35 35 35 35 35 35 36 36 36 36 37 37 37 37 37 37 37 37 37 37 37 38 38 38 37 37 38 38 37 37 37 38 38 38 38 38 38 38 38 38 38 38 39 38 38 38 39 39 39 39 39 39 40 40 40 40 40 41 41 ..41 41 41 41 41 41 41 41 41 41 41 40 41 41 40 40 41 40 42 41 41 41 42 41 41 41 42 41 41 41 42 41 41 41 41 42 42 42 42 42 42 42 42 42 42 42 43 42 42 43 43 43 43 43 43 43 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 44 45 45 45 44 44 44 44 44 44 44 44 44 44 45 44 45 44 44 44 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 46 47 46 47 47 47 47 48 48 48 48 48 48 48 47 47 48 48 48 48 48 48 48 48 48 49 49 49 SO 50 49 48 49 49 49 49 49 49 49 50 50 51 51 51 51 51 51 51 51 51 50 50 50 51 51 51 51 51 51 51 51 51 52 51 52 51 51 51 52 52 52 52 52 52 52 52 52 51 52 52 52 52 52 52 52 52 53 52 S3 S3 S3 53 53 53 53 53 53 53 53 53 54 54 53 S3 54 54 55 55 55 55 55 55 55 55 55 55 55 55 55 55 56 56 56 56 56 57 56 56 56 56 56 56 56 56 56 56 56 56 57 56 56 57 57 57 57 57 58 58 58 58 58 58 58 58 58 59 59 59 59 59 59 59 60 60 60 60 60 61 61 61 61 62 62 62 62 62 62 62 62 63 63 63 64 64 63 63 63 63 62 62 62 63 63 63 64 64 64 64 63 63 63 64 64 63 62 62 * 62 63 63 64 65 65 65 64 62 59 59 59 62 65 67 67 66 65 64 65 67 67 68 67 68 68 67 67 67 67 67 68 67 67 67 65 61 60 61 62 62 62 62 62 61 60 59 58 57 58 61 63 64 64 64 65 64 64 64 64 64 63 61 58 58 58 61 62 63 62 61 61 62 62 62 63 65 66 67 66 65 64 64 65 65 66 67 68 68 68 69 69 69 69 70 70 70 69 68 67 66 65 65 65 67 67 67 67 67 67 67 67 68 69 70 69 68 67 67 67 69 70 70 70 70 70 70 69 69 64 57 55 56 60 64 65 66 66 67 68 68 68 69 70 70 67 66 67 69 70 71 71 71 70 70 71 71 71 71 72 72 72 72 71 69 68 67 66 64 63 62 64 65 65 64 65 65 65 67 68 68 69 69 69 69 69 69 68 69 69 69 68 67 67 67 67 66 66 65 64 64 64 64 64 65 65 65 65 65 64 64 65 64 64 64 64 64 64 64 64 64 64 65 66 66 67 68 68 68 68 67 67 67 67 68 68 67 67 66 67 67 68 68 68 68 68 68 68 68 68 66 64 62 60 57 57 57 56 56 57 57 59 63 68 70 70 71 71 71 71 71 71 71 71 71 72 74 73 74 74 73 73 73 73 73 72 72 73 72 72 72 71 71 71 71 71 71 73 72 72 72 72 72 73 73 72 72 73 72 72 72 72 72 72 72 72 72 72 72 73 74 74 76 76 75 73 72 72 72 72 73 73 74 75 76 77 79 80 83 86 86 85 84 84 84 85 86 86 86 86 85 85 84 83 82 81 80 80 80 80 75 74 73 71 70 69 67 66 67 66 64 64 64 65 67 69 72 76 81 85 87 76 76 77 78 80 80 77 77 80 78 76 77 78 79 82 83 84 84 85 88 92 91 91 92 93 95 98 98 97 95 94 94 91 88 85 84 86 86 87 86 87 90 100 95 90 88 84 80 82 84 82 80 86 95 97 97 95 93 91 84 76 67 60 69 70 70 70 73 78 82 82 84 86 89 91 93 95 97 97 96 94 92 91 92 89 91 93 92 90 89 88 87 85 81 78 76 73 72 72 73 76 81 81 79 75 69 70 68 68 72 73 74 73 69 65 66 69 69 70 71 70 68 67 67 67 67 84 78 76 75 70 65 64 64 64 64 63 63 60 57 55 54 53 53 53 59 67 81 79 75 69 67 70 74 80 83 85 85 83 84 87 91 90 88 86 85 84 84 88 89 90 91 89 85‘ 80 78 77 82 84 88 90 89 86 82 78 76 75 74 75 74 PiP3iiS8!SSi38S3KJdCliCai3!SSK}a83eai!32S32£BBiSSS;S:C!tifigJSSfe{ttfc}y8SSKKBE2£C:c:5:GtJti»'' 80 76 73 73 73 74 75 76 75 75 74 74 75 78 80 77 73 72 73 73 68 61 56 51 47 46 48 51 55 62 64 64 63 58 55 51 46 45 47 54 57 50 52 55 56 56 56 56 56 59 62 64 65 64 64 65 68 70 71 70 70 70 CSSgKJ3£S832Sg3S^SK{p}2;g8£3PS288gPS£85S££aBgS5:£S;££fiSgggfc5ByBBK«BBg2S:SK5:t:c:t;"- 77 77 77 77 76 75 74 75 74 74 74 76 76 77 7B 76 72 68 66 67 67 Spits fj2t5dP3S£8R0S!3BS8Kirfs!S3£8!SSPl83SPaQ0?ig22t3i2gSggetti!5S28S!llM3}S!SSKKt3^S3S:s;d;i;KPP' 74 76 77 75 70 63 57 53 52 58 64 69 74 76 80 81 83 84 86 87 86 I 205 K2BCS!g!6!yyS2!3S:i5iSiSI)!g}st:«BKCBPKSSSBKS!3»fiSg«?JSi2!3C;a5i!65gi5I2iCSgSKiCB38S3ffi?ig888g:S8?8^S5S?!S2?i^88SSjeslS2»8g ^KKGyykSy!s!SSa!K!j;yb!lakj ,"l3*MKKt!K!jmSfciyo{yyf:4gg66gig£tei^i5ya54CSlcg8S!otjSS^08looS2a8!}Sg®Spi882g8£S£S8ICSol8o»8£13KQ B^gPKyfcjyisKKysiiijiiJ^iiPwtsBKycKayfeiwysaiiSisiigicessfisgyfscsisgfiaxsgsgBK^ssas^s^s^sEdBrfsisegssais^sgdidiffiifcss K'°K'0K®wySK5f^y}5{l3il(;SbiGyKKSSC{j!jKSiofcjyKS!5gSgfiMgggSt!gfiSKggS2ie3ijR!5Mo3BSo3S!ci2!3ISCdS5!SdSBSgg3i2si!j!gijS ja g g ^ g t i Ga K'ytoBo!gKK3yy22gy®!DCjKKS*t3S>Kt32yB«fct|5t35fi63l5Ht»C^3CSo®g2l3Q$3B8!!2Sa?ia?l2ffl^2l28£dsaas!i5S!SSffi5a^«*,-riSS2;??2t;'»i* : 24 220 222 223 15 29 32 25 36 1 18 35 21 31 42 33 38 22 23 33 35 41 28 24 23 222 22 18 27 24 36 26 25 18 34 29 34 41 39 44 46 42 43 46 42 34 52 50 42 26 50 40 50 40 17 52 42 70 55 55 63 62 50 62 73 65 69 64 63 69 64 53 70 76 84 63 77 61 64 80 81 80 85 86 95 76 73 73 64 92 90 99 95 96 90 75 44 81 66 88 ioi 95 . 19 26 233 14 29 32 22 25 34 18 35 35 30 36 32 41 39 25 28 23 22 43 33 23 20 25 35 25 526 28 25 36 29 34 42 44 40 45 45 43 46 40 40 33 38 42 17 24 48 52 42 52 70 54 50 62 55 72 60 62 64 55 969 69 65 69 70 62 64 83 77 84 61 80 78 260 60 62 82 85 64 93 91 98 77 94 089 90 81 76 68 89 43 89 99 27 14 32 3 224 230 29 22 19 27 26 26 27 29 43 32 36 31 31 31 39 41 28 28 23 23 22 20 21 20 28 26 24 37 37 29 243 33 42 44 43 41 33 44 46 37 46 42 37 42 36 22 52 40 56 43 55 51 18 59 44 S3 57 60 70 72 61 47 48 63 71 71 57 65 68 60 63 60 60 76 79 80 78 81 89 84 79 64 89 98 93 73 94 81 71 77 89 77 44 83 98 26 222 35 29 14 19 31 22 32 93 39 39 39 28 19 21 442 44 35 31 33 23 41 12 122 21 22 21 21 28 27 627 26 21 18 37 35 37 37 34 21 31 33 23 32 ■23 43 41 45 45 46 35 35 51 51 51 45 42 43 51 19 547 45 59 59 55 70 58 26 61 62 62 74 61 64 59 68 59 67 60 63 77 77 60 77 82 81 86 85 60 82 88 67 95 83 95 95 89 75 76 81 89 72 49 98 26 224 15 30 30 31 33 224 19 34 22 22 26 31 53 33 35 35 22 34 31 41 29 23 27 615 22 16 29 28 31 36 43 43 42 34 46 45 45 45 39 43 39 23 39 242 42 58 920 19 58 51 58 57 58 46 76 63 64 67 60 75 71 59 67 62 66 63 63 164 61 61 81 73 82 74 263 62 484 84 76 83 72 483 84 87 84 979 89 95 89 88 76 94 282 82 76 70 59 74 91 97 26 16 224 17 3 233 230 34 17 31 27 31 29 40 33 33 32 31 30 41 23 27 915 19 18 25 27 36 36 32 19 42 44 748 42 47 44 35 43 45 23 44 41 41 38 57 59 51 52 858 58 59 59 57 46 66 58 74 70 66 66 63 61 64 82 82 75 87 475 74 86 84 94 76 75 92 94 88 69 91 97 76 870 68 25 19 222 31 30 34 27 33 37 23 32 33 32 40 40 30 26 17 23 28 29 27 13 35 36 35 42 18 45 42 35 42 46 45 43 42 51 21 23 39 56 59 41 41 56 51 58 57 59 60 46 56 63 75 67 69 69 80 64 66 82 63 64 64 74 84 75 86 81 87 72 85 77 93 92 87 81 75 93 78 69 90 97 206 25 233 29 19 227 16 31 33 36 25 34 33 28 33 33 30 31 33 29 40 22 22 27 41 12 13 28 26 37 17 35 36 18 35 42 45 41 48 22 42 40 35 45 40 39 23 40 47 38 58 55 48 51 59 59 57 57 60 59 55 64 46 62 73 67 76 64 64 80 79 67 68 72 84 75 85 64 72 62 88 87 80 74 86 90 92 77 81 73 70 90 74 92 97 67 3 233 237 25 225 28 31 25 433 34 18 16 36 34 32 29 33 40 31 12 231 32 34 29 142 41 22 21 27 26 28 37 17 15 35 19 39 30 35 38 44 39 47 39 41 39 848 48 35 46 25 36 39 755 57 22 42 45 61 S3 50 57 57 56 56 62 66 45 60 77 60 78 869 68 70 64 66 65 64 77 79 74 75 61 84 83 864 68 79 89 88 SO 92 77 90 85 86 71 74 80 71 94 98 361 63 25 230 27 18 33 16 31 36 27 33 35 33 30 32 40 28 22 25 30 30 34 17 28 26 39 21 22 19 35 14 33 38 39 46 38 44 40 35 37 45 39 45 41 28 23 33 52 50 59 60 54 55 59 64 65 43 56 63 79 78 71 71 64 65 65 79 63 60 76 81 78 74 85 83 86 85 78 86 86 84 75 79 92 82 85 70 92 76 391 93 79 75 796 97 <->46 23 3 230 230 18 233 27 31 17 32 37 28 34 30 33 31 34 41 28 33 22 132 31 42 23 25 21 28 26 38 17 19 35 21 32 31 40 43 39 45 35 37 35 40 35 49 39 37 25 03 29 30 30 30 53 58 50 46 50 55 59 59 55 52 42 71 79 563 65 62 64 76 64 59 64 65 62 76 81 86 86 80 71 81 76 60 87 84 79 75 74 78 86 94 90 71 71 78 60 23 233 19 32 27 32 17 33 39 29 28 33 30 33 34 28 42 34 21 21 21 24 23 242 42 19 21 28 26 39 21 30 35 32 29 40 50 41 41 43 36 35 35 040 40 34 46 47 39 26 28 51 54 56 52 452 54 846 48 59 55 60 42 72 61 64 75 77 75 64 75 587 85 63 64 69 62 82 58 58 60 69 77 89 77 89 82 87 80 596 95 93 080 80 81 91 61 322 23 21 227 231 27 33 616 16 31 31 28 33 29 29 33 32 28 41 35 33 23 20 20 24 30 41 17 20 27 37 39 22 23 25 32 34 28 40 41 43 42 49 34 50 35 32 25 47 45 42 25 55 56 254 52 55 59 61 63 56 42 73 74 66 67 67 60 66 60 74 64 65 63 67 60 60 82 86 80 68 86 77 82 78 92 92 88 76 73 81 89 93 67 80 .32 21 225 231 26 32 33 28 31 345 31 43 32 33 32 28 30 25 536 35 242 42 42 21 23 22 26 29 30 27 24 34 21 39 27 42 45 42 33 41 36 36 45 29 56 322 23 43 46 31 24 55 57 750 47 59 152 51 46 60 59 76 73 62 74 66 64 76 59 65 59 66 60 60 85 84 78 85 67 79 88 93 84 78 94 97 87 76 81 76 74 79 93 88 81 22 21 225 26 230 31 27 34 29 16 928 30 29 27 26 42 28 21 22 23 21 23 26 34 23 22 15 35 38 33 21 33 27 242 42 47 42 232 42 32 37 39 39 846 48 41 29 31 45 48 22 55 56 56 59 58 59 55 53 50 67 465 64 68 77 74 71 74 S3 59 67 77 83 83 66 64 69 83 76 67 90 88 61 80 93 80 94 96 78 76 686 86 82 79 79 79 87 84 93 78 225 22 22 230 26 31 35 28 15 43 30 26 28 34 25 26 37 25 41 21 26 41 22 14 22 34 22 34 37 28 48 43 38 43 38 28 55 46 49 36 25 54 52 58 57 19 53 53 70 76 67 66 66 69 70 58 76 63 60 85 82 80 74 78 67 64 75 93 95 79 82 95 78 82 86 79 91 82 BBggKSyfe}ft}SSi35Sfc!iJgf!KtlliiBBBI2G8SB5s:t!!i!Bfc!y6fiC5gtig!glJigg8S5?5tJSB»ifSSlSSSISBtS838Di8c!RS8^B8a8?!s!SSi8liJl5i3SSS83SPBS?J8S!a KBKgC38iJ!?B3gS8bl^!gfiKBKBtiKBtS3Kfcih!!f!CSyfc{C3figbig|5S:igSSCSfcigSBaS8gK«!»88g5!£d388^l58Ji8^c!firf$^8»?liS80SB8BeiSB8^ BBKKt;S8l3lis{iJB8WablbiSg3SSII3t3Kl3PB33Bfc!biKe3egl!S8lsSg336g8WCC;Bg3gl5a8^B!3B2^rfcS£a838^88^2S5!c tJKGKi:S8bigb!SSkiSb;bl3g85:SGKKKP5;K8ytti;Kyai5g?gS£g3Cg82CSgi5S;gC3gS£SBCSdSari288S3!£^Birfj^cie5Si2S?2ffiS2dSS3a 23 23 23 23 22 22 22 21 21 21 20 20 19 19 18 18 18 17 17 17 17 18 18 17 ’l5 14 12 11 11 7 7 7 7 7 7 7 6 5 7 7 6 6 7 7 8 9 9 11 11 12 13 15 15 15 15 16 16 17 17 17 16 16 15 14 13 13 13 13 11 12 13 13 14 13 14 14 15 15 15 16 17 19 19 19 19 19 19 22 23 25 25 26 25 23 22 21 21 20 20 21 21 22 22 22 22 23 23 23 23 23 22 22 21 21 20 19 19 19 19 18 18 17 17 16 16 15 15 15 15 15 15 15 15 13 14 14 15 15 15 14 14 14 15 16 16 17 16 16 16 16 15 15 15 14 14 13 13 13 11 11 11 11 11 11 8 n 11 12 12 12 12 13 12 12 12 12 12 12 12 12 12 13 13 12 11 11 11 8 7 6 250 243 237 238 238 243 243 237 243 237 240 237 240 243 243 243 243 247 243 240 237 237 237 233 233 237 240 243 243 250 4 7 5 5 6 5 5 3 1 243 243 243 243 240 233 237 231 233 237 237 233 233 230 233 233 233 237 237 243 243 1 3 1 5 5 7 7 7 9 9 8 7 7 9 8 9 9 11 11 11 11 9 11 9 9 10 11 11 11 11 9 8 7 7 5 5 5 3 7 5 5 6 7 7 7 5 6 7 7 7 9 8 8 9 9 8 8 7 5 7 7 7 7 5 7 7 7 7 7 5 5 5 5 5 7 6 6 7 7 5 3 3 5 5 5 6 5 5 5 6 5 113 7 4 4 3 1 3 3 1 250 245 247 247 250 243 250 245 250 250 243 243 247 243 250 243 243 1 243 247 5 243 243 243 243 243 240 240 237 240 243 243 243 243 236 243 247 247 240 240 243 240 243 243 243 243 240 243 238 243 240 243 243 240 240 238 237 243 240 243 243 237 237 237 240 240 237 237 237 237 240 240 237 240 243 238 238 240 243 247 247 1 1 1 15 5 3 3 5 5 7 5 5 3 3 5 3 5 250 1 1 1 S 5 5 5 5 5 7 7 7 8 7 7 11 11 8 8 8 8 7 5 5 5 7 7 5 5 3 5 6 5 5 5 3 5 5 5 5 5 5 5 7 5 5 5 5 5 5 6 5 7 7 7 11 6 9 8 7 7 7 7 8 7 7 7 7 7 7 8 7 7 8 7 7 8 7 7 6 7 8 8 8 7 7 8 9 8 7 8 9 8 9 7 8 8 8 8 7 7 7 7 8 7 9 7 7 4 7 7 5 7 7 7 8 7 7 7 9 9 9 7 7 7 7 7 7 7 7 7 7 7 250 7 7 7 5 5 3 250 250 3 1 1 3 5 5 2 4 6 5 5 5 5 3 1 1 250 250 5 3 6 5 5 7 7 8 7 8 10 8 8 7 7 7 7 7 5 7 7 7 7 5 7 7 7 9 7 9 7 9 9 B 8 7 7 7 7 7 7 7 7 a 9 8 7 7 7 7 8 9 9 11 11 9 9 11 11 11 U 11 11 11 11 11 11 11 9 u 11 11 11 9 11 11 11 11 11 11 9 9 11 11 11 9 9 11 11 9 9 9 11 11 n 11 9 9 9 9 9 7 7 7 8 7 8 7 7 7 7 7 7 5 5 5 5 7 5 5 s 7 6 5 5 7 5 7 6 5 5 5 7 7 5 5 6 6 7 6 6 6 5 7 7 7 7 6 6 7 7 7 5 7 7 7 7 5 7 8 9 8 11 6 a 7 7 7 7 8 7 7 8 7 250 7 9 9 9 9 8 8 11 9 9 11 11 11 11 9 li 8 9 U 9 9 9 8 9 9 9 9 9 8 11 11 9 9 a 9 11 11 11 11 11 li 11 9 9 9 11 9 9 11 8 U U 11 11 11 11 11 11 10 9 8 9 11 11 U li 11 11 11 10 12 10 11 9 11 U 8 9 9 11 11 11 11 u 11 11 11 11 11 9 u 9 11 11 9 9 9 9 9 8 7 8 9 8 7 7 8 7 250 7 7 7 8 7 7 7 7 7 7 9 7 8 7 11 U 8 9 7 9 8 9 9 9 11 11 11 U 9 11 10 11 9 9 9 9 11 9 11 250 9 9 11 9 9 9 7 8 9 8 8 8 7 8 8 9 7 7 7 7 7 7 5 7 7 7 7 5 7 5 5 5 5 5 5 3 5 5 2 3 3 4 5 5 3 1 3 13 5 1 1 1 7 2 3 1 1 1 1 247 247 1 247 250 1 1 3 247 2 3 245 243 243 247 243 243 243 243 243 236 243 250 243 243 243 243 247 245 243 250 1 247 243 241 250 247 247 247 247 1 247 3 5 7 6 7 9 9 11 11 9 9 9 7 7 6 6 5 250 7 7 7 5 5 7 7 7 7 7 8 8 8 8 7 9 10 8 8 8 8 9 8 8 8 9 8 8 8 9 7 7 7 8 7 7 7 7 250 7 7 7 6 5 5 5 5 5 1 5 6 5 5 5 4 6 5 5 5 5 5 3 3 5 5 3 7 2 3 247 247 247 1 250 243 247 1 250 250 250 247 243 250 1 247 247 1 1 1 3 250 250 3 1 5 113 1 1 5 112 250 250 250 1 245 2 1 1 250 1 250 247 3 1 250 1 245 14 5 1 250 5 3 3 5 1 3 3 5 3 247 247 1 1 1 243 243 240 243 243 243 243 237 240 241 238 245 2*3 243 247 1 3 3 3 5 7 8 7 7 7 ' 6 7 7 5 5 5 5 5 6 3 3 6 5 6 5 3 5 5 5 5 3 2S0 250 1 1 241 247 247 247 247 247 243 245 243 1 5 1 3 6 3 3 1 1 250 247 1 1 3 3 3 5 4 7 5 5 7 7 7 7 7 7 7 6 6 7 7 7 7 7 7 5 5 6 5 7 U 7 7 8 8 7 8 8 7 7 9 7 7 7 7 8 7 7 5 5 5 6 7 5 7 7 5 5 6 1 5 250 3 1 245 245 1 250 250 247 243 247 241 3 247 247 243 247 243 243 243 243 1 243 240 243 241 240 243 240 243 243 237 243 243 240 247 243 243 240 247 245 243 247 243 243 243 243 243 240 1 243 243 236 243 240 243 243 243 243 237 237 243 243 247 243 243 241 238 250 240 240 243 240 243 243 240 243 240 240 250 243 238 240 237 243 243 240 240 240 243 247 243 243 243 243 247 243 243 247 243 247 250 250 243 243 243 245 243 1 3ATA VALID PCD1~PRKSSURE 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 1S6 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 1S6 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 1S6 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 207 5 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 356 156 156 156 156 156 156 156 156 156 356 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 356 156 156 156 356 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 1S6 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 356 156 156 156 356 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 1S6 356 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 1S6 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 357 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 157 157 156 157 157 157 157 156 156 156 156 156 156 157 157 156 157 156 156 156 157 157 156 156 156 156 156 156 157 156 156 156 157 157 156 156 156 157 357 156 156 156 156 157 156 156 156 157 156 157 156 156 157 156 157 156 157 156 157 157 157 157 156 156 157 157 156 157 156 156 157 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 157 156 156 156 156 156 157 157 356 156 156 157 157 156 156 156 158 156 156 156 156 156 156 156 157 156 156 156 156 157 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 356 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 157 157 156 156 156 356 157 156 156 156 156 156 156 156 157 157 156 157 157 156 157 156 157 157 156 156 156 156 157 157 156 156 157 157 157 157 156 157 157 157 157 157 156 156 156 156 157 156 157 156 157 157 157 156 156 157 157 156 156 157 156 157 157 157 157 156 157 157 157 157 156 156 157 157 157 156 156 157 157 157 356 156 156 157 157 157 156 156 156 156 157 157 156 157 157 156 157 157 157 157 157 157 157 157 157 357 157 157 156 157 157 157 157 157 156 356 157 157 157 157 157 157 156 357 157 156 157 156 157 157 157 357 157 157 157 157 156 157 157 157 356 156 156 157 157 157 157 157 357 156 157 157 157 156 156 157 357 157 156 157 157 157 157 157 157 157 156 157 157 157 157 157 157 357 157 156 157 157 157 156 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 1S7 157 156 156 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 357 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 357 157 157 157 157 157 157 157 157 157 357 157 157 157 157 157 157 157 157 157 156 156 157 157 157 357 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 357 157 157 157 157 157 157 157 157 157 157 157 157 157 157 357 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 158 157 157 157 157 157 157 157 157 157 157 157 158 157 157 157 157 158 158 157 157 157 157 158 158 157 158 158 158 158 158 158 158 157 157 157 158 157 157 158 157 157 157 158 158 158 158 158 158 158 158 158 158 158 158 158 158 153 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 157 157 158 158 158 158 158 158 158 158 158 158 158 157 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 157 157 158 158 158 158 158 158 158 158 158 158 157 157 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 358 158 158 158 158 158 158 158 158 158 158 158 158 158 158 1S8 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 1S8 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 159 158 159 159 158 158 158 158 158 15B 158 158 159 159 159 159 159 159 159 159 159 158 158 158 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 359 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 359 159 159 159 159 159 159 159 359 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 359 159 159 159 159 159 159 159 159 159 160 160 160 159 159 159 159 359 160 159 159 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 360 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 161 160 160 160 160 160 161 161 160 161 160 161 161 160 161 161 160 161 161 160 160 161 161 161 161 161 160 161 161 161 161 161 161 161 160 160 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 160 160 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 161 208 6 in 161 i6i in 161 161 in 161 161 in in i6i in i6i in in in in in in in 161 in 161 in i6i in in in 161 161 161 161 in in in 161 161 161 161 161 161 161 in 161 in in in in in in in in 161 161 i6i in in in 161 161 i6i in i6i in i6i i6i in in in in in in 161 161 i6i i6i in in i6i i6i in 161 in 161 in in in i6i 162 162 in in in in in 161 in 161 in 161 in in in 161 162 162 1 62 162 161 162 162 162 162 162 162 162 161 162 162 162 162 162 162 162 162 162 162 162 1 62 162 1 62 162 162 162 162 161 162 162 162 162 162 162 1 62 162 1 62 162 162 162 162 162 162 162 162 162 1 62 162 1 62 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 1 62 162 162 162 162 1 61 in 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 162 1 62 162 1 62 162 162 162 162 162 162 1 62 162 162 in 162 162 162 162 162 162 162 162 162 162 162 162 162 161 162 162 162 162 162 162 162 1 62 162 162 162 in 161 162 162 in 161 162 162 162 162 162 162 1 62 162 162 162 162 162 162 162 161 in 161 162 162 162 161 162 162 162 162 in 162 in in 162 162 1 61 162 161 161 in 162 161 in 161 162 162 in 161 in 161 in in in 161 in 161 in in in i6i i6i i6i in in 161 in 161 in 161 in i6i in in in in in in in in 161 i6i 161 in in in in in in 161 161 in i6i in i6i in i6i 160 i6i i6i in i6i in in in in 161 in in in in 162 ieo in in in i6i i6i in in in in in in in in 162 in in in in i6i in i6i in in i6i in in i6i 161 i6i in in in in in in i6i in in i6i in in in 162 160 in in in i6i i6i i6i in iei 161 in in i6i 160 in in i6i in in i6i in in in in in 161 in in in in i6i in in in in in i6i in in i6i in 162 in in in in in in 161 in i6i i6i in in i6i i6i in in i6i in in i6i in i6i in in in i6i in 162 160 in in 161 in in in in in in i6i in in i6i in in in in i6i in in in 161 in i6i 161 in in in 161 i6i in in in i6i i6i i6i in in 161 in i6i in i6i 160 in in 161 in in 161 161 in in in i6i in 160 in i6i i6i in i6i i6i in in in in i6i in in in in in i6i in 161 in i6o 160 in in in in i6i 160 in in 160 160 in in in 160 160 in in i6i in 160 160 161 in i6i in 160 in 160 in i6o 160 160 in in 160 160 160 160 160 161 160 160 1 60 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 1 60 160 160 159 in 161 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 161 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 1 60 160 1 60 160 1 60 160 160 160 1 60 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 160 160 160 1 60 1 60 160 1 60 160 160 159 159 159 160 159 159 159 160 160 159 159 159 160 160 160 159 159 159 159 159 159 159 159 159 160 159 159 159 159 160 160 159 159 159 159 160 159 159 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 158 158 159 159 159 159 159 158 158 159 159 159 159 158 158 159 159 158 158 158 158 159 158 158 158 158 158 159 158 158 158 158 158 158 158 158 159 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 157 159 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 159 158 159 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 159 158 158 158 158 158 158 159 158 158 158 159 158 158 158 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159-159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 159 159 160 159 158 159 159 159 159 159 159 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 159 159 160 159 158 159 159 159 159 1S9 159 159 159 159 159 160 160 158 159 159 159 160 159 159 160 160 160 160 159 160 158 160 160 160 159 159 159 159 160 160 159 160 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 159 in 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 159 161 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 in 160 160 160 160 160 160 160 160 160 160 160 160 in 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 in 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 160 160 160 160 160 160 160 160 160 160 160 159 161 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 159 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 160 160 1 60 160 160 160 160 160 160 160 160 159 160 160 160 160 160 1 60 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 160 160 160 160 1 60 160 160 160 160 160 160 160 159 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 1 60 160 160 160 160 1 60 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 160 159 160 160 159 159 160 160 159 160 160 160 160 160 159 160 159 159 160 160 159 159 160 160 160 160 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159' 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 160 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 159 158 159 159 159 159 159 159 159 159 159 159 159 159 158 158 159 159 159 159 158 158 158 159 159 159 1S8 158 158 159 159 158 158 158 158 158 158 158 158 158 159 158 159 158 158 1S8 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 159 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 157 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 157 158 158 158 158 158 158 158 158 158 158 158 158 158 158 209 7 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 158 157 157 158 158 158 158 158 157 157 158 158 158 157 157 157 158 158 157 157 157 157 157 157 157 157 157 158 157 158 158 157 157 158 157 157 157 157 157 157 158 157 157 157 157 157 157 157 157 157 157 157 157 157 157 158 157 157 157 157 157 157 157 157 157 157 157 157 158 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 158 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 156 156 157 157 157 157 157 156 157 156 157 157 157 156 156 157 156 157 157 156 157 156 156 157 157 157 156 155 156 157 157 157 156 156 157 157 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 157 157 157 156 157 156 156 157 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 ,156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 157 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 157 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 157 156 156 156 157 156 156 156 156 156 157 157 156 156 157 157 157 157 156 156 156 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 158 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157,157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 3? 157 156 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 IS 7 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 157 156 156 157 157 157 156 157 156 157 157 157 157 157 157 156 157 157 157 157 157 157 156 156 156 157 157 156 156 156 157 157 157 156 156 156 157 157 157 156 157 156 157 157 156 156 156 157 156 156 156 156 156 157 156 156 1S6 157 156 157 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 1S6 156 156 156 1S6 156 156 156 156 156 156 156 156 1S6 156 156 156 156 156 156 156 156 156 156 156 156 156 156 157 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 U56 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 156 155 156 156 1S6 156 156 156 156 156 156 156 156 156 155 156 156 156 156 156 156 156 156 156 156 155 156 155 156 156 155 155 156 156 156 155 156 155 156 156 156 156 156 156 156 155 156 155 155 156 156 156 156 155 155 156 156 155 155 155 155 155 155 155 155 155 156 156 356 156 155 155 156 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 156 155 155 155 156 155 1S5 155 155 155 155 155 155 155 155 155 155 155 155 156 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 1S5 155 155 155 155 155 155 155 155 155 155 155 156 154 155 155 155 155 155 155 155 155 155 155 155 156 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 154 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 154 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 154 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 154 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 154 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 156 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 210 8 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 LS5 155 155 155 155 155 156 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 156 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 1S5 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 156 154 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 155 DATA VALID pco2''VELocrnr 19 70 21 21 21 22 21 21 21 22 22 21 22 21 22 22 22 21 22 22 23 22 22 22 22 22 22 22 22 21 ZL 21 22 21 21 21 21 21 20 20 20 20 20 19 19 19 18 18 17 17 16 15 15 15 15 15 13 13 13 12 13 12 n 11 7 8 7 7 5 5 250 250 243 245 236 237 237 233 230 230 230 230 230 224 225 222 224 223 222 222 219 219 219 219 219 219 219 219 219 215 215 216 215 215 215 215 215 213 215 213 213 212 211 213 211 212 213 211 211 211 211 212 211 211 211 211 211 211 211 207 211 209 209 208 208 209 208 209 208 208 209 209 209 209 208 208 208 208 206 207 206 206 208 208 208 208 208 208 208 208 208 208 208 208 209 211 209 209 211 211 215 212 215 216 219 220 222 222 230 227 230 233 233 240 243 1 7 7 9 11 U 13 13 15 16 16 17 18 19 19 20 21 21 21 22 22 22 23 23 23 23 23 23 22 23 23 22 23 22 22 , 22 22 21 21 20 19 19 18 16 16 15 14 13 11 11 9 5 247 240 234 230 230 224 222 216 215 213 211 209 206 204 201 198 197 196 194 191 191 189 187 186 184 184 181 180 179 178 176 175 173 173 172 170 170 169 169 168 166 166 165 164 164 162 162 161 161 160 159 158 157 157 157 157 156 156 155 154 154 153 153 153 153 153 152 152 150 150 150 150 150 149 149 148 147 147 147 146 145 145 145 144 144 143 143 142 141 140 140 139 139 138 138 138 137 136 136 135 134 134 133 133 133 133 132 131 132 130 131 130 130 130 130 130 130 129 128 129 129 129 129 128 129 128 129 129 129 128 129 130 130 130 130 130 131 131 132 132 132 132 132 132 133 132 133 133 134 133 134 134 134 134 134 134 133 133 134 133 133 132 132 132 132 132 132 132 132 131 132 132 132 132 132 132 132 132 132 132 132 132 132 133 132 132 132 133 133 132 133 133 133 133 133 133 134 133 134 134 134 134 135 136 136 136 136 136 137 137 137 137 138 138 137 139 139 139 140 140 141 140 140 140 141 141 141 141 141 141 142 141 142 142 141 141 141 141 141 140 140 140 140 139 138 137 137 137 136 136 134 134 134 133 133 132 131 130 129 129 129 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 129 130 129 130 130 132 132 133 134 135 135 136 136 137 138 138 139 140 141 142 142 144 145 146 147 148 150 150 151 152 153 154 155 156 157 158 159 159 162 162 164 165 166 166 168 168 170 171 171 173 174 175 176 177 178 179 181 183 184 185 185 186 187 188 188 189 191 191 191 190 191 191 191 193 191 191 191 191 189 188 188 186 IBS 184 184 184 180 180 179 178 178 176 175 174 174 173 173 173 172 172 172 172 173 172 173 173 172 173 172 172 171 172 172 171 172 171 171 172 172 172 172 172 172 173 173 173 173 173 172 173 173 173 173 173 174 175 173 175 175 174 175 173 173 173 173 173 173 173 172 173 173 173 173 173 173 174 174 174 173 173 173 173 174 173 174 174 174 175 174 175 174 174 174 175 174 173 173 173 173 174 173 173 172 173 172 172 172 172 172 172 173 172 173 172 172 172 172 172 172 171 172 171 171 171 170 171 170 170 170 169 170 169 169 169 169 169 169 169 168 168 169 168 168 169 168 168 168 168 168 168 168 168 168 168 168 166 168 166 166 166 166 166 165 165 165 165 165 165 164 164 164 164 164 165 165 164 168 164 166 165 165 166 166 166 166 168 168 168 168 169 168 169 169 169 169 169 169 169 169 169 169 169 169 169 170 169 169 169 169 170 169 169 169 169 169 169 169 169 169 169 169 169 169 169 169 169 171 171 171 171 171 172 173 173 173 173 174 175 175 176 176 177 177 177 178 177 178 178 178 179 179 179 179 180 180 180 180 181 182 182 183 181 181 181 181 182 181 181 181 181 182 181 181 181 181 183 182 183 184 184 184 184 185 186 184 187 187 188 188 189 190 189 191 191 191 191 191 193 192 193 191 193 193 193 193 193 193 193 194 194 196 194 194 194 194 194 194 194 193 195 194 194 194 193 194 193 194 194 193 191 191 191 189 189 188 188. 188 187 186 186 185 185 IBS 184 184 182 182 181 181 180 180 179 178 178 177 177 177 176 176 176 175 175 174 174 174 173 172 172 171 171 171 171 171 171 170 171 171 170 171 170 170 170 169 170 169 169 170 169 169 169 169 169 169 169 168 169 168 168 168 168 168 168 168 168 168 168 168 169 169 169 169 169 169 169 170 171 171 172 172 173 173 175 175 176 176 178 179 179 180 180 181 182 182 184 185 185 186 187 186 188 188 188 189 189 189 189 189 191 189 191 189 189 189 188 189 189 188 188 188 188 187 186 185 185 185 184 184 184 181 183 181 181 181 180 180 180 180 179 179 178 178 178 177 177 176 176 176 176 175 175 175 174 175 175 174 175 174 175 174 174 175 174 175 175 176 176 176 177 178 179 179 180 180 181 184 184 185 186 187 188 189 191 193 194 196 197 199 200 201 204 205 208 211 213 215 219 224 224 224 230 230 238 243 250 1 5 7 7 U 11 n 12 12 13 15 15 15 15 15 16 15 16 17 18 19 19 20 21 21 22 22 23 23 24 25 25 25 25 26 27 27 27 27 27 28 28 28 28 29 29 29 30 30 30 31 32 32 32 31 32 32 32 32 32 32 31 30 30 30 29 28 27 27 25 25 25 23 23 22 22 21 21 21 21 22 22 22 22 23 23 24 25 26 26 27 27 28 28 28 28 29 29 29 29 29 29 28 27 27 26 25 25 23 23 22 21 21 19 18 16 16 14 13 12 12 11 11 11 11 9 9 11 9 8 7 7 7 5 3 243 237 237 231 230 231 230 230 224 220 219 215 215 213 209 208 205 205 205 204 201 200 199 196 196 193 192 189 189 188 186 185 184 182 184 184 184 183 182 184 184 184 184 183 184 182 184 184 184 184 184 185 186 185 185 185 184 184 184 184 184 184 184 184 184 184 185 186 188 189 191 193 193 195 195 200 200 202 206 211 213 216 219 224 230 237 247 5 7 11 12 12 13 13 15 16 15 15 14 13 13 12 11 7 7 5 247 243 237 237 237 233 237 243 237 238 245 3 5 5 5 247 1 1 7 7 5 5 247 250 250 240 240 240 237 233 230 227 224 220 219 215 214 209 208 206 204 204 201 201 200 197 196 1 94 194 193 194 196 194 193 194 193 193 191 191 191 189 191 189 189 188 188 188 188 188 188 188 189 188 189 189 189 189 189 191 191 191 193 191 193 193 193 193 193 194 196 194 196 196 196 196 194 194 194 194 196 196 196 196 196 196 194 194 196 196 196 197 197 199 196 194 194 193 192 191 189 188 187 186 186 184 184 184 182 184 184 184 182 180 180 178 176 175 173 171 170 169 166 165 162 162 161 159 157 156 155 153 153 152 153 153 154 154 154 154 156 156 157 158 159 159 160 161 162 164 165 166 168 169 170 169 171 172 174 175 175 176 178 179 182 1S4 185 185 185 186 186 186 188 189 193 194 194 194 193 193 191 189 189 188 187 186 185 184 184 184 184 180 180 178 178 176 176 175 175 175 172 170 169 169 169 168 167 168 166 167 168 165 165 165 165 166 166 166 166 168 169 170 170 171 172 174 175 176 178 180 182 184 187 189 191 193 199 201 204 208 209 213 215 219 222 230 233 237 250 5 7 11 12 15 16 17 19 21 21 22 22 23 24 25 25 25 26 26 28 28 29 28 29 29 29 29 28 29 28 28 28 27 26 26 25 25 24 22 20 18 15 12 n 5 237 225 219 211 208 200 196 191 187 184 179 176 171 169 166 162 159 156 153 150 147 145 143 140 139 138 136 133 132 129 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 12B 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 12 B 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 132 128 128 128 134 142 142 133 128 128 128 126 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 128 133 131 128 128 128 128 128 128 128 128 128 128 128 128 128 128 134 139 138 132 128 128 128 128 128 128 128 128 128 128 128 130 211 A.1 Wind Velocity Determination (uncorrected) from ‘Raw’ Wind Blast Data The 0-10 mA velocity signal from the sensor pod is proportional to the dynamic (velocity) pressure. This signal is converted by the logger to a number proportional to the square root of the dynamic pressure. The square root extraction is performed using a ‘lookup table’ whose values were determined from data collected during wind tunnel tests of the sensor pods at the University of Queensland (Fowler, 1997). In technical terms, the velocity signal is converted to a 12 bit binary number by an analog to digital converter in the logger. This 12 bit number is then used to represent an address or location in the lockup table which contains the correct 8 bit linearised number. The latter is then stored. This 8 bit signed number is proportional to the wind velocity. The following relationship yields the wind velocity in metres per second: V= K x N / 12.6 (A.1) where, V= wind velocity in metres per second; N= The absolute value of the stored data (-ve sign represents reverse air flow); and K is given by the relationship K = -J2*g/p (A .2) where, g = gravitational constant (assumed 9.81 m/s/s); p = the density of air as given by p = (p* 100)/(T*288) (A.3) where, p= baraometric pressure in millibars; and T= temperature in degree Kelvin. Note: As air temperature was not monitored, no temperature correction was applied for the online velocity determination (uncorrected) in the Wind Blast Data Logger. Standard air density was assumed at standard temperature and pressure. 212 A.2 Absolute Pressure Measurement from ‘Raw’ Wind Blast Data The output from the static pressure transducer of the sensor pod has been calibrated to 11.84 mA at 1 atmosphere (1000 hPa) The following relationship determines the static pressure from the actual value stored: p = 10.14 *(N- 51) (A.4) where, p = pressure in millibars (mB); and N= stored lookup value. Density correction (Dc) is given by the following expression: Dc = A/l/(Af-51)*0.01 (A-5) where, N= stored lookup value. A.3 Wind Velocity Corrections A.3.1 Zero correction equivalents (Baseline shift) The zero correction for the velocity time histories is according to the following relationships show in Table A. 1. Table A.l Zero correction equivalent calculation Stored lookup value Velocity correction Differential Pressure correction N V Pd (m/s) (Pa) 0.318 * (255 -N) 0.0618 * (255-N)2 255 0.0 0.0 254 0.3 0.1 253 0.6 0.2 247 2.5 4.0 237 5.7 20.0 213 A.3.2 Velocity correction for the compressibility effect and density The dynamic (velocity) pressure is the difference between the stagnation (total) pressure (p0) and static pressure (p). At high pressures and corresponding high densities, corrections needs to be made for the compressibility effect (Discussed in Chapter 4, section 4.2.1). For y = 1.4 equation 4.8 (Chapter 4) will be (A.5) where, M is the Mach value. The above equation is sometimes called the pitot tube equation because of its application in correcting pitot tube readings (Liepmann & Puckett, 1950). The wind blast loggers indicated velocity (Vi) responds to the difference between the static and stagnation pressures according to 1 T/2 Po~P = ~Po Vi (A. 6) where, pQ is standard air density The true air velocity (corrected) will be obtained using equations A.5 and A.6 in the form Pa V,2 = p V2.f(M) (A.7) where, M2 f(M) = \ + — + .... 4 The corrected wind velocity (V) is then (A 8) 214 APPENDIX B EXAMPLES OF PROCESSED OVERPRESSURE AND VELOCITY TIME HISTORIES, NEWSTAN & MOONEE COLLIERIES 215 300596#1.oppod1 Overpressure lime history Elapsed ime (seconds) Figure B. 1 Raw overpressure time history for wind blast event at LW 8, Newstan Colliery 300596# 1 oppoc Overpressure tir Peak Smoothed Overpressure = 9 kPa 3 4 5 Elapsed Time (seconds) Figure B.2 Smoothed overpressure time history for wind blast event at LW 8, Newstan Colliery 216 * 10 - 300596#1.oppod1 Overpressure time history ime (seconds) Figure B.3 Integrated overpressure time history for wind blast event at LW8, Newstan Colliery ® 0.980 - 300596#1.oppod1 Overpressure time history 3 4 5 Elapsed Time (seconds) Figure B.4 Air density correction factor time history for wind blast event at LW8, Newstan Colliery 217 30059601 oppodl Overpressure time history 3 4 5 Elapsed Time (seconds) Figure B.5 Differentiated overpressure time history for wind blast event at LW8, Newstan Colliery Peak Smoothed 300596# 1pod2 Wind Blast Wind velocity time history Velocity = 24 m/s § 3 4 5 Elapsed Time (seconds) Figure B.6 Smoothed wind velocity time history for wind blast event at LW8, Newstan Colliery 218 181000#1.oppod2 Overpressure time history u> 6 — in Peak Smoothed Overpressure = 13 kPa 3 4 5 Elapsed Time (seconds) Figure B.7 Overpressure time history, near face, for wind blast event at LW4b, MooneeColliery Peak Smoothed Wind Blast Wind velocity time history Velocity = 59 m/s 3 4 5 Elapsed Time (seconds) Figure B.8 Wind velocity time history, outbye in maingate (A hdg), for wind blast event at LW4b, Moonee Colliery 219 050400#1 oppod3 Overpressure time history Peak Smoothed Overpressure = 9 kPa Elapsed Time (seconds) Figure B.9 Overpressure time history, in tailgate, near face, for wind blast event at LW4a, Moonee Colliery Peak Smoothed Wind Blast 050400# 1.pod3 Velocity = 38 m/s Wind velocity time history 3 4 : Elapsed Time (seconds) Figure B. 10 Wind velocity time history, in tailgate, near face, for wind blast event at LW4a, Moonee Colliery 220 121099# 1.oppod2 Overpressure time history Peak Smoothed Overpressure = 3 4 5 Elapsed Time (seconds) Figure B. 1 lOverpressure time history, in maingate, near face, for wind blast event at LW3, Moonee Colliery Peak Smoothed Wind Blast Velocity - 66 m/s 3 4 5 Elapsed Time (seconds) Figure B. 12 Wind velocity time history, in maingate, near face, for wind blast event at LW3, Moonee Colliery 221 121099# 1.oppod3 Overpressure time history Peak Smoothed Overpressure = 3 4 5 Elapsed Time (seconds) Figure B. 13Overpressure time history, in tailgate, near face, for wind blast event at LW3, Moonee Colliery Peak Smoothed 121099#! pod3 Wind Blast Wind velocity time history Velocity - 68 m/s 3 4 : Elapsed Time (seconds) Figure B. 14 Wind velocity time history, in tailgate, near face, for wind blast event at LW3, Moonee Colliery 222 APPENDIX C WIND BLAST VELOCITY AND OVERPRESSURE PARAMETERS, MOONEE COLLIERY 223 MOONEE COLLIERY - LONGWALL 1 • INCIDENCES OF ROOF FALLS AND WIND BLAST VELOCI' (saj|auj) Peak Velocity Measurements Maximum Excursion Maximum rate of rise Com m ents (saj)aiu zero corrected .p ro cessed (maximum Integrated of w ind velocity (sm ooth E=0/ Binomial 50 w ind velocity) zero corrected,smoothed jaqiunu and density corrected zero & density corrected (smooth/E»0/B 501) a6eu|ego jaqiunu a & Integrated jenbs) density corrected and differentiated ||ej (m etres per second) (m etres) (m etres p er seco n d ) juaAg jeoo eajy auiu »tea aaej MG MG B hdg TG MG B hdg MG MG B hdg TG MG B hdg MG MG B hdg TG MO ■ hdg A hdg • on DCB (outbye) A hdg ■ on DCB (outbye) A hdg - on OCB (out bye) POD1 POD2 POD3 POD4 POD1 POD2 POD3 POD4 POD1 POD2 POD3 POD4 - 3 03 CO rvi 8 1 £ I 1.719 I 18,510 CN s o | Prior to Wind Blast ffl 1 1.410 l CO 3 03 CO S $ Logger Installation S 1 4,800 | - 3 03 CO LD 1.670 ! £ tn CO R £ to 1 22-Jan-98 1 § co S co U- AAA 'S *> T) ” CO O 130298 # 1 -3 10:24:28 1,611 370 3.6 r^- 03 U_ 03 1.9 V 1 6 3 Nose #1 blocked 19:34 Appx 1,540 4,810 co Li 03 03 Power off V lo kj- r- 190298 #1-7 23:19:15 1,540 700 2 8 (N kj- U. AAA 03 co 1,4 2 8 6 Nose#1 blocked V in CO 9 240298# 1-6 2:26:02 1,531 1.160 4 7 CN kj- U- 03 co 162 V 17 4 26.9 Nose #1 blocked 10 240298 # 7-C 3:37:05 1,531 1,100 5 7 7 .4 CN uS U. 03 CO V 14.6 30 Nose#1 blocked 11 09:57 Appx 1,507 2,050 2 Loaaer fuH s s I ■3 223 26-Feb-98 05:00 Appx I 2 8 CN § N CO CN 1 I 28-Feb-98 280298 # 1 -2 12:09:01 o I || o in CO CO CM co 11-Mar-98 110398 #1 1,401 CD Double event I 1 M o u 11039ft t n ■> si I 1:30 Appx 1.401 I 8 B (O CN i | 3 120398 #1 06:52:46 1 0 m ? 8 3 S 130398 #1 4:59:16 I 8 CD I § kT s 2 2 B B 140398 #1 1:01:25 1,384 | CD o S 6 03 CO V S? CN 5 CO 8 o r- co kf o in 2 CD CN kt in CN 5:57:54 a ; a ( 2 over vel range S 5 38 N n o s s ■ (