c12020 final report - coalminesinquiry.qld.gov.au...tb tb tb tb 17.9 2.9 1.6 76.2 3.2 16.7 24.5 49.5...

ACARP Aus t r a l i a n Coa l A s soc i a t i o n R esea r ch P r og r am

FINAL REPORT

Proactive Inertisation Strategies and

Technology Development

C12020 December 2005

BOI.036.001.0001

DISCLAIMER No person, corporation or other organisation (“person”) should rely on the contents of this report and each should obtain independent advice from a qualified person with respect to the information contained in this report. Australian Coal Research Limited, its directors, servants and agents (collectively “ACR”) is not responsible for the consequences of any action taken by any person in reliance upon the information set out in this report, for the accuracy or veracity of any information contained in this report or for any error or omission in this report. ACR expressly disclaims any and all liability and responsibility to any person in respect of anything done or omitted to be done in respect of the information set out in this report, any inaccuracy in this report or the consequences of any action by any person in reliance, whether wholly or partly, upon the whole or any part of the contents of this report.

BOI.036.001.0002

-

EXPLORATION AND MINING

EXPLORATION AND MINING REPORT P2006/26

ACARP Project C12020 Proactive Inertisation Strategies and Technology Development

Rao Balusu, Ting X Ren and Patrick Humphries

December 2005

OPEN REPORT

CSIRO Exploration & Mining PO Box 883, Kenmore, Queensland, Australia 4069

BOI.036.001.0003

PROJECT PARTICIPANTS The following people and organisations have participated in the ACARP project C12020 ‘Proactive inertisation strategies and technology development’ directly or indirectly through participation in discussions and various aspects of the project studies. ACARP

Paul O’Grady Neil Winkelmann Bevan Kathage Roger Wischusen Anne Mabardi

CSIRO

Rao Balusu Ting Ren Patrick Humphries Sheng Xue

MINES and other organisations

Bruce Robertson Tim Harvey Barry Robinson Tim Hobson Paul O’Grady Neil Winkelmann Dennis Black David Stone David Sykes Kelvin Schiefelbein Michael Loney David McMillan Russell Thomas Russell Packham Greg Smith Michael Barker

David Cliff Ron McKenna Mark Parcell Darren Brady Kevin Carey Bob Korczynski

ACARP Project C12020 Final Report, Proactive Inertisation, December 2005, Australia.

BOI.036.001.0004

ACARP Project: C12020

Proactive Inertisation Strategies and Technology Development Project Leader: Rao Balusu, CSIRO Exploration and Mining Industry Monitors: Paul O’Grady and Neil Winkelmann. ACARP Funding: $194,000 Total: $439,000 Completion: Dec 2005

ABSTRACT The frequency of heating incidents in longwall panels has increased significantly in recent years, leading to production losses and safety risks for a number of coal mines in Australia and around the world. Review of oxygen ingress patterns into longwall goaf at various mines has shown that in some cases the oxygen concentration in the goaf was well over 17% even at 300 m to 400 m behind the longwall face. Under those circumstances, the risk of the heating/spontaneous combustion incidents increases substantially during slow panel retreat or prolonged face stoppage periods. To reduce the risk of these goaf heatings, a major research project has been undertaken under the ACARP project 12020 “Proactive inertisation strategies and technology development”. The main objective of the project is to develop and demonstrate the proactive and effective inertisation strategies and techniques to reduce the risk of spontaneous combustion and heatings in active longwall panel goafs. The project has combined extensive field studies, together with computational fluid dynamics (CFD) models of goaf gas flow, to develop proactive inertisation strategies. The project has developed a detailed understanding of the effects of various inertisation strategies on goaf gas distribution in longwall panels. Field studies involving low flow inertisation of the working longwall panel goafs have also been conducted at two underground coal mines in Australia. The project results have shown that the proactive inertisation strategies implemented at the field sites were highly successful in reducing oxygen ingress into the goaf and the consequent risk of heatings in the longwall panels during the field demonstration periods. Results of various modelling investigations and field studies are presented in this report.

- - - - -


BOI.036.001.0005

TABLE OF CONTENTS

Page

SUMMARY............................................................................................................................................... i ACKNOWLEDGEMENTS ....................................................................................................................xi

1. INTRODUCTION 1.1 Background .............................................................................................................1 1.2 Objectives................................................................................................................2 1.3 Scope of work .........................................................................................................2 1.4 Project studies .........................................................................................................3

2. OXYGEN INGRESS AND HEATINGS ISSUES IN LW GOAFS 2.1 Introduction.............................................................................................................5 2.2 Brief review of heatings incidents in Australia ......................................................5 2.3 Oxygen ingress patterns in LW goafs – Mine A ....................................................7 2.4 Oxygen ingress patterns in LW goafs – Mine B ..................................................10 2.5 Heatings issues in longwall goafs ........................................................................12 2.6 Summary and conclusions.....................................................................................13

3. CFD MODELLING OF INERTISATION IN WORKING LW GOAFS 3.1 Introduction...........................................................................................................15 3.2 CFD models development ....................................................................................15 3.3 Base models results ...............................................................................................18 3.4 Goaf inertisation simulations – parametric studies ...............................................27 3.5 Development of proactive inertisation strategies .................................................38 3.6 Summary and conclusions ....................................................................................49

4. STUDIES ON OTHER OPTIONS TO REDUCE O2 INGRESS 4.1 Introduction...........................................................................................................51 4.2 Background of field sites – Mine C & Mine D .....................................................51 4.3 Foam injection studies...........................................................................................53 4.4 Foam and inert gas injection studies .....................................................................60 4.5 CFD modelling of foam plugs – preliminary studies ...........................................69 4.6 Summary and conclusions.....................................................................................70

5. FIELD DEMONSTRATION STUDIES 5.1 Introduction...........................................................................................................71 5.2 Field demonstration studies and results – Mine A ...............................................71 5.3 Field demonstration studies and results – Mine B ...............................................80 5.4 Proactive inertisation strategies ............................................................................85 5.5 Summary and conclusions ....................................................................................86

6. CONCLUSIONS 6.1 Conclusions and Recommendations .....................................................................87 6.2 Future research .....................................................................................................91

REFERENCES ......................................................................................................................93


BOI.036.001.0006

i

SUMMARY A number of new underground coal mines being developed in the Hunter Valley and Bowen Basin in Australia have a moderate to high susceptibility to spontaneous combustion (sponcom). The coal seams in these new areas are generally thick and the risk of sponcom/heatings increases significantly during longwall extraction due to the large quantities of broken coal left behind the chocks and its exposure to high oxygen levels in the goaf. This poses a major risk to the safety of the people, longwall face equipment and economic viability of the modern and highly capital intensive coal mines. There was an urgent need to develop and demonstrate pro-active inertisation technology to reduce the risk of heatings in working longwall panel goafs, particularly during the slow retreat or prolonged face stoppage scenarios. To address the above issue a major research project has been undertaken by CSIRO, in collaboration with Australian coal mining companies under the ACARP project C12020 entitled “Proactive Inertisation Strategies and Technology Development”. The main objective of the project was to develop and demonstrate effective proactive inertisation strategies to reduce the risk of heatings in longwall panels, particularly under modern longwall mining conditions in Australia that are characterised by high face ventilation quantities in the range of 50 to 80 m3/s and thick seam conditions. The project has adopted an integrated approach involving detailed monitoring and analysis of goaf gas distribution in working longwall panels for improved understanding of oxygen ingress patterns in the goafs; extensive Computational Fluid Dynamics (CFD) modelling of various inertisation options under different mining conditions and development of effective proactive inertisation strategies; and long term field demonstration studies at two underground longwall panels to evaluate the effectiveness of proactive inertisation on goaf oxygen ingress patterns. The details and results of the research work are presented in this report. (1) OXYGEN INGRESS PATTERNS AND HEATING ISSUES IN LONGWALL GOAFS In order to investigate the effect of proactive inertisation in longwall panels, it is very important to characterise the initial oxygen ingress patterns in longwall goafs under different mining conditions. This involves a detailed monitoring of the gas distribution behind the face under different ventilation and operational conditions. Details and results of the field studies carried out to characterise oxygen ingress patterns at two different mines are presented in the main report. A typical oxygen ingress pattern into the longwall goaf is presented in this section. The longwall panel at Mine A has access around the perimeter of the panel which enabled extensive monitoring of gas distribution in the goaf on both maingate (MG) and tailgate (TG) sides. Seven tube-bundle monitoring points were installed in the goaf to continuously monitor goaf gas distribution at varying distances behind the longwall face. A snapshot of the typical goaf gas distribution behind the longwall face at Mine A is presented in Figure 1. Results show that intake air ingress on the maingate side of the panel was very high, with the oxygen level at more than 17% even at 400 m behind the longwall face. Oxygen ingress on the tailgate appears to extend only up to 100 m behind the face due to higher goaf gas emissions in the panel.


BOI.036.001.0007

ii

17.2 2.3 4.4 73.1

TG

MG

Tb

Tb

Tb

Tb

17.9 2.9 1.6 76.2

3.2 16.7 24.5 49.5

6.3 8.8 21.9 54.8 Dip

Gas (%) O2

CO2 CH4 N2

Longwall start-up

20.1 0.8 0.2

2.2 22.5 29.1 41.2

14.4 8.2 6.4 67.8

Oxygen ingress into the goaf

Goaf Face

77.2

Figure 1 Typical gas distribution in the longwall goaf at Mine A

This level of high oxygen ingress could lead to heating development in the goaf, particularly during face stoppage or slow face retreat in the panel. Under these circumstances it was proposed to introduce proactive inertisation to reduce the length of the oxygen ingress zone in the goaf and prevent risk of heatings. It was also highlighted by Cliff (2005) that given the difficulty in detecting an active heating in longwall goafs, we should focus on preventing a heating from occurring. Cliff pointed out that early response should be triggered by detection of high oxygen in the areas of the goaf where it should not be. Although high oxygen does not immediately cause trouble, it will increase the likelihood of sponcom in the goaf if the longwall face stops or slows down for any reason. Remedial actions such as proactive inertisation to reduce the oxygen supply into the goaf can avert a heating in such cases. It is to be noted here that in fact, a small heating developed in the goaf within few weeks of the goaf gas condition shown in Figure 1. The goaf gas distribution at the other mine also showed similar trends with high oxygen ingress up to 400 m behind the longwall face. Although oxygen ingress into the goaf was high on the intake maingate side in both the cases, it was also observed that in some other mines oxygen ingress was high on the tailgate side of the panel. Detailed analysis of the various cases revealed that oxygen ingress depends on a number of factors including intake airflow, ventilation layout, pressure differentials, seal condition and leakages, caving characteristics, high permeability areas adjacent to the gateroads, goaf gas emissions, panel geometry, seam gradients and dip direction. Therefore, it is important to characterise the oxygen patterns in any new mine goafs before developing appropriate proactive inertisation strategies. The above goaf gas monitoring data was used to calibrate and validate the CFD models for developing appropriate proactive inertisation strategies. A brief summary of the modelling investigations and the results of field demonstration studies are presented in the following sections. (2) CFD MODELLING OF VARIOUS PROACTIVE INERTISATION STRATEGIES The focus of the CFD modelling exercise was to obtain a better understanding of the inert gas flow patterns in the active longwall goafs, and to perform a detailed analysis of various mining and operational factors for development of effective proactive inertisation strategies. A major advantage of CFD simulation is its capability to predict what will happen under a given set of circumstances, i.e. it can answer many ‘what if?’ questions before a proposed design can be implemented in the field.


BOI.036.001.0008

iii

(a) Base model simulations

A number of base case models have been developed to represent various longwall panel geometries under different mining and gas emission conditions based on the preliminary field investigations carried out at a number of mine sites. A commercial CFD code, FLUENT, has been used for development of these 3D longwall goaf gas flow models. The base models for the longwall inertisation studies were typically 1 km in length along the panel, 250 m in width and 80 m in height to cover the immediate high porosity caving regions in the goaf. Goaf gas emission was varied between 100 l/s and 3,000 l/s to represent different goaf gas environments. The results of the typical base case simulations for a longwall panel with face air quantity around 50m3/s are presented in Figure 2, showing the oxygen gas distribution in the goaf. In the colour coding scale of the figures, 0.21 represents 21% oxygen, i.e. fresh air composition. In these base cases, intake to the panel was through the maingate (MG), which was at 20 m lower elevation compared to the tailgate return airway. The goaf gas emission rate in this base case was around 600 l/s. Results show that oxygen ingress into the goaf was high with oxygen levels on the maingate side well over 14% at 350 m behind the face and over 10% even at 600 m behind the longwall face.

TG

MG

O2 c

once

ntra

tion

(0.2

1 =

21%

)

Figure 2 Oxygen distribution in the longwall goaf – typical base model results

Analysis of the results showed that intake airflow and ventilation pressures seem to have a major influence on gas distribution up to 50 to 150 m behind the face, and beyond that, goaf gases buoyancy seems to play a major role on goaf gas distribution. Base case simulation results presented in Figure 4 tallied well with the results of field goaf gas monitoring studies. (b) Inertisation simulations

The validated base case models were then used for extensive parametric studies involving changes in inert gas injection locations, inert gas flow rates and different inertisation strategies to investigate their effect on goaf inertisation in active longwall panels. The results of some typical inertisation simulations are presented here. (i) Effect of inert gas flow rate near the face: The effects of inert gas injection near the face at different flow rates were investigated in the first set of parametric studies. The results of the inertisation studies with inert gas injection at 30 m behind the face with flow rates of 0.5 m3/s and 1.0m3/s are presented in Figure 3, along with a comparison to base model results. Comparison of (a) and (b) in Figure 3 shows that inert gas injection at 30 m behind the face at a flow rate of 0.5 m3/s had negligible effect on oxygen ingress patterns


BOI.036.001.0009

iv

into the goaf, with oxygen levels well over 13% at 300 m behind the face. Figure 3(c) shows that inert gas injection even at 1.0 m3/s had only a minor effect on goaf gas distribution, with oxygen concentration around 10% at 300 m behind the face.

(c) inert gas injection @ 1.0 m3/s

(b) inert gas injection @ 0.5 m3/s

(a) no inert gas injection – base model results

TG

MG

Oxy

gen

conc

entra

tion

(0.2

1 =

21%

)

Figure 3 Effect of inert gas injection near the face on oxygen distribution in the goaf

Further simulations indicated that when inert gas was injected within 10 to 30 m behind the face, inert gas flow rates well over 5 m3/s were required to substantially reduce oxygen ingress into the longwall goaf. These results indicated that inert gas injection just behind the face was not an effective strategy and further simulations needed to be carried out to establish optimum locations in the longwall panel in order to achieve goaf inertisation with low inert gas flow rates. (ii) Effect of inert gas injection location behind the face: The effect of inert gas injection at a number of different locations behind the longwall face was investigated in this set of simulations. A comparison of the results of investigations with inert gas injection at 30 m and 200 m behind the face on maingate side is shown in Figure 4. In both the cases, inert gas was injected at the low flow rate of 0.5 m3/s. Oxygen distribution in the goaf for base model simulations is also shown in the Figure 4a for comparison purposes. Figure 4c shows that inert gas injection at 200 m behind the face resulted in substantial reduction in


BOI.036.001.0010

v

oxygen ingress into the goaf, with oxygen levels reducing below 5% within 100 m behind the longwall face. Results of the simulations indicate that inert gas injection deep in the goaf results in effective goaf inertisation, even at low inert gas flow rates of 0.5 m3/s.

MG

TG

(b) inert gas injection at 30 m behind the face


Oxy

gen

conc

entra

tion

(0.2

1 =

21%

)

(c) inert gas injection at 200 m behind the face Figure 4 Effect of inert gas injection location on oxygen distribution in the goaf

Further simulations with other inert gas injection location scenarios indicated that inert gas injection even at other deeper locations in the goaf, between 200 to 400 m behind the face, results in effective goaf inertisation for the modelled longwall panel conditions. Although the exact optimum location for inert gas injection for any longwall panel depends on site specific parameters, these modelling simulations indicated that inert gas injection at around 200 to 400 m behind the face would be far more effective than inert gas injection close to the face. Analysis of the various simulation results also indicated that longwall panel geometry, goaf gas emission rates and composition, ventilation layouts, pressures and flow rates, goaf characteristics, and gateroad conditions in the goaf would also have a significant influence on goaf inertisation. Although inert gas injection in the deep goaf at very low flow rates of 0.15 m3/s would also significantly reduce oxygen ingress into the goaf in some cases, modelling investigations indicated that an inert gas flow rate of around 0.5 m3/s would be required for goaf inertisation in most cases.


BOI.036.001.0011

vi

(iii) Effect of inert gas injection on MG/TG side: Additional simulations have been carried out to investigate the effect of inert gas injection on maingate vs tailgate sides of the longwall panel at the same distance behind the face. The results of the base case simulations along with the results of above simulations are shown in Figure 5. Oxygen distribution in Figure 5b shows that inert gas injection at 200 m behind the face on the maingate side substantially reduced oxygen ingress into the goaf and achieved successful goaf inertisation. This result also indicates that inert gas injection at 200 m behind the face would be an effective proactive inertisation strategy even in start-up areas of the longwall panels. However, results presented in Figure 5c indicate that inert gas injection on the tailgate side of the same panel at 200 m behind the face has no significant effect on oxygen distribution in the goaf. Most of the inert gas injected into the goaf on the tailgate side seems to migrate directly towards longwall return without significantly effecting oxygen distribution in the goaf.

TG

MG

(b) inert gas injection on maingate side


Oxy

gen

conc

entra

tion

(0.2

1 =

21%

)

(c) inert gas injection on tailgate side

Figure 5 Effect of inert gas injection on maingate vs tailgate side on O2 ingress into the goaf

These simulations indicate that in addition to the inert gas injection at optimum location behind the face, it is also important to inject inert gas on the appropriate side of the longwall panel to achieve effective goaf inertisation. These results highlight the importance of site parameters and the need to take into consideration all the parameters while developing effective inertisation strategies.


BOI.036.001.0012

vii

(3) FIELD DEMONSTRATION STUDIES Two operating mines with high oxygen ingress into the goaf were selected for field demonstration studies, as they present the worst case scenario and are therefore ideal sites for field investigation of proactive inertisation strategies. These two mine sites, referred to as Mine A and Mine B, are located at two different coal regions in Australia, one in New South Wales and other in Queensland. Proactive inertisation strategies for active longwall panels have been developed based on the analysis of the modelling results and review of the oxygen ingress patterns in the field at the mine sites. These inertisation strategies have then been implemented at the mine sites to investigate their performance in the field with respect to their effect on oxygen ingress patterns in the goaf. (a) Field studies at Mine A

The longwall face at mine A had progressed for about 450 m from the panel start-up line before the start of field trials of proactive inertisation. The normal goaf gas distribution in the panel showed that oxygen ingress into the goaf was very high with oxygen levels over 17% at 400 m behind the face. The proactive inertisation strategy for this panel consisted of injecting inert gas into the goaf through the first cut-through near the start-up area at the rate of about 0.15 m3/s. Inert gas was delivered to the cut-through via a 150 mm pipe connected to the bottom of a surface concrete drop hole in the perimeter roadway.

17.2 2.3 4.4 73.1

TG

MG

Tb

Tb

Tb

Tb

17.9 2.9 1.6 76.2

3.2 16.7 24.5 49.5

6.3 8.8 21.9 54.8

Gas (%) O2

CO2 CH4 N2

Longwall start-up

20.1 0.8 0.2

2.2 22.5 29.1 41.2

14.4 8.2 6.467.8

60 m3/s


Goaf

77.2

Dip (a) before inert gas injection

2.0 3.5 6.5 87.9

TG

MG

Tb

Tb

3.5 0.4 0.5 95.6

3.0 21.7 25.6 49.5

5.3 9.7 14.1 70.8

Gas (%) O2

CO2 CH4 N2

Concrete drop hole

5.2 2.9 6.1 85.6

4.0 25.0 25.3 45.6

5.0 15.6 18.5 60.8

0.6 6.7 17.4 75.3

20.6 0.4 0.1

Inert gas injection

Goaf

Oxygen ingress into the goaf (with proactive inertisation)

78.8

(b) after inert gas injection Figure 6 Effect of proactive inertisation on goaf gas distribution at Mine A


BOI.036.001.0013

viii

The gas distribution in the longwall goaf before and after injection of inert gas is shown in Figure 6. Results show that this proactive inertisation strategy was highly successful in the field during study period and had substantially reduced oxygen ingress into the longwall goaf on both sides of the panel. Field results in Figure 6b show that the oxygen level was below 6% at the second cut-through behind the face, i.e. 200 m behind the face. As can be seen from comparison of (a) and (b) in Figure 6, inert gas injection had also resulted in reduced oxygen ingress on the tailgate side of the panel. High oxygen ingress was restricted to only one cut-through behind the face on the maingate side. (b) Field studies at Mine B

The longwall panel at mine B also had progressed for about 550 m from the panel start-up line before the start of proactive inertisation field studies. The goaf gas distribution in the panel showed that oxygen ingress into the goaf was very high with oxygen levels over 17% at 370 m behind the face. The proactive inertisation strategy for this panel consisted of injecting inert gas into the goaf through two surface boreholes located on the maingate and tailgate sides of the panel at the combined flow rate of 0.6 m3/s. The gas distribution in the longwall goaf before and after injection of inert gas is shown in Figure 7, along with the location of inert gas injection boreholes.

TG

MG16.5 2.3 0.4 80.7

20.2 1.1 0.1 78.6

Dip Gas (%) O2

CH4 CO2 N2

Longwall start-up

20.7 0.7 0.1 78.5

19.9 1.1 0.1 78.7

17.8 1.6 0.4 80.1


Previous panel goaf

30 m3/s

Goaf

0.9 16.9 7.2 75.0

2.9 15.6 6.8 74.7

0.8 19.9 8.5 70.6

TG

MG 2.6 15.7 5.2 76.4

7.3 15.3 3.3 74.1

Gas (%) O2

CH4 CO2 N2

2.4 15.0 5.1 77.4

2.0 17.2 5.4 75.3

Inert gas injected through boreholes

Previous panel goaf

Goaf

(a) before inert gas injection

Oxygen ingress into the goaf(with proactive inertisation)

(b) after inert gas injection Figure 7 Effect of proactive inertisation on goaf gas distribution at Mine B


BOI.036.001.0014

ix

Results of the field studies show that this proactive inertisation strategy was highly successful in the second field site also and had substantially reduced oxygen ingress into the longwall goaf. Field results presented in Figure 7b show that the oxygen level was below 3% at the second cut-through behind the face, i.e. 100 m behind the face. This very high reduction in oxygen ingress was due to a combination of inert gas injection and reduction in intake airflow. It is to be noted here that longwall face was not retreating during the field studies period due to geotechnical problems and ventilation in the panel was reduced to 30 m3/s to assist in reducing oxygen ingress into the goaf. Field study results at both the mine sites demonstrated that the proactive inertisation strategies were highly successful in converting the general goaf environment into an inert atmosphere. These field studies also demonstrated that it is possible to reduce the oxygen ingress distance in the goaf to within 200 to 300 m behind the face by implementing appropriate proactive inertisation strategies. (4) CONCLUSIONS The longwall goaf gas distribution measurements at the mine sites showed that air ingress into the goaf was very high on the maingate side of the panels with oxygen concentration levels over 17% at 300 m to 400 m behind the face. This level of high oxygen ingress could lead to heating development in the goaf, particularly during face stoppage or slow face retreat periods. There was a need to develop and demonstrate proactive inertisation strategies to reduce the risk of heatings in active longwall panels. CFD modelling techniques have been used to investigate the effect of various inertisation strategies on goaf oxygen ingress patterns. The results of the base-case CFD models tallied well with the results of field monitoring studies. Inertisation simulations indicated that inert gas injection close to the face would be ineffective even at higher flow rates in the order of 1.0 to 2.0 m3/s. Modelling results indicated that inert gas needs to be injected at 200 m to 400 m behind the face at the rate of about 0.5m3/s to achieve effective goaf inertisation in most cases. It is recommended that proactive inertisation may be introduced in the following scenarios to reduce the risk of heatings development in the longwall goafs:

• Longwall panels extraction in the areas/seams with very high spontaneous combustion potential.

• Extraction in areas with severe geological disturbances, where it is expected that longwall retreat will be very slow for a period of weeks/months.

• Longwall stoppage for prolonged periods – e.g. due to tailgate/face collapse or other reasons.

• Abnormal CO readings behind the face or steady rise in CO readings at a number of seals in the active longwall goaf.

• When working longwall face retreating past suspected heating areas in the adjacent sealed goafs (proactive inertisation in adjacent sealed goaf).

• Abnormal CO readings or other indications of heatings in adjacent sealed goafs (proactive inertisation in adjacent sealed goaf).

The recommended guidelines for proactive inertisation strategy are:

i. inert gas should be injected into the goaf at 200 to 400 m behind the face, or inbye side of a suspected heating location in the goaf.

ii. inert gas flow rate of around 0.5 m3/s is recommended for most cases – this inertisation rate may need to be increased or decreased based on field conditions.


BOI.036.001.0015

x

iii. inert gas should be injected on intake side of the goaf in most cases. iv. Inert gas may be injected through goaf holes (preferable, if goaf holes are

available) or through cut-throughs on the intake side of the goaf. v. inert gas may need to be injected on both sides of the goaf if some heating is

suspected on return side of the goaf. vi. onsite nitrogen generation units or boiler gas are recommended for proactive

inertisation into longwall goafs. vii. inert gas injection to be continued until face resumes normal production in case of

prolonged stoppages or until face has retreated for more than 300 to 500 m past the suspected heating location, in case of heatings.

viii. ventilation system in and around the panel also should be designed to minimise oxygen ingress into the longwall goaf for effective inertisation.

Based on the above guidelines, appropriate inertisation strategies need to be developed for any specific field site based on the site conditions. Field studies carried out at two mine sites demonstrated that the proactive inertisation strategies developed during the course of the project were highly successful in reducing oxygen ingress into the goaf and in achieving effective goaf inertisation during the study periods. Field results showed that oxygen levels were below 5% to 6% at 200 m behind the face after implementation of the proactive inertisation. The project studies also greatly improved the fundamental understanding of the various parameters and strategies on goaf inertisation process. The proactive inertisation strategies developed during the course of this project have played a crucial role in containing the onset of spontaneous combustion and in reducing the risk of larger heatings development in both the longwall goafs during field demonstration periods. It is to be noted here that just injecting inert gas into the goaf does not ensure prevention of heating incidents in the entire length of panels or all longwall panels. A number of parameters such as major changes in panel ventilation system, goaf caving conditions, longwall retreat rate and goaf gas drainage can make a specific inertisation strategy ineffective and the inertisation strategies need to be modified based on the changed conditions at the field sites. Technology transfer to the industry through implementation of the newly developed techniques at the mine sites has been one of the highlights of the project. This project demonstrated that it is possible to reduce the oxygen ingress distance in the goaf to within 200 to 300 m behind the face by implementing appropriate proactive inertisation and ventilation strategies. The fundamental understanding of inert gas flow patterns in active goafs and proactive inertisation strategies developed during the course of the project can greatly reduce the risk of heatings in longwall panels and enhance the safety of coal mines.

- - - - -


BOI.036.001.0016

xi

ACKNOWLEDGMENTS The authors are highly grateful to the Australian Coal Association and ACARP management for funding this research project. Special thanks are due to industry monitors Dr Paul O’Grady and Mr Neil Winkelmann for their invaluable suggestions, support and management of the project. We also wish to thank Mr Roger Wischusen and Dr Bevan Kathage of ACARP for their contribution and support to the project. We wish to express our sincere appreciation to Mr Bruce Robertson and Mr Tim Harvey of Anglo Coal Australia for their valuable contributions, support and interest in the proactive inertisation research. The authors wish to record their grateful thanks to the management and staff of several collieries of Anglo Coal Australia and Xstrata for their support, permission to carry out the field studies and making the relevant data available for the project. We wish to express our special appreciation to Mr Barry Robinson, Mr Tim Hobson, Mr Dennis Black, Mr David Sykes, Mr Kelvin Schiefelbein, Mr David Stone, Mr Michael Loney, Mr David McMillan, Mr Russell Thomas, Mr Russell Packham, Mr Kevin Carey, Mr Mark Parcell, Mr Greg Smith and Mr Michael Barker for their valuable contribution and time taken to help the project team. We would also like to express our special appreciation and thanks to Dr David Cliff, Mr Ron McKenna, Mr Darren Brady, and Mr Bob Korczynski for their valuable input and support to the project. The authors would like to thank CSIRO Exploration & Mining for the permission and facilities provided in the conduct of the study. We would also like to thank our colleagues at CSIRO who contributed to different aspects of this work. In particular the contributions of Dr Sheng Xue, Mr Mick Kelly and Dr Hua Guo are very much appreciated.

- - - - -


BOI.036.001.0017

1

CHAPTER 1

INTRODUCTION

1.1 BACKGROUND The number of heating incidents in longwall panels has increased significantly in recent years leading to major production losses and safety risk for a number of mines in Australia. Most of the new underground coal mines in the Hunter Valley and Bowen Basin have a moderate to high susceptibility to spontaneous combustion (sponcom). The coal seams in these new areas are generally thick and the risk of sponcom/heatings increases significantly during longwall mining due to the large quantities of broken coal left behind the chocks and its exposure to high oxygen levels in the goaf. This poses a major risk to the safety of the people and economic viability of the modern highly capital intensive coal mines. The cost of dealing with such goaf heatings after the incident is very high in addition to the time lost and potential explosion risks involved. The risk of these heatings increases substantially during slow panel retreat or during prolonged face stoppage periods due to other problems. The main difficulty in preventing or dealing with the heatings after the incident is the lack of detailed understanding of various control options and inertisation strategies. In Australia, most of the work so far in the inertisation research area has concentrated on development and demonstration of inert gas generators. This solved one of the major problems of supplying inert gas to the mines located in remote parts of Australia. However, a critical review of the major heating control operations shows that the success or failure of the inertisation operations depends entirely on the design of inertisation strategies. Although the exact strategy to prevent heatings or to deal with heatings after the incidents depends on the location and panel conditions at the time of heatings, there is a need to improve our knowledge on the effects of various inertisation control strategies. It is better to try to minimise the occurrence of heating incidents in the first place by implementing proactive inertisation strategies. Literature survey shows that research in this area is going on in UK, France, Czech Republic and some other European countries over the last few years, and it has been reported that some mines have successfully used inert gas injection just behind the face to minimise the risk of sponcom in the longwall goafs. However, in the case of Australian longwall mines it was found that inert gas injection just behind the face does not have a significant influence on goaf oxygen ingress patterns. The mining conditions and ventilation systems in the Australian longwalls are different and therefore seem to require a combination of new and effective proactive inertisation strategies. In summary, there is an urgent need to develop and demonstrate proactive inertisation technology to reduce the risk of sponcom/heatings in the active longwall goafs, particularly to prevent development of heatings during the slow retreat or prolonged face stoppage scenarios. This research project has been undertaken to address the above major industry issues.


BOI.036.001.0018

2

1.2 OBJECTIVES The main objective of the project is to develop and demonstrate the pro-active inertisation strategies/ technology to reduce the risk of heatings in the active longwall goafs. The project will also develop an understanding of the effects of various inertisation and oxygen ingress control strategies, which will enable the mine operators to develop the most cost-effective strategies to deal with the spontaneous combustion risk or heating incidents. 1.3 SCOPE OF WORK To achieve the above objectives, the project research investigations have been carried out in parallel on two major fronts – CFD modelling and field trials. The project work was carried out in close collaboration with mine managements and involved trials and implementations of the various pro-active inertisation strategies developed during the course of the project. The effectiveness of the various techniques and options was measured by monitoring of the gas distribution and inertisation at critical points around the goaf. The scope of the project included:

• Characterisation of oxygen ingress patterns in active longwall goafs

• A brief review of heatings development in longwall goafs

• Development of CFD models of longwall panels to simulate oxygen gas distribution in the goafs for various mining geometries

• Extensive parametric studies to investigate the effects of various inertisation strategies under different airflow rates, goaf geometries and mining parameters

• Determination of the optimal locations and flow rates for proactive inertisation – to reduce oxygen ingress into the goaf under different mining conditions

• Preliminary investigation of other techniques such as foam injection to reduce oxygen ingress into the goaf

• Field demonstration studies – to evaluate and quantify the effects of various proactive inertisation strategies in the field

• Development of generic guidelines for proactive inertisation.


BOI.036.001.0019

3

1.4 OVERVIEW OF PROJECT STUDIES The total duration of the project was 24 months, commencing in July 2003 and completed by September 2005. The project has combined extensive field studies at mine sites together with CFD modeling investigations in order to characterise the effects of various mining parameters and develop effective proactive inertisation strategies. The following studies have been conducted during the course of this research project.

1. Characterisation of oxygen ingress and Review of heatings development in LW goafs

Goaf oxygen ingress characterization studies involve a detailed monitoring of the gas distribution behind the longwall face under two different mining conditions. Details and results of these field gas monitoring studies along with a brief review of heatings development in longwall goafs is presented Chapter 2. Results of these initial field monitoring studies have been used for setting up the base case CFD models and for initial calibration and validation of the numerical models.

2. Development of CFD models and Inertisation simulations

Computational Fluid Dynamics (CFD) models of longwall goafs have been developed with a number of different mine/panel geometries. The purpose of these base case studies was to simulate oxygen ingress patterns in the longwall goafs under changing mining conditions. These different base case CFD models will provide an understanding of the effects of goaf geometry on goaf gas distribution and inert gas flow patterns. The validated base case models were then used for extensive parametric studies involving changes in inert gas injection locations, inert gas flow rates, inert gas composition and inertisation strategies under different mining conditions with variations in goaf geometries, ventilation flow rates and goaf gas emissions. The main focus of these modeling studies was to obtain a qualitative understanding of the effects of various parameters and strategies on goaf inertisation and to assist in the development of effective proactive inertisation strategies. The details and results of these modelling studies are presented Chapter 3.

3. Preliminary studies on other options to reduce oxygen ingress into the goaf

The concept of additional measures such as foam injection, foam and inert gas combination on oxygen ingress patterns into the goaf was also trialed during the course of this project. The details and results of these preliminary field investigations on foam injection are presented in Chapter 4, along with basic simulation results of this concept.

4. Field demonstration studies

Results obtained from the above initial field studies and CFD modelling investigations were analysed in detail to obtain an understanding of the effects of various parameters and strategies on goaf inertisation. Optimum proactive inertisation strategies have been developed based on the above analysis and interpretations.


BOI.036.001.0020

4

These proactive inertisation strategies have then been implemented at two mine sites to investigate their performance in the field, particularly with respect to their effect on oxygen ingress patterns in the goaf. The field studies were carried out in the longwall panels over long periods up to a year to investigate the effect of proactive inertisation at various stages of longwall extraction in the panel. Details and results of these field investigations with proactive inertisation strategies are presented in Chapter 5, together with detailed discussions.

5. Conclusions and Recommendations

Chapter 6 describes the main findings and conclusions of this inertisation research project. Recommended proactive inertisation strategies and suggestions for further research are also listed in this chapter.

- - - - -


BOI.036.001.0021

5

CHAPTER 2

OXYGEN INGRESS AND HEATING ISSUES IN LW GOAFS

2.1 INTRODUCTION Hazards resulting from the spontaneous combustion/heating of coal remain a major threat worldwide to the safety and productivity of underground coal mines. The most problematic situation may occur when spontaneous heatings develop in the presence of an inherently explosive atmosphere which may lead a local heating incident into an explosion hazard involving the entire mine workings. Inquiry into the explosion incident at Moura No 2 underground mine in Australia concluded that the explosion was initiated by spontaneous combustion. The seriousness of such incidents and associated financial loss was again highlighted by the recent closure of Southland Colliery due to the breakout of a spontaneous heating fire (Gallagher, 2004). In addition to the intrinsic propensity of coals, the occurrence of spontaneous heating within a longwall goaf is closely associated with the airflow characteristics behind the face. Air leakage occurs across the permeable goaf due to the existence of pressure differentials. Current knowledge is still limited on the behaviour of inert gas in the longwall goafs. Inert gas is often injected with little knowledge of where it is needed, or where it is going in the goaf. A detailed understanding of the flow patterns and distribution of gas flow in the goaf is necessary not only to improve the control of goaf gas emissions but also for spontaneous heating prevention strategies such as the injection of inert gas. In order to investigate the effect of proactive inertisation in longwall panels, it is very important to characterise the initial oxygen ingress patterns in longwall goafs under different mining conditions. This involves a detailed monitoring of the gas distribution behind the face under different ventilation and operational conditions. Details and results of the field studies carried out to characterise oxygen ingress patterns at two different mines are presented in this Chapter. A brief review of heating incidents in coal mines and discussions on heatings issues in longwall goafs are also presented in this Chapter. 2.2 BRIEF REVIEW OF HEATINGS INCIDENTS IN AUSTRALIA A comprehensive review of spontaneous heating incidents in Australia by Cliff et al (1996) in a SIMTARS publication entitled ‘Spontaneous Combustion in Australian Underground Coal Mines’ showed that:

During the period 1960 to 1991, there were 125 incidents of spontaneous combustion reported by the Inspectorate in New South Wales

•

• During the period 1972 to 1994, there have been 68 recorded incidents of spontaneous combustion in underground coal mines in Queensland.


BOI.036.001.0022

6

Some of these disastrous incidents had led to loss of life and high social and financial costs in terms of mine closures, notably:

Box Flat (1972): An explosion following a fire killed 17 men in the Box Flat mines near Ipswich Queensland. The fire, which was started by spontaneous combustion in fallen coal in a cross cut between two intake airways, proved to be uncontrollable despite several attempts to put it out. The official inquiry concluded that explosive gasses built up in the recirculated air, eventually causing an explosion, which set off a larger coal dust explosion.

•

•

•

Moura (1994): The Moura No. 2 room and pillar mine in Queensland experienced an explosion that killed eleven people. One week before the disaster, management had suspected spontaneous combustion in an almost completed extraction panel and had sealed the section over the weekend. The area was monitored and Graham's ratio levels showed no sign of spontaneous combustion risk. Shortly after production resumed there was a rapid rise in carbon monoxide (CO) levels indicating combustion. This should have triggered a mine evacuation except the tube sample system had an inherent forty-minute delay. By the time the higher gas levels showed up on the surface monitoring equipment the mine had already exploded.

Southland Colliery (2003): A very serious underground fire occurred at the Southland Colliery during late December 2003. An emergency situation developed very quickly at Southland and culminated in the sealing and subsequent closure of the mine.

Since the Moura No.2 disaster in 1994, there has been no loss of human life resulting from these incidents in Australia. This is largely due to the increased awareness of the spontaneous heating hazards, the significant improvements in gas monitoring systems, the availability of a range of inertisation facilities, and the improved fundamental understanding of sponcom fires and gas flow mechanics in the longwall goafs. However, spontaneous combustion continues to occur in different sections of the coal mines, particularly in longwall goafs, and the costs incurred to the coal mining industry due to production delays or in the worst case scenario of mine closure runs into millions of dollars per year. Traditionally there has been a cherished myth that once a sign of a heating is detected there is time to deal with it before the situation gets out of control. The reality is often not the case, as it has been well demonstrated in the recent incident at Southland and those at Dartbrook and North Goonyella (Cliff, 2005). Large scale testing at SIMTARS using 16 tonnes of coal has clearly shown that a heating can gestate for months without being easily detected then within a 24 hour period rapidly escalate to reach the point of open flame (Cliff et al, 2000). ‘Prevention is better than cure’. The cost of dealing with such goaf heatings/fires after the incident is very high in addition to the time lost and potential explosion risks involved. Proactive inertisation is one of the potential prevention strategies if successfully developed and implemented can significantly reduce the risk of heatings development in longwall goafs. In order to develop effective proactive inertisation strategies, it is essential to first characterise and understand the oxygen ingress patterns in longwall goafs under different conditions. The following sections present the results of those initial field investigations.


BOI.036.001.0023

7

2.3 OXYGEN INGRESS PATTERNS IN LONGWALL GOAFS – MINE ‘A’ Mine A is located in the Hunter Valley in New South Wales. The mine has a production capacity of 4.0 million tonnes per year and operates a single longwall face and two continuous miner sections. Coal seam gas typically consists of carbon dioxide (CO2) and methane (CH4) gases with a CO2:CH4 ratio ranging between 65:35 and 35:65. The longwall face ventilation airflow was approximately 90 m3/s, with around 140 m3/s total flow in the panel, i.e. remaining 50 m3/s going through the perimeter roadway. The goaf gas emissions in the panel were high in the range 1,800 to 2,000 l/s, with around 1,200 l/s CO2 and 800 l/s CH4. The longwall layout is primarily a 2 entry gateroad system, as shown in Figure 2.1. The longwall panel has access around the perimeter of the panel which enabled extensive monitoring of gas distribution in the goaf on both maingate (MG) and tailgate (TG) sides. Seven tube-bundle monitoring points were installed in the goaf to continuously monitor goaf gas distribution at varying distances behind the longwall face.

TG MG

LW1

Figure 2.1 The longwall panel layout at mine A Figures 2.2 to 2.5 presents some snapshots of the typical goaf gas distribution patterns behind the longwall face as it retreated from the start-up line. The gas monitoring results indicate that the intake air ingress on the maingate side of the panel was very high, with the oxygen level at more than 17% even at 400 m behind the longwall face. However, oxygen ingress on the tailgate appears to extend only up to 100 m behind the face, due to the higher goaf gas emission in the panel at this mine.


BOI.036.001.0024

8

Tube bundle point

17.4 2.5 7.5 66.4

20.5 0.3 0.1 76.6

TG

MG

19.3 1.3 3.1 72.6

19.6 1.71.9

73.5

12.3 6.7

9.8 66.9

Gas (%) O2

CO2 CH4 N2 Longwall

start-up

11.3 12.6 9.9

60.8

Face


Goaf

Dip

Figure 2.22 Gas distribution in the longwall goaf at Mine A – case 1

Tube bundle point

17.4 2.1 8.8 66.4

20.9 0.3 0.1 78.6

18.8 1.4

TG

MG

18.4 3.24.9

70.5

15.6 4.2

6.8 71.2

Gas (%) O2

CO2 CH4 N2

Dip

Longwall start-up

8.9 13.7 13.4 59.2

Face


Goaf

5.4 70.8



BOI.036.001.0025

9

TG

MG

Tb 14

16.8 2.5 4.8 72.5

6.8 11.415.959.2

4.2 10.5

27.6 47.6

Dip Gas (%)

O2 CO2 CH4 N2 Longwall

start-up

19.3 1.7 0.7 77.2

6.9 15.4 19.1 53.3


Face

Tube bundle point

Goaf


17.2 2.3 4.4 73.1

TG

MG

17.9 2.9 1.6 76.2

3.2 16.7 24.5 49.5

6.3 8.8 21.9 54.8

Gas (%) O2

CO2 CH4 N2

Dip

Longwall start-up

20.1 0.8 0.2 77.2

2.2 22.5 29.1 41.2

14.4 8.2 6.4 67.8

Tube bundle point

Face


Goaf


This level of high oxygen ingress into the goaf could lead to the development of spontaneous heating behind the longwall face, particularly during face stoppage or slow face retreat in the panel. It is to be noted here that in fact, a small heating developed in the goaf within few weeks of the goaf gas condition shown in Figure 2.5. Signs of spontaneous combustion have been picked up by the tube bundle gas monitoring system when the face retreated about 450 m and proactive inertisation was required to suppress the onset of the low-level heating.


BOI.036.001.0026

10

2.4 OXYGEN INGRESS PATTERNS IN LONGWALL GOAFS – MINE ‘B’ Mine B is located in the Bowen Basin, west of Mackay in Central Queensland. The thickness of the working seam at the mine ranges between 4 and 6.5m and the depth of workings was around 125 to 300m. The mine has a production capacity of 6.0 million tonnes per year and operates a single longwall face and two continuous miner sections. Coal seam gas is predominantly methane (CH4), with insitu gas content in the range of 3 m3/t to 6 m3/t in the panel area. The goaf gas emissions in the panel were low in the range 300 to 400 l/s. The longwall face ventilation airflow was around 60 m3/s to 70 m3/s. As shown in Figure 2.6, the longwall panel has access only on the maingate side of the panel, as is the case in most of the longwall panels. Five tube-bundle monitoring points were installed on the intake maingate side of the goaf to continuously monitor oxygen ingress into the goaf at varying distances behind the longwall face.

Figure 2.6 Longwall panel layout at Mine B Figures 2.7 to 2.8 present two snapshots of the typical goaf gas distribution patterns behind the longwall face as it retreated from the start-up line. The gas monitoring data show that intake air ingress on the maingate side of the panel was very high with the oxygen level at more than 17% even at the 5th cut-through behind the longwall face, i.e. 370 m behind the longwall face. The oxygen ingress pattern on the tailgate side of the panel was based on the gas readings obtained from an experimental gas drainage borehole.


BOI.036.001.0027

11

Gas (%) O2

CH4 CO2 N2

Dip

Longwall start-up

Tube bundle point 20.7 0.3 0.1 78.8

20.6 0.3 0.1 78.9


MG

TG

Face

Goaf

Previous panel goaf

Figure 2.7 Gas distribution in the longwall goaf at Mine B – case 1

16.5 2.3 0.4 80.7

20.2 1.1 0.1 78.6

Gas (%) O2

CH4 CO2 N2

Longwall start-up

20.7 0.7 0.1 78.5

19.9 1.1 0.1 78.7

17.8 1.6 0.4 80.1


MG

Tube bundle point

Longwall Goaf

TG

Face

Previous panel goaf

Dip

Figure 2.8 Gas distribution in the longwall goaf at Mine B – case 2 As the longwall panel retreated for about 550m from the panel start-up line, a roof control problem in the tailgate slowed down its designed production rate and ultimately the face had to be stopped for the reinstallation of a new tailgate. The face ventilation was re-arranged and an airflow of 10m3/s delivered to the tailgate side by an auxiliary fan via ducting pipes. Figure 2.9 shows a snapshot of the goaf gas distribution behind the longwall face during this period in LW105. It can be seen from the goaf gas monitoring data that airflow ingress into the goaf was still quite deeper, and the oxygen concentration level was over 10% of the goaf gas, even at some 350 m behind the face line.


BOI.036.001.0028

12

18.8 2.8 0.2 78.2

17.4 2.4 0.4 79.8

11.54.4 0.8

83.2

10.43.7 1.0

84.8

8.8 4.6

1.5 85.0

Longwall Goaf

TG Closed

Dip

Gas (%) O2

CH4 CO2 N2

Longwall start-up

MG

Face

Previous panel goaf


Figure 2.9 Gas distribution in the longwall goaf at Mine B – case 3

2.5 HEATINGS ISSUES IN LONGWALL GOAFS The goaf areas in longwall mines are very complex in terms of ventilation flow patterns or caving processes. The goaf gas flow patterns depends on a number of factors including intake airflow, ventilation system, seals condition and leakage characteristics, pressure differential around the goaf, goaf gas emissions, panel geometry and seam gradients. Although oxygen ingress into the goaf was high on the intake maingate side in both of the above cases, it was also observed that in some other mines oxygen ingress was high on the tailgate side of the panel. This deep penetration of airflow provides a favourable condition for spontaneous combustion/heatings to develop in the goaf areas. Laboratory studies on bulk coal samples at SIMTARS (Cliff et al, 2000) and column testing at University of Queensland (Beamish and Jabouri, 2005) showed that self-heating warm spot starts near the return side of the sample and progresses slowly towards the intake side. The hot spot migrates towards the air intake source as the coal on the leading edge of the hot spot dries out and the hot spot chases the intake air to sustain the oxidation reaction. These laboratory studies indicate that the risk of heatings would be high in high oxygen environments, i.e. near high oxygen ingress areas on intake side of the goaf. Analysis of a number of heating incidents in longwall panels also indicated that a majority of heating incidents develop on the high oxygen ingress side of the goaf, except in special circumstances. The risk of heatings in longwall goafs increases if the face slows down or stops for prolonged periods due to any reasons. It is also to be noted here that prolonged stoppage of the longwall face at both the mines led to increased oxidation and low level heatings in the goafs. The cost of dealing with such goaf heatings/fires after the incident is very high in addition to the time lost and potential explosion risks involved.


BOI.036.001.0029

13

In a number of cases, it may be difficult to detect sponcom heatings at early stages of development, depending on the ventilation system and monitoring locations. It is also very difficult to locate the exact position of heating area in most cases. Another major difficulty in dealing with the heatings after the incident is the lack of detailed understanding of various control options and inertisation strategies. Although the exact strategy to deal with heating incidents in mines depends on the panel conditions at that time and heating locations, there is also a need to improve current knowledge on the effects of various inertisation strategies. Under these circumstances it was proposed to introduce proactive inertisation to reduce the length of the oxygen ingress zone in the goaf and prevent risk of heatings. It was also highlighted by Cliff (2005) that given the difficulty in detecting an active heating in longwall goafs, we should focus on preventing a heating from occurring. Cliff pointed out that early response should be triggered by detection of high oxygen in the areas of the goaf where it should not be. Although high oxygen does not immediately cause trouble, it will increase the likelihood of sponcom in the goaf if the longwall face stops or slows down considerably for any reason. Remedial actions such as proactive inertisation to reduce the oxygen supply into the goaf can avert a heating in such cases. While the earlier inertisation projects in Australia concentrated on development of inert gas generators (Bell et al, 1998), this project focused on development of effective inertisation strategies to reduce the risk of heatings in longwall goafs. 2.6 SUMMARY AND CONCLUSIONS In spite of the advancements in mining technologies and underground mine gas monitoring systems, the spontaneous combustion of coal in longwall goafs continues to be hazardous threat to production and mining safety due to the coal’s inherent liability to sponcom, the large quantities of broken coal left behind the chocks and its exposure to high oxygen levels in the goaf. It is paramount that such incidents can be detected and controlled at its earliest stage to avoid production delay and safety risks to the mine. The cost for dealing with such goaf heatings/fires following the incident can be very high, in addition to the time lost and potential explosion risks involved. Extensive longwall goaf gas monitoring has been carried out in longwall mines and the results have been analysed in detail. The longwall goaf gas distribution measurements at these two mine sites showed that air ingress into the goaf was very high on the maingate side of the panels with oxygen concentration levels over 17% at 300 m to 400 m behind the face. This deep penetration of airflow provides an ideal catalyst for spontaneous combustion to develop in the goaf. In particular, the high level of oxygen ingress could lead to spontaneous heatings if the face stops for a prolonged period due to any problems. In fact, the combination of the high oxygen ingress into the goafs along with prolonged panel stoppage and geological disturbances has contributed to the increased oxidation and low level heating in the goafs in both coal mines.


BOI.036.001.0030

14

Detailed analysis of the various cases revealed that oxygen ingress depends on a number of factors including intake airflow, ventilation layout, pressure differentials, seal condition and leakages, caving characteristics, high permeability areas adjacent to the gateroads, goaf gas emissions, panel geometry, seam gradients and dip direction. Therefore, it is important to characterise the oxygen patterns in any new mine goafs before developing appropriate proactive inertisation strategies. Analysis of the above goaf gas distribution studies coupled with heating incidents confirms the need for detailed investigation of various proactive inertisation strategies to reduce oxygen ingress in order to prevent the risk of heatings in longwall goafs. The following chapters will describe the modelling investigations carried out to develop optimum proactive inertisation strategies, and the results of field demonstration studies during the course of this project.

- - - - -


BOI.036.001.0031

15

CHAPTER 3

CFD MODELLING OF INERTISATION IN LONGWALL GOAFS

3.1 INTRODUCTION The occurrences of spontaneous combustion in longwall goafs has led to mine abandonment or production suspension in a number of underground coal mines worldwide. The use of inert gas such as nitrogen and boiler gas has become a preventive practice during longwall sealing, and more recently during normal working operations in some mines. However, current knowledge is still limited on the behaviour of inert gas in the working panel goafs, and the inert gas was often injected with little knowledge of where it was needed, or where it is going in the goaf. A detailed understanding of the flow patterns and distribution of gas flow in the goaf is necessary not only to improve the control of goaf gas emissions but also for spontaneous heating prevention strategies such as the injection of inert gas. Proactive inertisation of open goafs in active longwalls can be used to suppress the development of potential goaf heatings and ‘save’ time for the longwalls to advance beyond dangerous zones and to maintain normal production rate. This method is particularly important for reducing the risk of spontaneous heatings in active longwall goafs during slow face movement or stopping, in particular due to:

Slow face movement - due to difficult geological problems, i.e. faults/roof falls. Roadway collapse and associated secondary support. High oxygen ingress into the goaf, and Other production problems.

Supported by ACARP and in collaboration with Australian underground coal mines, CSIRO has developed an integrated approach incorporating CFD techniques to simulate a variety of pro-active inertisation strategies with the objective to reduce the risk of spontaneous heatings in active longwall faces. This chapter describes the CFD model development, base model simulation results and parametric studies for the formulation of proactive inertisation strategies for field demonstrations in two Australian underground longwall panels.

3.2 CFD MODELS DEVELOPMENT CFD is commonly accepted as referring to the broad topic embracing mathematics and numerical solution, by computational methods, of the governing equations which describe the motion of fluid flow, the set of the Navier-Stokes equations, continuity and any additional conservation equations, such as energy or species concentrations. Today CFD has grown from a mathematical curiosity to become an essential tool in almost every branch of fluid


BOI.036.001.0032

16

dynamics, from aerospace propulsion to weather prediction. The availability of robust commercial CFD codes and high speed computing has lead to the increasing use of CFD for the solution of fluid engineering problems across all industrial sectors and the mining industry is no exception. More recently, CFD simulation techniques have been used by researchers at CSIRO to investigate goaf gas migration mechanisms in longwall goafs for the development of effective goaf gas control strategies (Balusu et al, 2001), the optimum inertisation practice for longwall panel seal-off operations (Balusu et al, 2002), and airflow and dust dispersion patterns around longwall shearer (Balusu et al, 2004). The ACARP project C9006 ‘Optimisation of Inertisation Practice’ has greatly improved the fundamental understanding of the various site parameters and inertisation schemes for panel seal-off operations. Optimum inertisation strategies have been developed for specific site conditions and proved to be highly successful in goaf inertisation. A commercial CFD code Fluent has been selected for this study. Fluent is a finite volume computational fluid dynamics code that solves the Navier-Stokes equations for both incompressible and compressible flows. An elementary calculation of transfers to and from the neighbouring volumes is performed for each surface of the mesh. These exchanges depend on the incoming and outgoing flows and on the intrinsic characteristics of the flow regions. A key feature of this code is its user-defined function capability, or UDF, which allows the user to develop stand-alone C programs that can be dynamically linked with the Fluent solver to enhance the standard features of the code. Gas flow migration, including the injection of inert gases, is a complicated process in the longwall goaf, as many factors are involved, including ventilation layout and intensity, gas emission rate and compositions (e.g. the presence of methane and carbon dioxide), face (seam) orientation and dip, gas buoyancy and goaf permeability. A range of CFD models have been developed to achieve a detailed understanding of the gas flow mechanics and distribution in longwall goafs. In addition to innovative CFD modelling, the study also involved extensive validation and calibration of initial models using data obtained from field studies and parametric studies to investigate the effect of various parameters on goaf flow patterns. The CFD modeling work generally involves a number of key stages, including:

Initial field studies to obtain the basic information on panel goaf geometries and other parameters

Construction of 3D finite element model of the longwall goaf Setting up flow models and boundary conditions Base case model simulations Model calibration and validation using field goaf gas monitoring data, and Extensive parametric studies and development of optimum inertisation strategies.

A key part of the CFD models is the incorporation of longwall goaf permeability distributions and gas emissions via a set of UDFs that are linked to the solver. Flow through goaf was handled using custom written subroutines, which were added to the “flow through porous media” modules of the basic code. In these subroutines/modules, flow through the porous goaf regions was simulated by adding a momentum sink to the momentum equations. The sink had viscous part proportional to the viscosity and an inertial component proportional to


BOI.036.001.0033

17

the kinetic energy of the gases. A number of subroutines were written to represent different ventilation and goaf gas emissions scenarios, which were then combined with the main CFD solver to carry out the simulations. Typically the CFD models are in 3D with 500,000 to 1,000,000 cells in order to capture the behaviour of goaf gas flow in a 250m longwall panel up to 1,000 m in the direction of face retreat. Longwall CFD models are constructed according to the actual mine layouts. The mesh used in the models was ‘refined’ with higher density mesh in the areas of interest such as areas next to the face and roadways. A typical geometry and mesh used in longwall goaf gas flow models are shown in Figure 3.1.

TGFACE

Longwall start-up

GOAF

MG

GOAF

TG

MG

FACE

GOAF

TG

MG FACE

TG

MG FACE

GOAF

Figure 3.1 Typical model geometry and mesh used in the longwall CFD gas flow models


BOI.036.001.0034

18

Goaf Face Retreating Direction

TGH6 TGH5 TGH4 TGH3 TGH2 TGH1


Figure 3.2 A selection of typical longwall panel layouts in Australia Figure 3.2 shows a selection of typical longwall ventilation layouts in Australian mines. CFD models can be developed and boundary conditions defined on the basis of any ventilation layouts that are used in the mines. In this study, extensive CFD modelling simulations have been conducted specifically to

Understand goaf gas distribution patterns before inert gas injection – base models Provide insight into the inert gas injection and migration process within the goaf Carry out parametric studies – e.g., pro-active inertisation at different locations, i.e.,

MG/TG/surface goaf holes Identify the best strategies for field trials Develop and optimise inertisation strategies

The following sections provide details of the base model simulations. All the simulations were carried out in steady-state and the effect of temperature (strata and inert gas) was not considered. Unless specified, all the CFD models have a dimension of 1000m in face retreat direction and 80 m into the roof. The width of the longwall face is typically 240m and varies according to the real field data. 3.3 BASE MODEL RESULTS The occurrence of spontaneous heating within a longwall goaf is closely associated with the airflow and goaf gas characteristics behind the face. Air leakage occurs across the permeable goaf due to the existence of ventilation pressure differentials. To fully understand the inertisation process and hence develop the most effective strategies, it is essential firstly to establish the initial oxygen ingress/goaf distribution patterns before introducing inert gas


BOI.036.001.0035

19

injection at various locations. For the purpose of developing optimum proactive inertisation strategies, a set of generic base models have been developed based upon specific longwall panels to simulate goaf gas patterns. The objective here is to establish the oxygen distribution patterns (inversely the goaf gas patterns) for a range of representative longwall systems with different goaf gas compositions and emission rates such that some generic but effective goaf inertisation options can be further studied for the development of optimum optimisation strategies. All the base models presented below have been calibrated using available field gas monitoring data or field observations and good agreements obtained. 3.3.1 Base model 1 Figure 3.3 shows the model layout for Base Model 1. This model represents a longwall mining panel in which:

The maingate (MG) side was at 20m lower elevation than tailgate (TG) side; • • •

•

The face was at 25m higher elevation than the start-up line; The face was ventilated with a total airflow of 100m3/s, with TG (main intake) and MG providing 70m3/s and 30m3/s respectively, and the 3rd cut-through (C/T) behind the face was used as the back-return for gas control, as shown in Figure 3.3; Total goaf gas emission – 3000l/s with 20% of methane and 80% of carbon dioxide.

MG Q = 30m3/s

TG Q = 70m3/s

Goaf

C/Ts

C/Ts

Dip

Retreat towards higher elevations

Start-up

Face

Figure 3.3 The geometry/panel layout for Base Model 1


BOI.036.001.0036

20

(a) Plan view of the oxygen ingress pattern into the goaf at the working level

MG

TG

Face Start up

(b) Cross-section view of the oxygen ingress pattern along the longwall goaf

O2 level 0.21 = 21%

MG Face

3 C/T Return

Face start-up

TG

(c) 3D view of the oxygen ingress pattern in the longwall goaf

Figure 3.4 Oxygen ingress pattern in the longwall goaf – Base model 1


BOI.036.001.0037

21

The oxygen ingress pattern in the longwall goaf for base model 1 is shown in Figure 3.4. In the colour coding sale of the figures, 0.21 (red) represents 21% oxygen, i.e. fresh air, and 0.00 (blue colour) represents the accumulation of goaf gas (with oxygen levels below 2%). It can be seen from the modelling results that oxygen penetration is deeper on the TG side, with up to 12% of oxygen even at some distance over of 300 m behind the face. However, due to the presence of high portion of heavy gas CO2, i.e. 80% of carbon dioxide, the goaf gas tends to accumulate on the lower side of the goaf close to the MG and in general, the goaf floor, as shown in the plan view and cross-section view along the middle of the longwall panel. In terms of the occurrence of spontaneous combustion behind the longwall, it is likely that the air ingression on the TG side and around the TG cut-throughs (C/Ts) would provide the ideal condition for spontaneous combustion to develop and thus the focus of pro-active inertisation should be directed on the TG side in this case. The high goaf gas emission rate is useful in helping the self-inertisation process of the longwall goaf. 3.3.2 Base model 2 Figure 3.5 shows the model layout for Base model 2. This model represents a longwall mining panel in which:

The MG side was at 20m lower elevation than the TG side; • • • •

Longwall face and start-up line was at the same elevation; U-ventilation pattern with a total airflow of 50m3/s; and Total goaf gas emission was around 600 l/s of methane (CH4).

This model layout represents the most common longwall geometry in a number of mines. Base models 3 and 4 represent variations of model 2 with MG at higher elevation and face retreating towards lower elevations in model 4.

Goaf Retreat Direction

C/Ts

C/Ts

Start-up line

Face

MG

TG

dip



BOI.036.001.0038

22

TG

MG


Start-up Face

(b) Cross-section view of the oxygen ingress pattern along the middle of the longwall goaf

O2 level 0.21 - 21%

Face

TG

MG




BOI.036.001.0039

23

Figure 3.6 shows the oxygen ingress pattern in the longwall goaf for base model 2. The airflow ingress into the goaf on the maingate side of the goaf is much deeper than on tailgate side of the panel. The modelling results indicate that oxygen ingress into the goaf on the maingate side was up to 12% of oxygen even at some distance over 500 m behind the face. Due to the dip of the seam and the effect of buoyancy, the goaf gas (methane) tends to migrate towards the high elevation side of the goaf close to the tailgate and in general, the high sections of the goaf, as shown in the plan view (a) and cross-section view along the middle of the longwall panel (b). Again, the layering of goaf gas can also be clearly observed in Figure 3.6c in which the airflow penetrates along the lower sections of the goaf, particularly immediately behind the face and further into the maingate side of the goaf. The deep air ingress on the maingate side would provide a much favourable condition for spontaneous combustion to develop and therefore pro-active inertisation and other precautions should be practiced more vigorously on this side of the longwall goaf. 3.3.3 Base model 3 Figure 3.7 shows the model layout for Base Model 3. This model represents a longwall mining panel in which:

The MG side was at 20m higher elevation than the TG side; • • • •

Longwall face and start-up line was at the same elevation; U-ventilation pattern with a total airflow of 45m3/s; and Total goaf gas emission was around 600 – 1,000 l/s of methane (CH4).

This model is a variation of base model 2 with elevations of MG and TG reversed, i.e. MG was at higher elevation in this model to investigate the effect of panel gradient on goaf gas distribution.

C/Ts

Goaf Retreat

C/Ts

Start-up line

Face

MG

TG

dip



BOI.036.001.0040

24

TG

MG


Start-up

Face

(b) Cross-section view of the oxygen ingress pattern along the middle of the longwall goaf

O2 level 0.21 = 21%

Face TG MG




BOI.036.001.0041

25

Figure 3.8 shows the oxygen ingress pattern in the longwall goaf for base model 3. Although a similar ventilation system was used in this study, the air ingress pattern into the goaf is significantly different from base model 2. The modelling results indicate that oxygen ingress into the tailgate side is much deeper than that in the maingate side, mainly due to the dip of the seam and the effect of buoyancy. The goaf gas (methane) tends to accumulate on the high elevation side of the goaf close to the MG and in general, the high parts of the goaf, as shown in the plan view (a) and cross-section view along the middle of the longwall panel (b). The layering of goaf gas can also be clearly seen in Figure 3.8 in which high level of oxygen settles at the lower sections of the goaf, particularly immediately behind the face and further into the lower side (TG) of the goaf. The deep air ingression on the TG side would provide the favourable condition for spontaneous combustion to develop and pro-active inertisation should be considered on this side of the longwall goaf under these modelled conditions. 3.3.4 Base model 4 The model layout was similar to the layout shown in Figure 3.7, except that longwall face has retreated towards lower elevations in this model. This model represents a longwall mining panel in which:

The MG side was at 10m higher elevation than the TG side; • • • •

Longwall face was at 30m lower elevation than face start-up line; U-ventilation pattern with a total airflow of 45m3/s; and Total goaf gas emission 600-1000l/s of methane (CH4).

Figure 3.9 shows the oxygen ingress pattern in the longwall goaf for Base model 4. The modelling results indicate that oxygen ingress into the goaf is almost at the same distance along both sides of the goaf, with up to 12% of oxygen at some distance over 300 m behind the face. Due to the dip of the seam and the effect of buoyancy, the oxygen penetration on the tailgate side tends to be slightly deeper, as shown in the plan view (a) and vertical cross section along the TG side of the longwall goaf (b). The buoyancy effect of the goaf gases can also be clearly observed in Figure 3.9 (c) in which high level of oxygen tends to settle at the lower sections of the goaf closer to the face, allowing deep penetration of ventilation airflow into the goaf at the working seam level. As far as spontaneous combustion is concerned behind the longwall, the air ingression on both sides of the goaf and around the cut-throughs (C/Ts) would provide the favourable condition for spontaneous combustion to occur and thus pro-active inertisation should be considered on both sides of the longwall goaf for these modelled conditions. The above understanding of the oxygen distribution patterns within the longwall goafs under different mining and geological conditions is important for the identification and development of optimum proactive goaf inertisation strategies by using inert gas, as demonstrated in the following sections.


BOI.036.001.0042

26

TG MG


Start-up Face

(b) Cross-section view of the oxygen ingress pattern along the TG side of the longwall goaf

TG MG

O2 level 0.21 = 21%




BOI.036.001.0043

27

3.4 GOAF INERTISATION SIMULATIONS – PARAMETRIC STUDIES CFD simulations were carried out with the above base models to investigate the effect of a range of factors on the inertisation of longwall goafs to suppress the development of spontaneous combustion. Extensive parametric studies were conducted during the course of this project, including:

The use of different inert gases, i.e., Nitrogen (N2), Carbon Dioxide (CO2) and Boiler gas (14% CO2, 85% N2 and 1% O2);

•

•

•

Inert gas injection rates, ranging from 0.15m3/s up to 3m3/s

Injection locations, i.e., both MG and TG sides at different distance behind the face through cut-throughs (C/Ts) as well as via surface boreholes

The following sections provide a summary of the parametric studies of goaf inertisation simulations. 3.4.1 Inert gases - N2, CO2 and Boiler gas CFD simulations were carried out with Base models 1 and 3 respectively to investigate the effectiveness of goaf inertisation with different inert gas at a rate of 0.5m3/s. In all cases, the inert gas was injected via the cut-through(C/T) at some 220m and 115 m behind the face on the maingate side. Figure 3.10 shows the oxygen distribution pattern in the goaf after the injection of boiler gas, nitrogen (N2) and carbon dioxide (CO2) with Base model 1. For the purpose of comparison, the predicted result for Base model 1 (without inert gas injection) was given in 3.10 (a). The modelling results indicate no difference on the effect of goaf inertisation between the injection of boiler gas and nitrogen in this case. There is however some marginal differences between the injection of carbon dioxide and nitrogen/boiler gas, as demonstrated in this case, the injection of CO2 seems to work better when the goaf gas consists of a high portion of carbon dioxide, i.e. heavy gas which has similar density. Figure 3.11 illustrates the oxygen distribution pattern in the goaf after the injection of boiler gas, nitrogen (N2) and carbon dioxide (CO2) with Base model 3. Again, for the purpose of comparison, the oxygen ingress pattern for Base model 3 (without inert gas injection) was also given in 3.11 (a). The modelling results indicate that there is no major difference on the effect of goaf inertisation between the injection of boiler gas and nitrogen in this modelled case. However, there are significant differences in goaf inertisation between the injection of carbon dioxide and nitrogen/boiler gas, as shown in Figure 3.11 (d). The injection of CO2 does not appear to have a good inertisation effect in the goaf in this case, instead, oxygen penetration on the tailgate side moves deeper into the goaf, which is likely to be problematic for spontaneous combustion management in the longwall panel.


BOI.036.001.0044

28

TG

MG

(a) Oxygen distribution in the goaf - the Base model 1

Injection point

MG

TG

(b) Oxygen distribution in the goaf – boiler gas injection at 220m MG side

(c) Oxygen distribution in the goaf – N2 injection at 220m MG side

O2 level %

Injection point

TG

MG

Injection point

MG

TG

(d) Oxygen distribution in the goaf – CO2 injection at 220m MG side Figure 3.10 Oxygen distribution pattern in the goaf after injection of different inert gases - Model 1


BOI.036.001.0045

29

MG

TG

(a) Oxygen distribution in the goaf - the Base model 3

TG

Injection point

MG

O2 level %

(b) Oxygen distribution in the goaf – boiler gas injection at 115m MG side

TG

MG

Injection point

(c) Oxygen distribution in the goaf – nitrogen injection at 115m MG side

TG

MG

Injection point

(d) Oxygen distribution in the goaf – carbon dioxide nitrogen injection at 115m MG side Figure 3.11 Oxygen distribution pattern in the goaf after injection of different inert gases - Model 3


BOI.036.001.0046

30

3.4.2 Inert Gas Injection Rates Obviously, the maximum volume of inert gas and hence the injection flow rate is limited by the current available inert gas generating systems and economic factors. There is a need to identify a range of inert gas injection rates that are sufficient to achieve the inertisation effect in the goaf under different conditions by considering both the economical constraints and other technical parameters.

(a) inert gas injection at 0.15 m3/s

MG

Injection point

TG

O2 level %

(b) Inert gas injection at 0.5 m3/s Injection point

MG

TG

(c) Inert gas injection at 2.0 m3/s Injection point

MG

TG

Figures 3.12 The effect of inert gas injection rates on goaf inertisation – base model 1

The predicted the goaf oxygen distribution patterns with different inert gas injection flow rates of 0.15m3/s, 0.5m3/s, and 2.0m3/s at 115 m behind the face on the MG side of the goaf for Base Model 1 conditions are shown in Figure 3.12. The modelling results indicate that there was no significant effect on goaf inertisation for injection rates at the lower range of 0.15m3/s to 0.5m3/s. The inertisation is taking effect only when the injection rates increased to a high flow rate of 2m3/s in the modelled case. It is to be noted here that the model in this case


BOI.036.001.0047

31

has high goaf gas emission rates with high proportion of CO2 gas and represents a case with highly favourable conditions for goaf inertisation. The results of another modelled case with low goaf gas emissions and high oxygen ingress are presented in Figure 3.13. The inert gas was injected at 30 m behind the face at flow rates of 0.5 m3/s to 2.0 m3/s. The modelling results show that inert gas injection even at higher flow rates of 2.0 m3/s did not result in effective goaf inertisation in this case. These results indicate that just increasing the inert gas flow rate would not achieve goaf inertisation in all cases and the other design parameters such as location of inert gas injection needs to be optimised to achieve effective goaf inertisation.

(a) Inert gas injection at 0.5 m3/s

Injection point

MG

TG

O2 level %

(b) Inert gas injection at 1.0 m3/s

TG

MG

Injection point

(c) Inert gas injection at 2.0 m3/s Injection point

TG

MG

Figures 3.13 The effect of inert gas injection rates on goaf inertisation – base model 2


BOI.036.001.0048

32

3.4.3 Inert Gas Injection Locations CFD simulations were carried out to identify the optimum inert gas injection locations behind the face to achieve the most effective goaf inertisation under a fixed injection flow rate of 0.5m3/s. Figures 3.14 and 3.15 shows the effect of inert gas injection at different locations behind the face on oxygen ingress patterns in the longwall goafs with Base model 1 and Base model 2 under high and low goaf gas emissions conditions respectively.

(a) Inert gas injection at 10m behind the face

MG

TG

Injection location

(b) Inert gas injection at 60m behind the face

(c) Inert gas injection at 115m behind the face

MG

TG

MG

Injection location

TG

O2 level 0.21 = 21% MG

TG

Injection location

Injection location

(d) Inert gas injection at 220m behind the face

Figures 3.14 Effect of inert gas injection locations on goaf inertisation - base model 1


BOI.036.001.0049

33

MG

TG

(a) Base model – no inert gas injection

(b) Inert gas injection at 60m behind the face

(c) Inert gas injection at 220m behind the face Injection location

MG

TG

MG

TG

Injection location

O2 level 0.21 = 21%

(d) Inert gas injection at 320m behind the face

TG

MG

Injection location

Figures 3.15 Effect of inert gas injection locations on goaf inertisation – base model 2

The modelling results indicate that inert gas injection at a location immediately behind the face line will only have negligible impact on goaf inertisation, as most of the inert gas injected will simply disperse into the main ventilation stream and disappear into the return airflow. Even at some 60m behind the face line, the injection of inert gas at a rate of 0.5m3/s would only have marginal effect on goaf inertisation. The results presented here (and other


BOI.036.001.0050

34

simulations carried out during the course of this project) indicate that the most optimum injection locations for the modelled cases should be within the range of 200-400 m behind the face line on the maingate side. The results indicated that inert gas flow rate of 0.5 m3/s would be required for most cases, although in some cases lower flow rates might be sufficient or in some other cases higher flow rates might be required to achieve the desired effect. The optimum location and appropriate inert gas injection site depends upon the understanding of goaf gas distribution patterns within the longwall goaf, as shown in Figure3.16 with Base model 3 following the injection of inert gas at 220m and 325 m behind the longwall face line at a flow rate of 0.5m3/s. It can be seen from the simulation results that by simply injecting inert gas on the maingate side at some 200~300m behind the face without a good knowledge of the oxygen ingress behaviour in the goaf may not be able to achieve the expected inertisation result. In this case, due to the gradient of the seam and hence the buoyancy effect of goaf gas, oxygen ingress into the tailgate side was much deeper, and the injected inert gas on maingate side was not reaching the high oxygen areas to deplete the high level of oxygen.

(a) No inert gas injection

(b) Inert gas injection at 220m behind the face on MG side

TG

MG

Injection location

MG

TG

MG

Injection location

TG

O2 level %

(c) Inert gas injection at 320m behind the face on MG side

Figures 3.16 Effect of inert gas injection on maingate side in base model 3


BOI.036.001.0051

35

(a) Inert gas injection at 115m behind the face on TG side

O2 level %

Injection point

MG

TG

TG

MG

Injection point

(b) Inert gas injection at 220m behind the face on TG side

Figures 3.17 Effect of inert gas injection on tailgate side in base model 3 The simulation results of injecting inert gas on the tailgate side in Base Model 3 are shown in Figure 3.17. The results indicate that a much better goaf inertisation effect can be achieved if the inert gas was injected on the tailgate side in this modelled case, even at some 115m behind the face line. The results indicate that a detailed understanding of oxygen ingress patterns into the goaf is necessary for design of effective proactive inertisation strategy. Figure 3.18 shows the simulation results of inert gas injection at different locations with Base Model 4. The modelling results indicate that by injecting inert gas on either MG side or TG side alone will not achieve a good goaf inertisation effect, due to the high oxygen ingress in both sides of the goaf in this case. However, with the same total inert gas injection volume (0.5m3/s), a better goaf inertisation result could be achieved by simultaneously injecting inert gas on both sides of the goaf, even at a lower rate of 0.25m3/s on each side. In brief, modelling results show that in most cases, the optimum inert injection locations would be within 200~400m behind the goaf on maingate (intake) side. It is equally important to have a prior knowledge of the goaf gas (air ingress) distribution patterns so that the optimum injection points can be correctly chosen to achieve the most effective goaf inertisation result. Inert gas injection on tailgate side or on both sides could also produce a desirable inertisation result in some cases, depending upon specific site conditions and the availability of inert gas generating systems.


BOI.036.001.0052

36

MG

TG

(a) no inert gas injection

(b) Inert gas injection at 220m behind the face on MG side - @ 0.50 m3/s

(c) Inert gas injection at 220m behind the face on TG side - @ 0.50 m3/s

O2 level %

Injection point

MG

TG

Injection point

MG

TG

Injection point

MG

TG

Injection point

(d) Inert gas injection at 220m behind face on both MG and TG sides - @ 0.25 m3/s each side

Figures 3.18 Effect of inert gas injection locations on goaf inertisation –base model 4


BOI.036.001.0053

37

3.4.4 Inert Gas Injection by Using Surface Goaf Holes In case of access problems in underground workings, such as after withdrawal of personnel from underground workings, the only possible means of goaf inertisation may be via surface drilled boreholes or those existing surface goaf holes for goaf gas drainage. CFD simulations were conducted to investigate if goaf inertisation can be achieved by injecting inert gas through surface goaf holes or in combination with inert gas injection in underground workings. Figure 3.19 shows the simulation results of goaf inertisation via goaf hole injection of boiler gas at various locations and flow rates.

(a) Inert gas injection at 0.25m3/s via surface borehole – 120m behind face MG side

(b) Inert gas injection at 0.50m3/s via surface borehole – 120m behind face TG side

TG

Injection point

MG

TG Injection point

MG

O2 level %

(c) Inert gas injection at 0.25m3/s via each borehole – 120m behind face MG/TG side

Injection points

TG

MG

(d) Inert gas injection at 0.25m3/s via each borehole – 220m behind face MG/TG side

TG

MG

Injection points

Figures 3.19 Effect of inert gas injection via surface goaf holes on goaf inertisation


BOI.036.001.0054

38

The modelling results indicate that an improved goaf inertisation effect could be achieved by injecting inert gas via surface boreholes. In this case, the modelling results show that by injecting inert gas on both side of the goaf via surface goaf holes would produce a better inertisation result than simply injecting inert gas on one side of the goaf. Similarly it is also important to understand goaf gas distribution patterns so that the correct injection locations can be identified to achieve the maximum goaf inertisation effectiveness.

3.5 DEVELOPMENT OF PROACTIVE INERTISATION STRATEGIES Taking into consideration the results of above CFD simulations, further modelling studies were conducted to develop effective and optimum pro-active inertisation strategies appropriate for Mine A and Mine B field site conditions. The results of these modelling studies are presented in this section and the results of field studies are presented in Chapter 5. 3.5.1 Mine A simulations The goaf gas at this mine typically consists of carbon dioxide (CO2) and methane (CH4) with a CO2:CH4 ratio between 65:35 and 35:65. The normal goaf gas distribution pattern in the panel after face retreat of about 500 m is shown in Figure 3.20. The gas monitoring data indicate that oxygen levels were high all the way up to the face start-up line on the MG side of the goaf. This high oxygen ingress led to low level heating in the goaf and an effective proactive inertisation strategy was required to reduce the oxygen ingress into the goaf.

Figure 3.20 Typical goaf gas distributions in longwall panel of Mine A

Gas (%)O2

14.4

67.8

3.2

49.5

6.4

54.8

2.2

41.2

8.216.7

CO2 CH4 N2

20.1

77.2

0.8 0.2

17.9

76.2

2.9 1.6

17.2

73.1

2.3 4.4

6.4 29.5 22.5

29.1 8.8

21.9

Goaf

MG

TG


BOI.036.001.0055

39

CFD simulations were conducted in conjunction with field demonstration at mine site to develop the most optimum pro-active inertisation strategies. The following parameters were used during CFD model development:

• MG at 6m higher elevation than TG and face 60m higher than start-up area • Face ventilations around 80 m3/s • Goaf gas emission at 1500 l/s with 65% of CO2 and 35% CH4

Figure 3.21 shows the CFD model layout and the perimeter ventilation arrangement. The arrow shows the direction of panel ventilation and face advancing direction. Goaf holes (e.g. TG1 to TG6 and MG1 to MG6) were incorporated into the models at various distances for parametric studies of goaf inertisation injection.

TG6 TG5 TG4 TG3 TG2 TG1

MG6 MG5 MG4 MG3 MG2 MG1 Maingate

Tailgate


Figure 3.21 CFD model layout and the ventilation arrangement at Mine A The results of base model simulations showing the oxygen distribution in the goaf under normal conditions for the Mine site A are presented in Figure 3.22. The modelling results indicate that airflow penetrates into the goaf up to some 400 m behind the face on intake side of the goaf. The simulations were carried out with uniform pressure drop along the perimeter roadway (i.e., no regulators in the perimeter roadway behind the face), as a worst-case scenario with respect to sponcom risk in the goaf. The result also indicates that the perimeter ventilation arrangement is a good option for goaf gas management but particular attention is needed to cut-through seals to minimise air leakage and the use of proactive inertisation to control the occurrence of spontaneous heating. The predicted oxygen distribution pattern in the longwall goaf agreed well with the field goaf gas monitoring data given in Figure 3.20 and those in Chapter 5. A number of CFD simulations were carried out to investigate the effect of inert gas injection at various locations and at different flow rates under the mine site conditions. The results of some typical studies under the following two scenarios are presented in this section.

Inert gas injection via cut-through on the MG side • • Inert gas injection via surface boreholes


BOI.036.001.0056

40

(a). Oxygen contours at seam level – plan view

TG

MG

TG

MG

O2 level %

(b). Oxygen contours at seam level – 3D view

Figure 3.22 Oxygen ingress pattern in longwall panels at Mine A (a) Scenario 1 - Inert gas injection via cut-through(s) on the MG side

The most practical access for goaf inertisation at the earlier stage of the longwall panel is via the cut-through(s) on the maingate side, because of the use of perimeter ventilation system. CFD simulations were conducted to study the injection of inert gas via these cut-through(s) at some varying distances the longwall face. Figure 3.23 shows the effect of goaf inertisation injection by injecting boiler gas at a flow rate of 0.5m3/s via cut-through’s at 300 m and 500m behind the longwall face on the maingate side. The modelling results show that an effective goaf inertisation has been achieved in both the cases as shown by the significant reduction in


BOI.036.001.0057

41

oxygen ingress area in the goaf behind the chocks. Considering the site access conditions at the kick-off stage of LW1 panel, i.e., the existence of a concrete drop hole, the most optimum goaf inertisation strategy appears to be injecting inert gas via the 23rd/22nd cut-through, at some 400m to 500m distance from the face line. The critical issue for the success of goaf inertisation is to minimise the air leakage through the cut-through seals to avoid oxygen ingression and excessive dilution of the inert atmosphere.

TG

MG

Injection point

O2 level %

(a) Inert gas at 300 m behind face line @ 0.5m3/s

Injection point

TG

MG

(b) Inert gas at 500 m behind face line @ 0.5m3/s

Figures 3.23 Effect of inert gas injection through cut-throughs deep in the goaf at Mine A. (b) Scenario 2 - Inert gas injection via surface boreholes

As the longwall panel retreats further from the start-up line, proactive goaf inertisation may need to be continued to suppress spontaneous combustion heating in the goaf and inert gas injection via surface goaf holes would be a practical option, particularly on the TG side due to access problems in underground. CFD simulations were carried out to study this option and assess the optimum strategies for goaf inertisation. Figure 3.24 shows a summary of the various inert gas injection options in this scenario.


BOI.036.001.0058

42

(a) Base model

(b) Inert gas at 0.5m3/s through MG2 – 250m behind face line

(c) Inert gas at 0.15m3/s through MG3 – 420m behind face line

MG

TG

O2 level %

MG

TG MG2

MG

TG MG3

MG

TG

MG4

(d) Inert gas at 0.5m3/s through MG4 – 600m behind face line

(e) Inert gas at 0.5m3/s through TG3 – 420m behind face line

MG

TG TG3

Figures 3.24 Effect of inert gas injection into the goaf through surface boreholes at Mine A.


BOI.036.001.0059

43

The modelling results indicate that injection of inert gas at a rate of 0.5m3/s via a surface hole at 250m behind the face would produce a good goaf inertisation effect (Figure 3.24b). The zone of high oxygen level has been considerably reduced to a narrow zone just behind the face in this case. Figure 3.24c shows the effect of injecting inert gas at a low rate of 0.15m3/s via borehole MG3 at some 420m behind the face. The modelling indicates that this option will help deep goaf inertisation but only marginal effects on areas immediately behind the chocks. The results indicate that injection of inert gas at a rate of 0.5 m3/s through borehole MG4 (600m behind face line), or through TG3 (420m on TG side behind the face line) would both produce some good inertisation deep in the goaf. A combination of these options would be the optimum strategy for preventing heatings in the goaf, which has been the case in longwall at Mine A. In addition to the correct inert gas injection points, the modelling results also indicate that it is critical to have a tight control of goaf leakage over these cut-through seals for the ultimate success of proactive goaf inertisation at low inert gas flow rates. 3.5.2 Mine B simulations Due to geotechnical and roof control problems, the tailgate (TG) was collapsed and as a result the longwall retreat was stopped for the installation of a new tailgate. While the face retreat was at a very low rate and eventually put on hold, signs of spontaneous heating (high CO and H2) were detected and pro-active goaf inertisation was considered to suppress the heating development. To assist the inertisation process in the field, CFD models were developed to model the goaf gas flow mechanics at some 500m from the start-up line, with the objective of optimising the various inertisation parameters, in particular, the injection points and injection rates behind the goaf. Two scenarios were simulated in this study:

Scenario 1 - TG closed due to roof fall and; • • Scenario 2 - TG open after reinstallation

(a) Scenario 1 - TG closed due to roof fall

The goaf gas distribution before the tailgate collapse is shown in Figure 3.25. It can be seen that oxygen ingress into the goaf was very high with O2 levels over 17% even at 400 m behind the face. The typical gas distribution in the goaf with just 10 m3/s auxiliary ventilation on the face after collapse of tailgate is shown in Figure 3.26. During this period, face ventilation was provided via auxiliary ventilation ducting system to the tailgate side and then travelled back along the face to the maingate. It can be seen from the goaf gas monitoring data that airflow ingress into the goaf was quite deeper even in this case, and the oxygen concentration level was over 10% of the goaf gas, even at some 350 m behind the face line. This deep penetration of airflow provides a favourable situation for spontaneous combustion to develop in the goaf.


BOI.036.001.0060

44

16.5 2.3 0.4 80.7

20.2 1.1 0.1 78.6

Gas (%) O2

CH4 CO2 N2

Longwall start-up

20.7 0.7 0.1 78.5

19.9 1.1 0.1 78.7

17.8 1.6 0.4 80.1


Previous panel goaf

Face

TG

MG

Tube bundle point

Longwall Goaf Dip

Figure 3.25 Typical goaf gas distributions at Mine B – before TG collapse

Dip

Longwall start-up

18.8 2.8 0.2 78.2

6.1 5.4 2.9

85.6

17.4 2.4 0.4 79.8

11.54.4 0.8

83.2

10.43.7

1.0 84.8

8.84.6 1.5

85.0

TG

Q=10m3/s

Longwall Goaf

Auxiliary ventilation ductFace Retreat Direction

MG

Gas (%) O2

CH4 CO2 N2

Figure 3.26 Typical goaf gas distributions at Mine B – TG closed

Start-up

Face TG

MG

MG C/Ts

TG C/Ts

GOAF TG1 – sborehole

urface

MG1 – surface gas drainage borehole

MG2 – surface gas

drainage borehole

Figure 3.27 CFD model layout for Mine B - TG closed


BOI.036.001.0061

45

Figure 3.27 shows the layout out of the CFD models. Based upon field data, the following was assumed for the CFD models:

TG at 15m higher elevation than MG and the face at 5m higher than the start-up line • • • •

Total goaf gas emission = 350l/s with 100% of methane (CH4) only Face ventilation = 10m3/s delivered by auxiliary ventilation. Three surface boreholes in the goaf, two were originally drilled for goaf gas drainage and the third for goaf inertisation.

Figure 3.28 shows the base model simulation results for Mine B. The modelling results indicate that airflow penetrates deep into the goaf, in particular on the MG side, oxygen level is about 10% at the 4th C/T in the model, i.e. about 350~400m behind the face line. The predicted goaf oxygen distribution pattern tailed well with the field goaf gas monitoring data given in Figure 3.26. The base model was then used to evaluate and optimise goaf inertisation options in conjunction with field operations.

Face

TG

MG

O2 level

0.21 = 21%

(a) Plan view of the oxygen ingress pattern into the goaf

TG Face MG

(b) 3D view of the oxygen ingress pattern in the longwall goaf

Figure 3.28 Oxygen ingress pattern at Mine B - TG closed


BOI.036.001.0062

46

TG1 TG

MG

MG1

TG

MG

(a). Inert gas - MG1 @ 0. 5m3/s (b). Inert gas – TG1 @ 0. 5m3/s O2 level

%

MG2

TG1 TG

MG

(c). Inert gas – MG2 @ 0. 5m3/s (d). Inert gas – TG1 & MG2 @ 0.15m3/s each

MG2

TG

MG

MG

TG1

5th C/T

TG TG1

MG2

TG

MG

(e). Inert gas – TG1 & MG2 @ 0.25m3/s each (f). Inert gas – TG1 0.5m3/s & 5c/t @ 0.15m3/s

Figures 3.29 Optimisation of inert gas injection options at Mine B – TG closed CFD simulations were carried out to evaluate a range of options for injecting inert gas into the goaf, particularly via the surface boreholes which were the most practical channels to pump inert gas into the goaf at the time of the incident. Figure 3.29 shows the effect of goaf inertisation injection at different locations with boiler gas and nitrogen on oxygen ingress pattern within the goaf. The red arrow indicates the inert gas injection point(s) in the simulations. The modelling results indicate that inertisation injection via a single borehole, i.e., MG1 or TG1 alone with boiler gas at 0.5m3/s would not achieve the effective goaf inertisation, although the injection is having some good impact around the injection point. The optimum inertisation option would be a combination of inert gas injection via borehole MG2 or TG1 and MG2 or the 5th C/T, as shown in Figure 3.29 (c), (e) and (f). It the can be seen that with this combination even at a low inertisation flow rate, i.e., 0.15m3/s via each borehole, a reasonably satisfactory goaf inertisation result could be achieved (Figure 3.29d).


BOI.036.001.0063

47

The optimum goaf inertisation strategy appears to be injecting inert gas simultaneously via TG1 and MG2 with boiler gas at 0.25m3/s in each or via TG1 and 5th C/T with boiler gas at 0.5m3/s and nitrogen at 0.15m3/s respectively. Both inert gas flow rates are achievable by current inert gas generation systems on mine site. (b) Scenario 2 - TG open after reinstallation

Following the development of a new tailgate, face ventilation was restored to its original U system with a total airflow of 30m3/s. Figure 3.30 shows the typical gas distribution in the longwall goaf with standard U ventilation. It can be seen that airflow ingress into the goaf was much deeper in comparison with Scenario 1, with the oxygen concentration level well over 16% of the goaf gas, even at some 350 m behind the face line. This may be due to the air leakage across the cut-through seals as well as other effects such as buoyancy of goaf gas. Again this deep airflow ingress into the goaf provides a favourable situation for spontaneous combustion to develop when the face was retreating at a slow rate.

Longwall start-up

20.2 1.1 0.1 78.6

19.9 1.1 0.1 78.7

16.52.3 0.4

80.7

17,81.6 0.4

80.1

20.7 0.7

0.1 78.5

Dip

TG

MG

Longwall Goaf Face Retreat Direction Q=30m3/s

Gas (%)O2

CH4 CO2 N2

Figure 3.30 Typical goaf gas distributions at Mine B – with U ventilation On the basis of above studies in Scenario 1, CFD simulations were conducted to investigate additional options for injecting inert gas into the goaf, as a result of the change-back in face ventilation system.The same model layout as in Scenario 1 was used, with the only addition of the restored tailgate. All the assumptions for Scenario 1 remained the same except the ventilation system was changed to U pattern with an airflow rate of 30m3/s. Figure 3.31 shows the base model simulation results in this scenario. The modelling results indicate that airflow penetrates much deep into the goaf, with oxygen levels around 15% at the 5th C/Ts in the model, i.e. about 350m behind the face line. The predicted goaf oxygen distribution pattern was in good agreement with the field goaf gas monitoring data given in Figure 3.30. Figure 3.32 shows the effect of goaf inertisation injection at different locations with boiler gas and nitrogen on oxygen ingress pattern within the goaf. The red arrows show the inert gas injection point(s) in the simulations. The modelling results show that inert gas injection via the 3rd and 5th cut-through at some 250-350m behind the longwall face can lead to a good goaf inertisation effect, as already demonstrated in other simulations presented in this chapter.


BOI.036.001.0064

48

However, inert gas injection via a single borehole at the same flow rate (0.5m3/s), e.g. TG1 on the TG side in this case, is unlikely to achieve a total goaf inertisation effect. Again an improved inertisation result could be achieved by injecting boiler gas and nitrogen simultaneously via TG1 and MG2 or via C/Ts behind the 3rd C/T on the MG side. The optimum goaf inertisation location appears to be injecting inert gas simultaneously via boreholes TG1 and MG1 or TG1 and MG2 at a total flow rate of 0.5 m3/s to 0.7 m3/s. The inert gas flow rates are critical to this option, as illustrated in Figure 3.32 (e) and (f), a flow rate of 0.25m3/s via borehole MG1 was insufficient to provide a good inertisation on the MG side, unless the flow rate is increased to 0.5m3/s.

TG

MG


O2 level %

TG

MG

(b) 3D view of the oxygen ingress pattern in the longwall goaf

Figure 3.31 Oxygen ingress pattern at Mine B – after TG restoration


BOI.036.001.0065

49

MG

TG

5th C/T

(a). Inert gas – 3c/t @ 0. 5m3/s (b). Inert gas – 5c/t @ 0. 5m3/s

(c). Inert gas – TG1 @ 0. 5m3/s (d). Inert gas – TG1 & MG2 @ 0.15m3/s each

O level 2%

MG

TG

3rd C/T

TG1

MG

TG

MG2

MG

TG TG1

(e). Inert gas – TG1 & MG1 @ 0.25m3/s each (f). Inert gas – TG1 0.25m3/s & MG1 @ 0.5m3/s

TG

MG

MG1

TG1

MG

TG

MG1

TG1

Figures 3.32 Optimisation of inert gas injection options at Mine B - after TG restoration

In brief, CFD modelling simulations were conducted to assist in the development of optimum goaf inertisation strategies for reducing the oxygen ingress into the goaf for both Mine A and Mine B conditions. The oxygen reduction in the goaf is an effective management strategy for reducing spontaneous combustion risk in the longwall panels. The optimum proactive inertisation options identified by CFD simulations have been implemented at both the field sites during the field demonstration studies. Details and results of the field demonstration results are presented in Chapter 5 of this report. 3.6 Summary and Conclusion Extensive CFD simulations were conducted for a range of longwall layouts and gas emission conditions based upon several Australian underground coal mines. These base models were calibrated with available field gas data and used to investigate the best inertisation strategies


BOI.036.001.0066

50

that could be deployed to narrow down the high oxygen level zones which are potentially liable to spontaneous combustion in the goaf. These studies have greatly improved the fundamental understanding of goaf gas flow patterns and inert gas dispersion mechanism in the longwall goaf and thus assisted in the development of innovative goaf inertisation strategies for active longwall panels. The investigations involved extensive parametric studies on inert gas compositions, injection locations, inert gas flow rates as well as the impact of goaf gas emissions, seam dips, face orientation and ventilation systems. In general, the studies indicate that inertisation through the cut-through seals at some 200m behind the face would be more effective than that at close range immediately behind the face line. Further simulations with other inert gas injection location scenarios indicated that inert gas injection even at other deeper locations in the goaf, between 200 to 400 m behind the face, results in effective goaf inertisation for the modelled longwall panel conditions. Although the exact optimum location for inert gas injection for any longwall panel depends on site specific parameters, these modelling simulations indicated that inert gas injection between these ranges (around 200 to 400 m behind the face) would be far more effective than inert gas injection close to the face. Analysis of the various simulation results also indicated that longwall panel geometry, goaf gas emission rates and composition, ventilation layouts, pressures and flow rates, goaf characteristics, and gateroad conditions in the goaf would also have a significant influence on goaf inertisation. Although inert gas injection into the goaf at low flow rate of 0.15 m3/s would also significantly reduce oxygen ingress into the goaf in some cases, modelling investigations indicated that an inert gas flow rate of around 0.5 m3/s would be required for goaf inertisation in most cases. Results also indicated that higher inert gas flow rates might be required in some cases, depending on the site specific conditions. Goaf inertisation can also be carried out with surface boreholes when underground access becomes prohibitive or impossible. The modelling results indicate that inertisation via surface goaf holes would be more effective in narrowing down the sponcom liable zones, a combination of inert gas injection via cut-through(s) and surface borehole(s) would further improve the effectiveness of goaf inertisation. A good understanding of the goaf gas flow mechanics, the selection of the correct injection points and inert gas flow rates, as well as a good management of longwall face ventilation drop across the goaf and goaf leakage, are all critical to the success of goaf inertisation. Knowledge obtained from the CFD modelling studies was used in conjunction with field conditions to develop the optimum pro-active goaf inertisation schemes for two Australian coal mines. The details and results of these field studies are described in Chapter 5.

- - - - -


BOI.036.001.0067

51

CHAPTER 4

STUDIES ON OTHER OPTIONS TO REDUCE OXYGEN INGRESS

4.1 INTRODUCTION Goaf gas distribution monitoring studies carried out at different mines have clearly shown that oxygen penetration into the goaf area behind the face occurs in most of the longwall mines. The degree of penetration depends largely on ventilation layouts, leakage around seals, goaf breakage and consolidation parameters and barometric fluctuations. The traditional methods for reducing the size of the air fringe on the maingate or tailgate perimeter of the goaf involves mostly ventilation changes (i.e. adjusting face or perimeter road quantities) or installing intake regulators close to the face. Longwall faces also employ brattice curtains in between the first chock and the rib line to prevent air penetrating into the goaf. The main objective of the project is to develop and demonstrate the proactive inertisation technologies involving injection of inert gas at optimum locations to reduce oxygen ingress into the longwall goafs. The secondary aim of this project is to conduct a preliminary investigation of the effect of other alternative technologies such as high expansion foam plugs on air ingress patterns in the goaf areas. This chapter presents the details and results of these preliminary investigations conducted at two different mine sites on alternative technologies. The first set of field trials involved injection of high expansion foam through maingate seals to investigate its effect on goaf oxygen ingress patterns. The second set of trials involved injection of both foam and inert gas to investigate the effect of foam on goaf inertisation. The background information on field sites and results of investigations are presented in the following sections.

4.2 BACKGROUND OF FIELD SITES – MINE C & MINE D 4.2.1 Mine C Mine C is one of the gassiest mines in Australia and is located in the Hunter Valley in New South Wales. The mine has a production capacity of 4.0 million tonnes per year and operates a single longwall face and 2 continuous miner sections. The coal seams of the working section, floor and immediate roof average a total of 20 m in thickness. The depth of the workings ranges from 200 to 400 m. All seams in the mine contain a multi-component seam gas comprising mainly carbon dioxide and methane. Gas contents range from 6.0 m3/t to 11.0 m3/t with CO2/CH4 ratios ranging from 90/10 to 60/40. The coal seams at this mine are classified as having medium to high propensity for spontaneous combustion. The mine layout is primarily a two entry gateroad system, except near the mains.


BOI.036.001.0068

52

The longwall faces are 200 m wide with panel lengths in the range of 1,800 to 2,300 m. The extraction height of the longwall is about 4.5 m. Gateroad pillar width ranges from 30 m to 45m with cut-throughs normally spaced at 100 m intervals. The mine’s primary ventilation circuit comprises two intake drifts and a single exhaust shaft. Three surface exhaust fans are used to supply about 340 m3/s of air into the mine at 1.2 kPa pressure. To control the gas levels at return end of the face, a back-return ventilation system was adopted at the mine. In this system, most of the intake air traversing the face goes towards the longwall return through the cut-through behind the face. 4.2.2 Mine D The Mine D is located in the Bowen Basin in Queensland. The mine operates a single longwall face employing 2 leg high reach 1000 T capacity chocks and produced about 5.5 Mt in 2000. The longwall mining height is about 4.5 m. The width of the longwall panels was about 250 m and the length ranged from 1,600 m to 2,500 m. The depth of the longwall panels ranges from 80 m to 250 m in the current mining block. The mine layout was primarily a 6-roadway system for mains and a 2-entry gateroad system for longwall panel development. The gas content of the coal seams ranges from 4 m3/t to 7 m3/t and consists mostly of CH4 gas. The mine employed a long hole pre-drainage system and was able to extract a higher proportion of the in-situ gas before longwall extraction due to high permeability of the coal seams in that region, which ranges from 10 to 30 millidarcy (md). Therefore, goaf gas emissions in the longwall were relatively lower, ranging from 100 l/s to 500 l/s. The mine did not employ any post-drainage system in the longwall panels. All the pre-drainage holes were connected to a surface borehole and pre-drainage gas vents to the atmosphere with positive pressure. No gas drainage plants were installed at the mine. Seam gas pressure and high gas desorption rate seemed to be high enough for free venting of the pre-drainage gas. The mine has a history of heatings in the pillars and a high rate of CO production in the longwall goafs. The orientation of the north-side longwall panels was such that outbye tailgate corner was the point of lowest elevation in the panel. The mines primary ventilation circuit comprises of four intake drifts from the highwall and a single exhaust/return airshaft. Two surface exhaust fans are used to supply about 240 m3/s of airflow at 1.1 kPa pressure. During longwall panel extraction, approximately 50 m3/s of airflow was supplied to the face through maingate and during the trial. A 30 point tube bundle monitoring system was installed at the mine for continuous environmental monitoring in the longwall panels. The longwall panel extraction at this mine started with 4th panel on south side and continued towards 3rd, 2nd and 1st panels. Similarly panel extraction on the north side started with 4th panel extraction. The mine then moved down dip to extract 7th, 6th and 5th panels. In view of this panel extraction, the bottom gateroad in the panels, which was located adjacent to the old goafs, was used as tailgate return airway. Top gateroads in the panel, which were located adjacent to the virgin blocks, were used as maingate intake airways. The 5th panel on the Northern side was located in between 2 previously mined longwall panels.


BOI.036.001.0069

53

4.3 FOAM INJECTION STUDIES Foam injection trials were conducted at two underground coal mines by injecting high expansion foam into the goaf perimeter at select locations behind the retreating longwall face to test it’s effectiveness in reducing the oxygen penetration into the goaf. The effect of the foam injection was measured using the mines’ tube bundle monitoring systems with monitoring points located in key locations around the panels being tested. High expansion foam was generated using the mines compressed air and water supplies. Hitron high expansion foam from Orion was used as this foam has a high expansion ratio (40 to 1), long drain time, can be evenly distributed and has good stability in non ventilated applications. Notably Hitron foam has a good tolerance of hard or salty water. The foam is mixed at a ratio of 2% foam concentrate to water then compressed air is introduced. The foam was delivered to the sites through 150mm victaulic pipe. Typical foam delivery rates were around 140 m3/hr. To achieve this production rate the following air, water and foam were required.

• Water flow rate 60 litres per minute • Foam Flow rate1.3 litres per hour • Airflow of 19 to 20 litres per second

4.3.1 Foam Trials at Mine C The first trial was conducted at Mine C in August 2004. At the time of testing the air fringe present on the maingate side of the longwall goaf extended back a distance of 400m. High expansion foam was injected into 6 c/t in the maingate of the longwall panel, the face position at the time was 50 m outbye of this cut through. Figure 4.1 shows the location of the trial in the longwall panel.

(a) Details of foam trial

(i) Monitoring: Tube bundle monitoring points were located at 7c/t, 9c/t and 15c/t inbye of 6c/t on the maingate and at two locations 12c/t and 10c/t in the tailgate (ii) Panel Ventilation: The Mine A longwall face is ventilated with around 80m3/s of air supplied through the maingate and tailgate and returning through the second tailgate roadway. A perimeter roadway is maintained around the panel and is ventilated with around 20m3/s of air. (iii) Foam Volume: Cut through volume behind seal 7x5x4 = 140m3

Vol adjacent to cut through 2.5x5x4 = 50m3 (assume half roadway width) Expansion vol either side of c/t = 200m3 (assumed) The target minimum foam volume to fill any void in the cut through and plug up the goaf perimeter was set at 390m3.


BOI.036.001.0070

54

Figure 4.1 Location of the trial site in longwall panel and tube bundle monitoring points

Foam injection location

Tube bundle points

(b) Results at Mine C

The results from the trial show a marked change in the composition of the goaf gases inbye of the foam injection site. In total a volume of 440m3 of foam was injected in to the 6c/t seal site over a four hour period. As atmospheric pressure is known to have a large impact on goaf gas migration it was recorded during the trial and is shown in Figure 4.2.

984

986

988

990

992

994

996

998

1000

04:00:01

06:30:02

09:00:02

11:30:02

14:00:02

16:30:02

TIME

hPa

BARO

Foam injection

Figure 4.2 Barometric pressure changes during the Mine C trial


BOI.036.001.0071

55

Maingate Results The changes in gas concentration values recorded on maingate side during foam trial are presented in Figures 4.3 to 4.5. Analysing the results obtained from the tube bundle system we can distinguish a number of changes occurring due to the introduction of a foam plug at the maingate. At the tube bundle monitoring points located on the maingate side changes in the gas composition were different closer to the injection site as compared to the monitoring point located deeper in the goaf. The monitoring point showing the most change from the foam injection was located at 9c/t some 300m inbye of the injection location (assuming a linear change in the effects of the foam plug making the results from 8c/t less than those recorded at 9c/t). In this case we see the CO2 level rise from 6.5% to 18% (see Figure 4.4) and correspondingly the O2 level drops from 18.5% to 15%. From the monitoring results we also see an effect on the monitoring results recorded at 15c/t in the maingate some 900m inbye of the foam injection however in this instance CO2 drops from 19% down to 17% and O2 rises from 7.0% to 9.5% (see Figure 4.5). The area in between the face and 9c/t does show some effects as shown in Figure 4.3; however it would seem that the influence of the face ventilation and scouring of the immediate goaf area behind the face is still dominant in the area up to 200m behind the face. Monitoring results over the following days showed the goaf area at 9c/t went inert due to normal goaf action 72 hours after the foam injection trial and remained stable. To enable the 9c/t seal to become inert by natural means (self inertisation) the longwall face had to retreat a further 60m from its position during the foam trial.

LW110 7C/T

0

5

10

15

20

25

3:40:05AM

5:44:27AM

7:48:50AM

9:53:12AM

11:57:34AM

2:01:57PM

4:06:19PM

TIME

%

CH4 CO2 O2

Foam injection

Figure 4.3 Gas levels recorded at 7 c/t during foam trial


BOI.036.001.0072

56

LW110 9C/T

02468

101214161820

3:45:05AM

5:49:27AM

7:53:50AM

9:58:12AM

12:02:34PM

2:06:57PM

4:11:19PM

TIME

%

CH4 CO2 O2

Foam injection

Figure 4.4 Gas levels recorded at 9 c/t during foam trial (Foam injection showed the biggest influence at this point 300m inbye of the test location)

0

5

10

15

20

25

3:53:05AM

5:57:27AM

8:01:50AM

10:06:12AM

12:10:34PM

2:14:57PM

4:19:19PM

TIME

%

CH4 CO2 O2

Foam injection

Figure 4.5 Gas levels recorded at 15 c/t during foam trial


BOI.036.001.0073

57

Tailgate Results The gas concentration values recorded at monitoring points located in the tailgate (Figures 4.6 and 4.7) showed no changes in the trends during the foam trial.

LW110 12C/T TG

05

101520253035404550

3:45:02 6:15:02 8:45:02 11:15:02 13:45:02 16:15:02

TIME

%

CH4CO2O2

Foam injection

Figure 4.6 Gas levels recorded at TG 12 c/t during foam trial

LW110 10C/T TG

05

101520253035404550

3:45:02 6:15:02 8:45:02 11:15:02 13:45:02 16:15:02

TIME

%

CH4CO2O2

Foam injection

Figure 4.7 Gas levels recorded at TG 10 c/t during foam trial


BOI.036.001.0074

58

4.3.2 Foam Trials at Mine D A second foam only trial was conducted at Mine D in May 2005. The purpose of this trial was to measure any change in the pressure acting on the goaf seals adjacent to the injection area due to the presence of a foam plug as experience had shown that the installation of belt flaps in the maingate as an intake regulator were capable of inducing a 10 to 15pa pressure change on the seals. The trial began at approximately 10am and concluded at 4pm. The injection site was located at 3c/t on the maingate side of the panel as shown in Figure 4.8. Several different circumstances were encountered in this foam injection trial. Firstly the distance from the end of the injection pipe to the goaf edge was higher than that previously attempted at both the Mine C and Mine D trials. Secondly the presence of both a high gas make and an intake regulator in the maingate travel road (located between 8 and 9 cut through) had brought the goaf fringe up to within 100 to 150m of the retreating face. With the gas fringe being this close to the retreating face a significant alteration in the mixture of gases present due to foam injection was not expected. However due to the likelihood that the placement of a foam plug in the goaf does create a restriction to ventilation and therefore a pressure differential it was the aim of the trial to measure this pressure change by measuring the pressure at the seals before and during the injection of foam.


Tube bundle points Pressure measurement points

Regulator

Figure 4.8 Location of the Mine D trial and tube bundle monitoring points


BOI.036.001.0075

59

(a) Details of foam trial

(i) Monitoring: Tube bundle monitoring points were located on the maingate side of the panel at 5c/t, 6c/t and 7c/t. Seal pressure readings were recorded at 5c/t, 6c/t, 7c/t, 8c/t and 9c/t

(ii) Panel Ventilation: The longwall panel was ventilated with a standard U type system delivering 50m3/s to the face via the two maingate roadways. The travel road inbye of the face was ventilated with 25m3/s.

(iii) Foam Volume: The target foam volume for the trial was set at 550m3 as the delivery point of the foam was further away than in the first trial due to the chain pillar being 50m wide instead of the usual 30m wide chain pillar.

(b) Results at Mine D

Results from the Mine D foam only trial show that during the trial the barometer fluctuated significantly falling 350 Pascals in 6 hours (see Figure 4.9). During the foam injection trial the initial seal pressure readings were recorded for the 5 seals immediately inbye of the test location, all seals were found to be breathing in or under negative pressure. During the foam trial the seal pressures were measured at two and three and four hours after commencement of foam injection the pressure results are displayed in Figure 4.10. Note that the negative pressure on the seals inbye of the regulator located in between 8 and 9c/t increased during the trial.

985

986

987

988

989

990

991

12:00

AM

2:00 A

M

4:00 A

M

6:00 A

M

8:00 A

M

10:00

AM

12:00

PM

2:00 P

M

4:00 P

M

6:00 P

M

8:00 P

M

10:00

PM

12:00

PM

Time

hPa

Figure 4.9 Barometer change during the test at Mine D


BOI.036.001.0076

60

S6 Maingate Seal Pressures

0

5

10

15

20

25

30

35

40

45

5 6 7 8 9

Cut through Location

Pre

ssur

e in

Pas

cal i

nto

the

Goa

f

Initial12:30PM1:30PM

Figure 4.10 Maingate seal pressure reading during the foam trial

4.4 FOAM AND INERT GAS INJECTION STUDIES Mine D has used a Tomlinson Inert gas Generator to inertise its longwalls goaves since the first panel was completed in 1999. Since that time several advances have been made in the way the inert gas was applied to the goaf and the corresponding sealing sequence. Logically the next step is to formulate a strategy where by the inert gas can be applied to an active goaf situation where the longwall face is still operating. The Tomlinson inert gas generator produces 0.5 m3/s of inert gas at 20 degrees above ambient temperature and the inert gas is composed of the following

Oxygen 1.0 to 2.0% Carbon Dioxide 14.0% Carbon monoxide 2 to 10ppm Nitrogen 84%

The gas was generated on the surface and piped through borehole and a six inch range for 800m before finally being delivered through 200m of 4 inch poly line. This line was then connected to the mines 6 inch galvanised pipe range for final delivery to the seals. Figure 4.11 shows the location of the trials in the panel.


BOI.036.001.0077

61

Tracer gas injection location Tube bundle points

Figure 4.11 Location of the foam and inert gas trials in the longwall panel (The tracer gas injection point remains the same for both tests)

Two different trials were conducted at the Mine D:

(a) The first involved injecting inert gas (0.5 m3/s of boiler gas) at a location some 120m behind the longwall face and measuring its effect at the maingate seal sites. In both cases the inert gas was injected through a 150mm pipe in the maingate seal. During this trial a tracer gas study was conducted to record the airflow patterns present in the goaf prior to any injection of high expansion foam.

(b) The second combined inert gas injection and the placement of high expansion foam to

reduce the size of the area behind the face required for inertisation. During this stage of the trial the second tracer gas study was conducted to track the path of the inert gas after a plug of high expansion foam had been generated.

The panel is monitored via a tube bundle system and instantaneous electronic monitors. For the trial 7 points in the working panel goaf and 1 point in the adjacent sealed goaf were utilised. Around 300 bag samples were collected and analysed by SIMTARS for the presence of tracer gas. Several other bag samples were collected from the maingate seals of the adjacent goaf during the tracer trial. A total of eight points were monitored during the tracer


BOI.036.001.0078

62

gas studies. These monitoring points were located as follows: 6c/t to 10c/t and 18c/t on maingate side of the working panel goaf, a tailgate general body monitoring point and one point located at 5c/t in the maingate of the adjacent sealed panel. These monitoring points are shown in red in Figure 4.12.


Figure 4.12 Foam injection location after the face has retreated an additional 100m The longwall face is ventilated via a U system with 55 m3/s of air on the face. Due to the sequence of mining, the working panel is flanked by goaves on both the main and tailgate sides. A barrier pillar of 50m width was left on the maingate side by the development of a third roadway. The panel is being sealed by installing 2 and 20 psi rated stoppings across the B heading roadway on retreat. Total volume of foam injected is about 700 m3. 4.4.1 Inert Gas Injection along with Tracer Gas The first trial was conducted at Mine D in July 2004. The trial consisted of injecting inert gas into the working panel goaf via the maingate seals. The inert gas was injected at two locations (6c/t and 7c/t) during the trial this allowed for inert gas injection tests to be conducted over a range of distances behind the retreating face from effectively 0m to 150m behind the retreating longwall face. Inert gas was injected firstly through 6c/t when the face was 29m outbye of the cut through (chainage 371m). Inert gas was injected there for 3 days as the face retreated. After 3 days the injection point was switched to 7c/t for a period of a further 3 days.


BOI.036.001.0079

63

The first of the tracer gas trials was conducted at this time by injecting SF6 tracer gas into the inert gas delivery range and redirecting the inert gas injection point back to 6c/t. The redirection to 6c/t from 7 c/t took place at 13:00 hrs with the release of tracer gas at 13:10 hrs. The longwall chainage at this point was 316m. The results from the first tracer gas tests at Mine D are shown in Table 4.1. The tracer gas was released into the inert gas being injected through the 6c/t seal. The face at this time was located 30m outbye the 6c/t seal.

Table 4.1 Results of First Tracer Gas Test at Mine D

Date Time TG Time 6c/t Time 7c/t Time 8c/t Time 9c/t Time 10c/t Time 18c/t Time 5c/t N619/07/2004 13:21 <0.01 13:12 <0.01 13:15 <0.01 13:17 <0.01 13:19 <0.01 13:24 <0.01 - - - -

13:37 <0.01 13:27 <0.01 13:30 <0.01 13:33 <0.01 13:35 <0.01 13:40 <0.01 - - - -13:58 <0.01 13:48 <0.01 13:50 <0.01 13:53 <0.01 13:55 <0.01 14:00 <0.01 14:40 <0.01 - -14:25 <0.01 14:07 7.50 14:17 <0.01 14:18 <0.01 14:23 <0.01 14:26 <0.01 - - - -14:33 0.06 14:27 31 14:51 0.07 14:33 <0.01 14:35 <0.01 14:38 <0.01 - - - -14:57 0.02 14:49 76 15:09 0.08 14:53 0.01 14:55 <0.01 14:59 <0.01 15:01 <0.01 14:51 <0.0115:15 <0.01 15:07 111 15:11 0.03 - - 15:13 0.01 15:17 <0.01 15:19 <0.01 - -15:45 <0.01 15:37 108 15:39 0.11 15:41 0.03 15:43 0.01 15:47 <0.01 15:49 <0.01 - -16:15 <0.01 16:07 110 16:09 0.07 16:11 0.02 16:13 <0.01 16:17 <0.01 16:19 <0.01 - -16:45 0.03 16:37 109 16:39 0.09 16:41 0.01 16:43 <0.01 16:47 <0.01 16:49 <0.01 - -17:14 0.26 17:07 99 17:10 0.06 17:12 0.01 17:14 <0.01 17:16 <0.01 17:19 <0.01 - -17:45 1.00 17:37 93 17:39 0.05 17:41 0.02 17:43 <0.01 17:47 <0.01 17:49 <0.01 17:51 <0.0118:15 1.30 18:07 75 18:09 0.06 18:11 0.02 18:13 <0.01 18:17 <0.01 18:19 <0.01 - -21:51 2.10 21:43 18 21:45 2.50 21:47 <0.01 21:49 <0.01 21:53 <0.01 21:55 <0.01 21:57 2.5

20/07/2004 3:15 0.27 3:10 3.80 3:15 4.10 3:15 0.18 3:15 <0.01 3:15 <0.01 3:15 <0.01 3:05 0.048:15 0.09 8:00 1.10 8:03 2.20 8:06 0.70 8:10 <0.01 8:17 <0.01 8:12 <0.01 8:19 28.6

10:04 0.07 - - 10:00 2.00 10:02 1.00 10:04 <0.01 10:06 <0.01 10:08 <0.01 10:10 26.212:15 0.09 - - 12:05 1.80 12:07 1.10 12:10 <0.01 12:17 <0.01 12:12 <0.01 12:18 27.214:06 0.12 - - 14:00 2.40 14:02 <0.01 14:04 <0.01 14:08 <0.01 14:10 <0.01 14:12 23.516:06 0.12 - - 16:00 2.50 16:02 1.50 16:04 <0.01 16:08 <0.01 16:10 <0.01 16:12 23.517:55 0.04 - - 17:50 2.60 17:22 1.50 17:54 <0.01 17:58 <0.01 17:57 <0.01 18:00 21

First Tracer Gas Test

In Figure 4.13 we can see that the arrival time of the peak for the tracer gas to the return side monitoring point was 7.9 hours. Along the maingate side it can be seen that it took between 1 and 2 hours for the first signs of tracer gas to reach the maingate cut throughs 7, 8 and 9 (300m inbye from the injection location). The arrival time for the peak concentrations of tracer gas at the monitoring points was an order of magnitude greater taking up to 24 hours at the more inbye seals of 8 and 9c/t. Another important observation from the first tracer gas study is that tracer gas was also detected in the adjacent goaf on the tailgate side within 5 to 9 hours of its release. This result clearly indicates the existence of a leakage path into the old goaf through the chain pillar seals separating it from the active goaf and consequently that inert gas injected from a working maingate can reach adjacent goaves. A diagrammatic explanation is shown in by the red arrows in Figure 4.13.


BOI.036.001.0080

64

6 c/t INJECT

1 hr 401 hr 402 hrs

5-9hrs

7 hr 55

7 C/T9 C/T

5 C/T

TAILGATE

MAINGATE

GOAF

8 C/T

TG GB

1 hr Nil

10 C/T

Figure 4.13 Tracer gas arrival times at different locations during the first tracer gas test 4.4.2 Combined Foam and Inert Gas Injection along with Tracer Gas The trial of combined high expansion foam and inert gas injection was conducted in conjunction with a second tracer gas trial when the longwall chainage was 278m, or 22m outbye of 5c/t. At this point the foam plug was injected into 5c/t and inert gas was being injected into the goaf at 6c/t, once again tracer gas introduced into the delivery range of the inert gas. It is to be noted that a larger amount of tracer gas was injected during the second test, i.e., 20kg as opposed to 10kg in the first test, which increases the quantities of tracer gas recorded at the monitoring points but would not affect the arrival time of the tracer gas at the monitoring points. The results of the second tracer gas test are shown in Table 4.2. Figure 4.14 shows the arrival times of the tracer gas at different monitoring points in the goaf. The main difference between the first and second tests is that the longwall face had retreated a further 100m from the first test site to be located outbye of 5c/t and that high expansion foam was injected at 5c/t during the second trial. The second test results show a different inert gas flow pattern due to the injection of high expansion foam at 5c/t immediately behind the longwall face. The arrival time of the tracer gas peak at the tailgate monitoring point for the second trial (after the injection of high expansion foam) was recorded at 4.9 hours after release compared to 7.9 hours in the first trial. The arrival times of tracer gas at the monitoring points on the maingate side vary significantly from the first trial with tracer gas being present much sooner than the first trial. In the case of 10c/t and 18c/t no tracer gas at all was detected in the first trial however the second trial detected tracer gas at the seal sites within 1 hour.


BOI.036.001.0081

65

Table 4.2 Second Tracer Gas Test Results from Mine D

Date Time TG Time 6c/t Time 7c/t Time 8c/t Time 9c/t Time 10c/t Time 18c/t Time 5c/t N622/07/2004 10:26 <0.01 10:15 <0.01 10:17 0.06 10:20 1.10 10:24 0.10 - - - - - -

10:42 0.01 10:35 <0.01 10:37 0.05 10:39 1.10 10:40 0.09 - - - - - -11:04 <0.01 10:55 <0.01 10:57 0.04 11:00 1.10 11:02 0.10 11:06 0.01 - - - -11:23 <0.01 11:15 1.80 11:17 0.05 11:19 1.20 11:21 0.10 11:25 0.01 - - 11:27 1.811:45 0.03 11:35 84 11:37 0.10 11:39 1.00 11:41 0.09 11:48 0.01 11:43 <0.01 - -12:02 0.36 11:55 274 11:57 0.22 12:00 1.00 12:02 0.10 12:08 0.01 12:05 <0.01 - -12:28 0.05 12:19 372 12:20 3.90 12:24 1.70 12:26 0.11 12:30 0.02 12:27 0.02 - -12:52 0.02 12:45 348 12:46 1.30 - - 12:49 0.10 12:54 0.02 12:50 0.02 - -13:23 0.01 13:15 292 13:17 1.40 13:19 1.50 13:20 0.09 13:29 0.01 13:21 0.01 - -13:57 0.02 13:50 252 13:52 1.50 13:53 1.50 13:54 0.06 13:59 0.01 13:55 0.01 - -14:24 0.02 14:20 206 14:17 1.70 14:20 1.50 14:22 <0.01 14:26 <0.01 14:23 <0.01 - -14:53 0.15 14:56 154 14:45 1.40 14:47 1.40 14:50 0.04 14:55 <0.01 14:58 <0.01 - -15:55 2.30 15:45 103 15:48 1.60 15:50 1.50 - - - - - - - -17:10 2.00 17:00 75 17:05 1.20 17:07 1.90 17:08 0.04 17:13 <0.01 - - 17:15 1.6

Second Tracer Gas Test

6 c/t INJECT

Inst Inst Nil

N/A

4hr 55

7 C/T9 C/T

5 C/T TAILGATE

MAINGATE

GOAF

8 C/T

TG GB

Inst Nil

10 C/T

Figure 4.14 Tracer gas arrival times at different locations during the second tracer gas test Plotting the arrival times of the peaks on the same plot shows the earlier arrival of the peak of tracer gas for the second test at all of the monitoring points in particular the maingate seal sites show tracer gas arriving almost instantaneously . The shape of the peak for the tailgate monitoring point shown in Figure 4.15, also suggests that the path taken by the gas to reach the tailgate was altered significantly from that of the first test. The fact that the tracer gas readings rose steeply to there maximum in the second test indicates that the majority of the gas arrived at the monitoring point within a short space of time suggesting that less air paths through the deep goaf (driven by differential pressures across the goaf) were available for the inert gas travel along as opposed to the first test which showed a gradual rise to its peak reading indicating numerous air paths through the deep goaf due to the ingress of air.


BOI.036.001.0082

66

Peak arrival time Comparison at Tailgate

0.00

0.50

1.00

1.50

2.00

2.50

2 22 37 61 87 139

176

229

294

475

1099

1339

1570

Elapsed time in Minutes

Trac

er G

as in

ppm

Tracer 1

Tracer 2

Figure 4.15 Comparison of arrival times and peak shapes at the tailgate monitoring point The results obtained from the maingate side monitoring points close to the injection point must be used to measure the peak arrival time only as a larger amount of tracer gas was injected in the second test distorting the magnitude of the results for comparison. Figure 4.16 above shows that tracer gas was detected at 6c/t during both trials, the peak in the second trial arriving earlier than the first. The results for 7c/t (Figure 4.17) once again show the peak arrival time for the second trial to be faster than the first by some 800 minutes (13 hours). Figure 4.18 shows the comparison of peak arrival times at 8c/t, a large difference of some 27 hours is observed between the peaks residual tracer gas from the first trial can still be seen at the commencement of the second test.

0

50

100

150

200

250

300

350

400

1 20 40 62 81 111

141

171

201

231

295

425

804

Elapsed Time in Minutes

Rea

ding

in p

pm

Tracer 1

Tracer 2

Figure 4.16 Comparison of tracer gas arrival times at 6c/t


BOI.036.001.0083

67

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 25 45 65 81 109

139

169

200

229

259

373

805

1210

1450


Read

ing

in p

pm

Tracer 1 <0.01Tracer 2 <0.01


0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

1 24 41 61 89 139

169

200

229

259

372

803

1210

1440

1700


Rea

ding

in p

pm

Tracer 1Tracer 2

Trial 2 gas peak

Residual tracer gas

Figure 4.18 Comparison of tracer gas arrival times at 8c/t The results for 10c/t (figure 4.19) show that during the time period for the first trial that no tracer gas was detected at the 10c/t monitoring site. However with the introduction of high expansion foam tracer gas was detected within ten minutes of monitoring, and peaking within 1.5 hours.


BOI.036.001.0084

68

0.00

0.00

0.00

0.01

0.01

0.01

0.01

0.01

0.02

0.02

9 28 51 71 93 117 152 182 209 243 261


Rea

ding

in p

pm

Tracer 2


0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0 48 70 92 115 146 180 188 223

El a pse d Ti me i n M i nut e s

Tracer 2

Figure 4.20 Comparison of tracer gas arrival times at 18c/t Similarly at 18c/t shown in Figure 4.20 no tracer gas was detected during the sampling period allowed for the first test. With the introduction of the high expansion foam plug in the second test we see that the second tracer gas test has been detected at the monitoring site, peaking within 2 hours. This phenomenon of changing the deep goaf gas flow regime by introducing a foam plug was also noted during the Mine C foam trial where the goaf gas composition changed from being stable and inert to a condition where CO2 was reduced and O2 increased (both by approximately 2%) during the trial as illustrated in Figure 4.5.


BOI.036.001.0085

69

4.5 CFD MODELLING OF FOAM PLUGS – PRELIMINARY STUDIES Preliminary CFD modelling was performed to investigate the performance of the inert gas and foam injection trials carried out at the Mine D. The models were developed and calibrated using tube bundle gas monitoring data from the working panel to represent goaf gas composition in the panel accurately, mine survey data was used to adjust for the new panel geometries. This new base model was used to simulate various options of inert gas and high expansion foam injection combinations. The base model results presented in Figure 4.21 shows the goaf gas distribution for the working panel with the air fringe clearly present on both maingate and tailgate sides of the panel due to gas buoyancy and the dip from maingate to tailgate. A blue colour represents low oxygen and red colour represents high oxygen at around 21% or fresh air.

TG

MG

Figure 4.21 Oxygen distribution in the longwall panel goaf at Mine D

Inert gas injection point

Figure 4.22 Effect of inert gas injection immediately behind the working face

Figure 4.22 shows the oxygen distribution in the goaf with inert gas being injected into the goaf from the maingate side at approximately 100m behind the working face. It can be seen that high oxygen levels are still present around the injection site due to the overwhelming effects of scouring of the area immediately behind the face due to high ventilation quantities on the longwall face.


BOI.036.001.0086

70

Figure 4.23 shows the final combination of a high expansion foam plug in place immediately behind the retreating longwall face coupled with inert gas being injected into the goaf at the next cut through inbye. It is clearly observed that this combination reduces the extent of the oxygen ingress into the goaf by reducing the amount of oxygen ingress at the maingate and in turn reducing the dilution effects on the inert gas.

Foam plug location Inert gas injection point

Figure 4.23 Effect of a foam plug in combination with inert gas on goaf O2 distribution 4.6 SUMMARY AND CONCLUSIONS Preliminary investigations were also conducted during the course of this inertisation project to study the effect of additional measures such as foam injection, foam and inert gas combination on oxygen ingress patterns into the goaf. The initial foam trials indicated that the introduction of high expansion foam as a plug into the perimeter of a longwall goaf effects the goaf gas flow patterns potentially aiding in the self inertisation process of the goaf. The second set of trials with foam and inert gas combination indicated that the effectiveness of the inert gas has increased due to the decrease in dilution and alteration of the goaf gas flow patterns. However, the drawback of using high expansion foams in these circumstances is the durability or stability of the foam when exposed to airflow, breaking down the foam’s bubble structure. These preliminary results indicate that further investigations to increase the foam stability and detailed studies to investigate the effects of foam injection into the goaf under mining and operational different conditions are highly warranted.

- - - - -


BOI.036.001.0087

71

CHAPTER 5

FIELD DEMONSTRATION STUDIES

5.1 INTRODUCTION An integrated part of this research project was the field demonstration studies of the optimum goaf inertisation strategies identified by the CFD simulations presented in previous chapter. Two Australian underground coal mines, Mine A and Mine B, participated in this project. Both mines experienced high oxygen ingress problems and presented as ideal sites for field investigation of proactive inertisation strategies. These two mine sites are located at two different coal regions in Australia, one in New South Wales and other in Queensland. Optimum inertisation strategies for active longwall panels have been developed based on the analysis of the modelling investigation results and review of the oxygen ingress patterns in the field at the mine sites. These proactive inertisation strategies have then been implemented at the mine sites to investigate their performance in the field, particularly with respect to their effect on oxygen ingress patterns in the goaf. This chapter presents a summary of the field demonstration studies in these two mine sites. 5.2 FIELD DEMONSTRATION STUDIES AND RESULTS – MINE A As introduced in previous chapters, this mine is operating longwall panels in the new mining area in a new seam prone to spontaneous combustion. As the first panel progressed some 450m from the start up line, signs of spontaneous heating were detected and pro-active goaf inertisation was needed to maintain normal production and safeguard the longwall face. Figure 5.1 shows the longwall panel layout from start-up line at the earlier stage of the block.

TG MG

Figure 5.1 Longwall panel layout at Mine A


BOI.036.001.0088

72

The longwall panel has access around the perimeter of the panel which enabled extensive monitoring of gas distribution in the goaf on both maingate (MG) and tailgate (TG) sides. Seven tube-bundle monitoring points were installed in the goaf to continuously monitor goaf gas distribution at varying distances behind the longwall face. Prior to inertisation, the goaf gas distribution in the start-up area of the panel showed that oxygen ingress into the goaf was very high with oxygen levels over 17% at 400 m behind the face, as shown in Figure 5.2. The dotted blue line shows the oxygen ingress zones within the goaf according to the tube bundle monitoring data, which are also shown in Figure 5.2.

17.2 2.3 4.4 73.1

TG

MG

Tb 8

Tb 7 Tb 14

17.9 2.9 1.6 76.2

3.2 16.7 24.5 49.5

6.3 8.8 21.9 54.8

Tube bundle points

20.1 0.8 0.2 77.2

2.2 22.5 29.1 41.2

14.4 8.2 6.4 67.8

Gas (%) O2

CO2 CH4 N2

Face


Goaf Dip

Figure 5.2 Goaf gas distribution behind the face before inert gas injection

13.9 3.6 6.9 70.5

TG

MG

Tb 8

Tb 7

14.5 4.5 4.5 76.1

2.5 16.6 22.2 50.0

5.2 13.1

24.5 56.5

1.5 27.4 30.5 36.2

7.3 12.5 14.0 61.4

1.4 10.1 64.6 24.4

20.3 0.1 0.1 76.5

Inert gas injection

Goaf

Gas (%) O2

CO2 CH4 N2

Oxygen ingress into the goaf (just after start of proactive inertisation)

Figure 5.3 Goaf gas distribution behind the face just after start of inert gas injection By referring to the results of CFD simulation studies and specific site conditions, the proactive inertisation strategy for this panel consisted of injecting inert gas into the goaf through the first cut-through near the start-up area. Inert gas was delivered to the cut-through via a 150 mm pipe connected to the bottom of a surface concrete drop hole available in the perimeter roadway. Inert gas generation equipment was installed on the


BOI.036.001.0089

73

surface and connected to the top of the concrete drop hole through 150 mm steel and flexible pipes combination. Figure 5.3 shows the goaf gas distribution patterns just after start of the injection of inert gas deep into the goaf on MG side of the panel. Figure 5.4 shows the goaf gas distribution patterns in the goaf after 1 day inert gas injection via the 1st cut-through (23ct) close to the start-up line. The goaf gas monitoring results indicate that the oxygen ingress fringe has been pushed back towards the face and therefore narrowed down the sponcom liable zones in the goaf.

Figure 5.4 Goaf gas distribution behind the face - 1 day after inert gas injection Figure 5.5 shows the goaf gas distribution pattern after 3 days inert gas injection. It can be seen clearly that the injection of inert gas has pushed the high oxygen zone was further back towards the face, resulting in most of the goaf area being inert.

4.1 6.6 13.8 74.4

TG

MG

Tb 8

5.7 7.5

14.9 69.0

3.010.57.1

82.5

5.7 4.3

3.1 95.1

3.1 9.9 7.0

85.3

7.0 12.6 14.7 61.9

2.9 3.6 8.8

88.4

17.5 3.0 2.1 78.0

Inert gas injection Oxygen ingress into the goaf

(with proactive inertisation)

Goaf

3.8 5.5 8.0

91.4

4.4 0.7 0.7 94.1

13.4 4.5 4.2 80.2 Tube bundle points


TG

MG

Tb 8

Tb 7

2.812.49.9

76.5

4.8 8.1

8.9 82.8

2.3 17.0 17.9 58.6

6.8 13.7 15.7 59.2

Inert gas injection

Goaf

Gas (%) O2

CO2 CH4 N2

Gas (%) O2

CO2 CH4 N2

Figure 5.5 Goaf gas distribution behind the face - 3 days after inert gas injection


BOI.036.001.0090

74

Figure 5.6 presents the goaf gas distribution pattern after 10 days inert gas injection. The gas monitoring results show that the oxygen level was below 6% at the second cut-through (19ct) behind the face, i.e. 200 m behind the face. The injection of inert gas injection has also resulted in reduced oxygen ingress on the tailgate side of the panel. High oxygen ingress was restricted to only one cut-through behind the face on the maingate side.

0.6 6.7 17.4 75.3

2.0 3.5 6.5 87.9

TG

MG

Tb 8

3.5 0.4 0.5 95.6

3.0 21.7 25.6 49.5

5.3 9.7 14.1 70.8

5.2 2.9 6.185.6

4.0 25.0 25.3 45.6

5.0 15.6 18.5 60.8

20.6 0.4 0.178.8

Inert gas injection

Goaf

Gas (%) O2

CO2 CH4 N2


Figure 5.6 Goaf gas distribution behind the face - 10 days after inert gas injection Figure 5.7 shows the goaf gas distribution pattern in the longwall goaf after 40 days inert gas injection. The gas monitoring results indicate that the continued injection of inert gas has successfully kept the oxygen ingress level down at most part of the goaf, e.g., at some 9% at the second cut-through (18ct) behind the face, i.e. 200 m behind the face on the MG side. The injection of inert gas injection has also helped restrict the oxygen penetration on the tailgate side of the panel. The goaf inertisation has also helped the panel to resume normal production within a short period following the signs of spontaneous combustion.

0.4 13.6 20.5

65.5

TG

MG

1.2 10.1 11.5 77.1

0.5 18.114.067.4

0.2 20.4

17.0 62.3

5.8 5.1 1.987.0

0.6 20.1 15.3 64.0

2.4 12.78.1

76.7

0.4 16.3 22.8 60.4

9.3 0.8 0.289.6

Inert gas injection

Oxygen ingress into the goaf Goaf

20.4 0.5 0.178.8

4.2 11.0 6.5

78.1

Gas (%) O2

CO2 CH4 N2

Figure 5.7 Goaf gas distribution behind the face - 40 days after inert gas injection


BOI.036.001.0091

75

The proactive inertisation was conducted throughout the life of this longwall panel to prevent the onset of spontaneous combustion in the goaf. Figure 5.8 presents the goaf gas distribution and inertisation scenario when the face was at around 6 c/t, i.e. after the face has retreated by more than 1,500 m in the panel. Inert gas was pumped into the goaf via the 8ct, at some 200m behind the face on the MG side, and through the goaf hole at 400 m behind the face on the TG side. Both points are within the optimum injection range identified by the CFD modelling studies presented in previous chapter. The gas distribution in the goaf after 10 days of inert gas injection at the two locations indicted by the red arrows is shown in Figure 5.8. The gas monitoring results show that the oxygen levels were below 5 - 6% at 200 - 300 m behind the face on both sides of the goaf. The injection of inert gas at these two locations has greatly narrowed the high oxygen zones and thus helped suppress the onset of spontaneous combustion in the goaf.

TG

MG

20.75

0.21 0.870.1518.07 18.80

0.82 1.6813.7315.8313.96 14.09

0.10 78.93

19.02

0.9979.12

3.60

0.2396.02

0.61

10.02 61.24

0.28

19.65 61.25

20.33

0.51 78.96

17.64

1.4479.24

2.40

11.6972.17

0.61

12.7270.83

2.38

9.83 73.82

1.61

10.92 73.38

Goaf

Gas (%) O2

CH4 N2

Figure 5.8 Goaf gas distribution pattern in the goaf with proactive inertisation

Figures 5.9 Goaf gas composition changes at 7CT on the TG side

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

18.00

20.00

1-Apr-

05 0:

00

3-Apr-

05 0:

00

5-Apr-

05 0:

00

7-Apr-

05 0:

00

9-Apr-

05 0:

00

11-A

pr-05

0:00

13-A

pr-05

0:00

Oxy

gen,

Met

hane

and

Car

bon

Dio

xide

(%

)

0.0000

0.0100

0.0200

0.0300

0.0400

0.0500

0.0600

0.0700

Caa

rbon

Mon

oxid

e (%

)

OXYGEN

METHANE

CARBON DIOXIDE

CARBON MONOXIDE

CO2 Inert gas injection

Inert gas injection


2.21

19.1864.72

13.88


BOI.036.001.0092

76

To further demonstrate the results of proactive inertisation operations in this panel, Figures 5.9 to 5.17 show a summary of the goaf gas composition changes at different cut-through since the start of inertisation. The results indicate that during the implementation of the proactive strategies at various locations in the panel, oxygen ingress into the oxygen has been successfully restricted to a narrow zone immediately behind the face, with most part of the goaf having oxygen level below 5% to 6%.

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

18.00

0.0000

0.0100

0.0200

0.0300

0.0400

0.0500

0.0600

0.0700

Car

bon

Mon

oxid

e (%

)

OXYGEN

M ETHA NE

CA RB ON DIOXIDE

CA RB ON M ONOXIDE

%)

e (

n D

ioxi

drb

o C

ane

&et

hayg

en, M

Ox


0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

1-Au

g-04

20-S

ep-0

4

9-N

ov-0

4

29-D

ec-0

4

17-F

eb-0

5

8-Ap

r-05

28-M

ay-0

5

17-J

ul-0

5

0.0000

0.0100

0.0200

0.0300

0.0400

0.0500

0.0600

0.0700

0.0800

0.0900

0.1000C

AR

BO

N M

ON

OXI

DE

(%)

OXYGEN

METHANE

CARBON DIOXIDE

CARBON MONOXIDE

%)

XI

DE

(


OXY

GEN

, MET

HA

NE,

AN

D C

AR

BO

N D

IO


BOI.036.001.0093

77

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

23-A

pr-0

4

12-J

un-0

4

1-Au

g-04

20-S

ep-0

4

9-N

ov-0

4

29-D

ec-0

4

17-F

eb-0

5

8-Ap

r-05

28-M

ay-0

5

17-J

ul-0

5

OXY

GEN

, MET

HA

NE,

AN

D C

AR

BO

N D

IOXI

DE

(%)

0.00000

0.01000

0.02000

0.03000

0.04000

0.05000

0.06000

0.07000

CA

RB

ON

MO

NO

XID

E (%

)

OXYGEN

METHANE

CARBON DIOXIDE

CARBON MONOXIDE


0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

45.00

1-Apr-0

5 0:00

11-A

pr-05

0:00

21-A

pr-05 0

:00

1-May

-05 0:

00

11-M

ay-05 0:

00

21-M

ay-05 0

:00

31-M

ay-05 0

:00

10-Ju

n-05 0

:00

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

Car

bon

Mon

oxid

e (%

)OXYGEN

METHANE

CARBON DIOXIDE

CARBON MONOXIDE

Figures 5.13 Goaf gas composition changes at 11CT on the MG side

Oxy

gen.

Met

hane

& C

arbo

n D

ioxi

de (%

)


BOI.036.001.0094

78

0.00

5.00

10.00

15.00

20.00

25.00

1-Apr-

05 0:

00

11-A

pr-05

0:00

21-A

pr-05

0:00

1-May

-05 0:

00

11-M

ay-05 0

:00

21-M

ay-05 0

:00

31-M

ay-05 0

:00

10-Ju

n-05 0

:00

0.0000

0.0100

0.0200

0.0300

0.0400

0.0500

Car

bon

Mon

oxid

e (%

)

OXYGEN

METHANE

CARBON DIOXIDECARBON MONOXIDE

)e

(%xi

d D

ioC

arbo

ne

&

han

, Met

enO

xyg


0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

1-Au

g-04

20-S

ep-0

4

9-N

ov-0

4

29-D

ec-0

4

17-F

eb-0

5

8-Ap

r-05

28-M

ay-0

5

17-J

ul-0

5

5-Se

p-05

E

XID

DIO

N

A

RB

O

ND

C

AN

E, A


OXY

GEN

, MET

H(%

)

0.000

0.010

0.020

0.030

0.040

0.050

0.060

OXYGEN

METHANE

CARBON DIOXIDE

CARBON MONOXIDE

C

AR

BO

N M

ON

OXI

DE

(%)


BOI.036.001.0095

79

0.00

5.00

10.00

15.00

20.00

25.00

30.00

12-J

un-0

4

1-Au

g-04

20-S

ep-0

4

9-N

ov-0

4

29-D

ec-0

4

17-F

eb-0

5

8-Ap

r-05

28-M

ay-0

5

17-J

ul-0

5

5-Se

p-05

OXY

GEN

, MET

HA

NE,

AN

D C

AR

BO

N D

IOXI

DE

(%)

0.0000

0.0100

0.0200

0.0300

0.0400

0.0500

0.0600

CA

RB

ON

MO

NO

XID

E (%

)

OXYGEN

METHANE

CARBON DIOXIDE

CARBON MONOXIDE


0.00

20.00

40.00

60.00

80.00

100.00

120.00

0.0000

0.0050

0.0100

0.0150

0.0200

0.0250

0.0300

0.0350

0.0400

0.0450

OXYGEN

METHANECARBON DIOXIDE

CARBON MONOXIDE

23-A

pr-0

4

12-J

un-0

4

1-Au

g-04

20-S

ep-0

4

9-N

ov-0

4

29-D

ec-0

4

17-F

eb-0

5

8-Ap

r-05

28-M

ay-0

5

17-J

ul-0

5

CA

RB

ON

MO

NO

XID

E (%

)

)XI

DE

(% D

IOB

ON

CA

RN

DE,

AH

AN

Figures 5.17 Goaf gas composition changes at 23CT on the MG side The field demonstration studies carried out at Mine A site were highly successful in reducing oxygen ingress into the goaf and in achieving effective goaf inertisation. The field results showed that oxygen levels were below 5% to 6% at 200 - 300 m behind the face during the study periods after implementation of the proactive inertisation. These

OXY

GEN

, MET


BOI.036.001.0096

80

field studies demonstrated that it is possible to reduce the oxygen ingress distance in the goaf to within 200 to 300 m behind the face. The project studies also greatly improved the fundamental understanding of the various parameters and strategies on goaf inertisation. The proactive inertisation operations played a key role in containing the small heating in the goaf at the start-up area of the longwall panel. However, it was noted that one standard inertisation strategy was not sufficient for the entire length of the longwall panel and resulted in onset of another small heating near the geologically disturbed area of the panel. It was realised that proactive inertisation strategies need to be modified with major changes in mining conditions such as ventilation system, geological structures, caving conditions, longwall retreat rates and goaf gas drainage.

5.3 FIELD DEMONSTRATION STUDIES AND RESULTS – MINE B This section presents the details and results of field demonstration studies of proactive inertisation carried out at another Mine B. The longwall face in panel 105 at Mine B has advanced for about 550 m from the panel start-up line during the field trials. Due to roadway roof control problems, the tailgate (TG) was collapsed and as a result the longwall retreat has to be halted for the installation of a new tailgate. As the face was stoped, there were signs of spontaneous heating (high CO and H2) developing in the goaf, and consequently pro-active goaf inertisation was considered to reduce oxygen ingress into the goaf and to prevent the heating development. Figure 5.18 shows the longwall panel layout at this stage in Mine B. The TG side of the panel was adjacent to an old longwall goaf and MG side in virgin coal block.

Figure 5.18 Longwall panel layout at Mine B during field demonstration studies The longwall panel has access only on the maingate side of the panel, as is the case in most of the longwall panels. Five tube-bundle monitoring points were installed on the intake maingate side of the goaf to continuously monitor oxygen ingress into the goaf at varying distances behind the longwall face.


BOI.036.001.0097

81

A snapshot of the goaf gas distribution behind the longwall face before the collapse of the tailgate is given in Figure 5.19. The gas monitoring results show that intake air ingress on the maingate side of the panel was very high with the oxygen level at more than 17% even at the 5th cut-through behind the longwall face, i.e. 370 m behind the longwall face. The oxygen ingress distance on the tailgate side of the panel was based on the gas readings obtained from an experimental gas drainage borehole. This level of high oxygen ingress in low gas environments could lead to heatings if the coal seam has moderate to high propensity to sponcom. The risk of heatings increases significantly if the face stops for a prolonged period. It is also to be noted here that prolonged stoppage of face at this mine led to increased oxidation and low level heating in the goaf. Figure 5.20 shows the goaf gas distribution in behind the longwall panel after the collapsed tailgate was closed and the face was ventilated with auxiliary fans at 10m3/s. Again the airflow penetration into the goaf was high on the maingate side with oxygen level over 10% at some 300 m behind the longwall face.

Previous panel goaf

TG

MG 16.5 2.3 0.4 80.7

20.2 1.1 0.1 78.6

Gas (%) O2

CH4 CO2 N2

Longwall start-up

20.7 0.7 0.1 78.5

19.9 1.1 0.1 78.7

17.8 1.6 0.4 80.1

Tube bundle point

Goaf

Oxygen ingress into the LW goaf

Dip

Figure 5.19 Goaf gas distribution behind the longwall face - before the closure of the tailgate

TG

MG 11.5

4.4 0.8

83.2

18.8 2.8 0.2

78.2

Dip Gas (%) O2

CH4 CO2 N2

Longwall start-up

8.8 4.6

1.5 85.0

17.4 2.4 0.4

79.8

10.43.7

1.0 84.8

10m3/s TG

Goaf

Previous panel goaf


Figure 5.20 Goaf gas distribution behind the longwall – after tailgate closure


BOI.036.001.0098

82

Proactive inertisation at Mine B was carried out through the existing concrete drop holes and additional goaf holes specifically drilled for this goaf inertisation. In addition, both Tomlinson boiler and Floxal nitrogen units were used for the inertisation operations. The proactive inertisation strategy for this panel consisted of injecting inert gas into the goaf through the surface boreholes located on the maingate and tailgate sides of the panel at a combined flow rate of 0.6m3/s. A combination of inertisation strategies involving inert gas injection either via a single hole or a combination of holes on both sides of the goaf simultaneously were adopted for inertisation at this mine site. These combinations have been identified in the CFD modelling studies as the most optimum options for this case. Figures 5.21 to 5.23 shows the results of goaf gas distribution in the longwall goaf after injection of boiler gas via the boreholes on the tailgate and maingate sides of the panel. The goaf gas monitoring data indicate that the goaf inertisation operation has worked well in narrowing down the size of the high oxygen ingress zones in the longwall goaf.


0.9 16.9 7.2 75.0

2.9 15.6 6.8 74.7

0.8 19.9 8.5 70.6

TG Closed

2.6 15.7 5.2 76.4

7.3 15.3 3.3 74.1

Gas (%) O2

CH4 CO2 N2

2.4 15.0 5.1 77.4

2.0 17.2 5.4 75.3

Previous panel goaf

Longwall Goaf

Inert gas injection boreholes Tube bundle point

Inert gas injection

Dip

Figure 5.21 Goaf gas distribution behind the longwall – 1 day after start of inertisation


2.5 21.5 7.1 68.8

2.0 0.0

9.7 88.2

1.1 24.1 6.9 67.9

10.6 12.1 3.7

73.5

Gas (%) O2

CH4 CO2 N2

1.5 22.1 6.9 69.5

2.1 14.2 8.1 75.6

Dip

Longwall Goaf

Inert gas injection

TG Closed

Previous panel goaf

Figure 5.22 Goaf gas distribution behind the longwall – 2 weeks after start of inertisation


BOI.036.001.0099

83

1.5 17.7 8.8 71.9

2.9 13.2 9.9 73.9

1.9 15.3 8.2 74.6

2.7 15.6 7.2

74.4

Gas (%) O2

CH4 CO2 N2

1.4 15.2 8.1 75.2

0.920.6 8.270.2

Previous panel goaf

Longwall Goaf

Inert gas injection

TG Closed


Figure 5.23 Goaf gas distribution behind the longwall – 5 weeks after start of inertisation To reduce oxygen ingress into the adjacent sealed panel, some of the boiler gas was also injected into the adjacent sealed goaf through another concrete drop hole available in that panel. To cater for these increased inert gas requirements another inert gas generation unit ‘Floxal’ was procured and installed at the mine. Floxal unit generates nitrogen gas at flow rates of about 150 l/s. Floxal nitrogen generation unit was used to inject nitrogen gas via the surface borehole close to the start-up line on the maingate side. Figure 5.24 shows the goaf gas distribution pattern following the injection of inert gas via the two surface goaf holes as indicated by the red arrows. The results indicate that this proactive inertisation strategy was highly successful and had substantially reduced oxygen ingress into the longwall goaf, and as such has helped the containment of spontaneous combustion during the stoppage of the longwall face while a new tailgate was being re-installed.

3.6 4.9

4.2 87.2

3.6 7.3

4.0 85.1

MG 2.5

8.4 5.883.2

4.3 11.3 6.7

77.7

Gas (%) O2

CH4 CO2 N2

2.7 10.0 6.8 80.4

3.46.6

4.985.1

Boiler gas injection


Previous panel goaf

Longwall Goaf N2 gas injection

TG Closed


Figure 5.24 Goaf gas distribution pattern – 7 weeks after start of inertisation The goaf inertisation operation was continued after the re-installation of the new tailgate in the panel. Figure 5.25 and 5.26 respectively provides a snapshot of the goaf gas distribution patterns following the restoration of longwall face ventilation system at 30m3/s to assist in reducing oxygen ingress into the goaf. The gas monitoring results of the


BOI.036.001.0100

84

field studies show that the proactive inertisation strategy continued to maintain a highly successful goaf inertisation environment behind the face, with the oxygen level dropping below 3% at the second cut-through behind the face, i.e. 100 m behind the face. This very high reduction in oxygen ingress was due to a combination of inert gas injection and ventilation strategies including reduction in intake airflow.

4.2 0.7

2.2 92.8

0.64.9

4.290.2

19.6 0.7 0.4

79.2

Gas (%) O2

CH4 CO2 N2

1.4 5.8

4.388.4

0.73.5

3.891.9



Previous panel goaf

Longwall Goaf

N2 gas injection TG Q=30m3/s MG


1.1 1.6

2.8 94.4

Figure 5.25 Goaf gas distribution pattern following restoration of face ventilation and continued

inert gas injectioin – 13 weeks after start of inertisation

0.73.1

3.892.4

17.5 1.3 2.2

78.9

Gas (%) O2

CH4 CO2 N2

1.2 4.1

5.189.5

1.11.6

2.894.4


Previous panel goaf TG Q=30m3/s MG


Longwall Goaf

Inert gas injection boreholes

N2 gas injection

1.1 1.6

2.8 94.4

Tube bundle point

Figure 5.26 Goaf gas distribution pattern following face ventilation restoration and continued inert gas injectioin – 17 weeks after start of inertisation

In summary, the field demonstration studies of proactive inertisation at Mine B also have proved to be a very successful operation. Inert gas, generated by Tomlinson Boiler and the Floxal Nitrogen unit, has been injected into the goaf on both sides of the longwall goaf via the surface goaf holes. This operation was carried out with an effective management of the ventilation system and the injection of inert gas via the three goaf holes. The goaf gas


BOI.036.001.0101

85

monitoring results show that the high oxygen ingress zones have been narrowed down significantly towards the face, with the oxygen level reducing below 3% at some100 m behind the face, which has proved to be effective in suppressing the onset of spontaneous combustion/heating in the goaf. The proactive inertisation strategies developed with the combination of CFD modelling studies and field work have played a crucial part in containing the onset of spontaneous combustion in the longwall goaf during the stoppage period of the face before the reinstallation of a new tailgate and the resumption of normal production.

5.4 PROACTIVE INERTISATION STRATEGIES Based on the results of CFD modelling and field studies conducted during the course of this project, the recommended guidelines for proactive inertisation strategies are:

• inert gas should be injected into the goaf at 200 to 400 m behind the face, or inbye side of a suspected heating location in the goaf.

• inert gas flow rate of around 0.5 m3/s is recommended for most cases to reduce oxygen ingress into the goaf (higher inert gas flow rates are required in the case of advanced heatings in the goafs).

• inert gas should be injected on intake side of the goaf in most cases to reduce oxygen ingress into the goaf. (different strategy may be required in case of heatings/ suspected heatings depending on their location and intensity).

• inert gas should be injected through goaf holes (preferable option, if goaf holes are available) - or - through cut-through seals on the intake side of the goaf.

• inert gas may need to be injected on both sides of the goaf if some heating is suspected on return side of the goaf or along major fault areas.

• onsite inert gas generation units are recommended for proactive inertisation (other inert gas supplies may also be required in case of advanced heatings).

• inert gas injection to be continued until face resumes normal production in case of prolonged stoppages or until face has retreated for more than 300 to 500 m past the suspected heating location, in case of heatings.

• ventilation system in and around the panel also should be designed to minimise oxygen ingress into the goaf for effective inertisation.

A good understanding of the goaf gas flow patterns, the selection of the appropriate injection location points and inert gas flow rates, as well as a good management of longwall face ventilation and pressure drop across the goaf and minimisation of goaf leakage are all critical to the success of proactive inertisation in the longwall goafs. It is to be noted that most of the above strategies are suitable for proactive inertisation to reduce oxygen ingress into the goaf. The specific inertisation strategies for control of any heatings/suspected heatings may differ from the above generic strategies and depends on the heating intensity, size, location, ventilation system and other mine site conditions.


BOI.036.001.0102

86

5.5 SUMMARY AND CONCLUSIONS Proactive inertisation demonstration studies have been carried out in working longwall panels at two different mines. Both the mines experienced deep air ingress into the goaf with oxygen concentration levels over 17% at 300 m to 400 m behind the face. This level of high oxygen ingress has contributed to the initiation of spontaneous heating in the goaf, particularly during the face stoppage/ very slow retreat periods. The field demonstration studies carried out at Mine A site were highly successful in reducing oxygen ingress into the goaf and in achieving effective goaf inertisation. Inert gas was injected initially via the cut-through close to the face start-up line on the maingate slide, and later on both sides of the goaf via cut-throughs and surface goaf hole(s). The field gas monitoring results showed that oxygen levels were dropped below 5% to 6% at 200 to 300 m behind the face after implementation of the proactive inertisation. Field demonstration studies of proactive inertisation at Mine B have also proved to be a success. Because of underground access constraints, the inert gas has to be injected into the goaf via the surface goaf holes on both side of the longwall panel. The monitoring results show that the high oxygen ingress zones have been reduced significantly, with the oxygen level below 3% at100 m behind the face, which has proved to be effective in controlling the development of heating in the goaf while the face was in stoppage. Field studies results at both the mine sites demonstrated that the proactive inertisation strategies were highly successful in converting the general goaf environment into an inert atmosphere. These field studies also demonstrated that it is possible to reduce the oxygen ingress distance in the goaf to within 200 to 300 m behind the face by implementing appropriate proactive inertisation strategies. The proactive inertisation strategies developed with the combination of CFD modelling studies and field work have played a crucial part in containing the onset of spontaneous combustion and in reducing the risk of larger heatings development in both the longwall goafs during the study periods. It was also noted that one standard inertisation strategy was not sufficient for prevention of heating incidents in the entire length of the panels or all longwall panels. It was realised that proactive inertisation strategies need to be modified with major changes in mining conditions such as ventilation system, geological structures, caving conditions, longwall retreat rates and goaf gas drainage. The project studies also greatly improved the fundamental understanding of the various parameters and strategies on goaf inertisation. The fundamental understanding of inert gas flow patterns in active goafs and proactive inertisation strategies developed during the course of the project can greatly reduce the risk of heatings in longwall panels and enhance the safety of coal mines.

- - - - -


BOI.036.001.0103

87

CHAPTER 6

CONCLUSIONS

The principle objective of this ACARP research project was to develop and demonstrate effective proactive inertisation strategies to reduce the risk of spontaneous heatings in active longwalls that are characterised by high face ventilation quantities in the range of 40 to 80 m3/s and thick seam conditions in Australia. The project was carried out with an integrated approach combining detailed analysis of oxygen ingress patterns in the goafs, extensive Computational Fluid Dynamics (CFD) modelling studies of various proactive inertisation options, and field trials of these strategies at two underground longwalls. This chapter provides a summary of the main conclusions and recommendations from this research project, and some areas for future research to improve the current pro-active inertisation practices. 6.1 CONCLUSIONS AND RECOMMENDATIONS (a) Oxygen ingress and heating issues in longwall goafs

Goaf gas monitoring data from different longwall panels showed that air ingress into the goaf was very high in a number of cases, with oxygen concentration levels over 17% at 300 m to 400 m behind the face. This deep penetration of airflow provides an ideal catalyst for sponcom/heatings development.

•

•

•

•

•

•

The combination of the high oxygen ingress into the goafs along with prolonged panel stoppage and geological disturbances contributes to the increased oxidation and low level heating in the longwall goafs.

Analysis of a number of heating incidents in longwall panels indicated that a majority of heating incidents develop on the high oxygen ingress side of the goaf, except in special circumstances.

The heating incidences in longwall goafs continues to be a threat to production and safety due to the coal’s inherent propensity to sponcom and the exposure of the large quantities of broken coals to high level oxygen in the goafs behind longwall faces.

Oxygen ingress depends on a number of factors, including intake airflow, ventilation layout, pressure differentials, seals condition and leakages, caving characteristics, gateroad support systems, goaf gas emissions, panel geometry and seam gradients. Therefore, it is important to characterise the oxygen patterns in initial longwall goafs of any new mines before developing appropriate proactive inertisation strategies.

(b) CFD modelling of inertisation options

CFD models were developed for a range of longwall goaf geometries and gas emission conditions based upon several Australian underground longwall panel layouts. The results of these base case models tallied well with field measured goaf gas concentration values.


BOI.036.001.0104

88

CFD modelling results indicated that there are only minor differences between the boiler gas and nitrogen with respect to their effect on goaf inertisation in the simulated cases. However, there could be significant differences between the effects of different inert gases in some cases depending on the specific site conditions.

•

•

•

•

•

•

•

•

•

•

•

•

Longwall panel geometry, goaf gas emission rates and composition, ventilation layouts, airflow rates and pressure differentials, goaf characteristics, and gateroad conditions in the goaf would also have a significant influence on goaf inertisation.

Inertisation simulations indicated that inert gas injection close to the face, i.e. within 50 m behind the face, would not be effective even at higher inert gas flow rates of 1.0 m3/s in active longwall panels with face air quantities over 50 m3/s.

Inert gas injection at inbye locations in the goaf, i.e. between 200m to 400m behind the face, resulted in more effective goaf inertisation in the simulated cases.

Inert gas flow rate of around 0.5m3/s would be required for goaf inertisation in most cases, although in some cases inert gas injection at low flow rates of 0.15 m3/s would also result in effective goaf inertisation and in some other cases inert gas flow rates up to 1.0 m3/s may be required for effective inertisation of the goaf.

Simulation results indicated that inert gas needs to be injected on intake (maingate) side of the goaf, i.e. on high oxygen ingress side, for effective goaf inertisation. Inert gas injection just on return (tailgate) side of the goaf does not reduce oxygen ingress on the intake side of the goaf in most cases.

In case of suspected heatings on return side, inert gas may need to be injected on return side or both sides of the goaf depending on the extent and intensity of heating, specific site conditions and the available capacity of inert gas generating systems.

CFD simulations indicated that inert gas injection though surface goafholes located at 20 to 30 m from gateroads into the goaf would be more effective when compared with inert gas injection through cut-throughs/seals in the gateroads. Inert gas injection through a combination of surface goafholes and cut-throughs can further improve the effectiveness of goaf inertisation in some cases.

Results also indicated that in addition to the optimum inert gas injection points and flow rates, it is critical to have a tight control over ventilation system in and around the goaf for the success of proactive inertisation at low inert gas flow rates.

(c) Preliminary studies on other options to reduce oxygen ingress in longwall goafs

Preliminary trials on foam injection indicated that the introduction of high expansion foam as a plug into the perimeter of a longwall goafs significantly effects the goaf gas flow patterns, potentially aiding in the goaf self inertisation process.

Another set of trials with foam and inert gas combination indicated that the effectiveness of the inert gas has increased due to the decrease in air dilution and alteration of the goaf gas flow patterns.

The stability of high expansion foam used in the trials was very low with foam’s bubble structure breaking down within 24 hours of injection. The preliminary studies confirmed the need for further detailed investigations on foam durability/stability and its effectiveness under different mining conditions.


BOI.036.001.0105

89

(d) Field demonstration studies – Proactive inertisation

Field demonstration studies showed that the implementation of proactive inertisation strategies at Mine A has successfully reduced the oxygen ingress into the longwall goaf, with goaf oxygen levels reduced from 20% to 6% at the second cut-through behind the face, i.e. at around 200 m behind the face.

•

•

•

•

•

•

•

Results of field demonstration studies at Mine B showed that the oxygen levels in the longwall goaf reduced from 20% to 3% at the second cut-through behind the face, i.e. at around 100 m behind the face.

Field study results at both the mine sites demonstrated that the proactive inertisation strategies were successful in reducing oxygen ingress into the goaf and in converting the general goaf environment into an inert atmosphere during the study periods.

These field studies also demonstrated that it is possible to reduce the oxygen ingress distance in the goaf to within 200 to 300 m behind the face by implementing appropriate proactive inertisation strategies.

The proactive inertisation strategies have played a crucial role in containing the onset of spontaneous combustion and in reducing the risk of larger heatings development in both the longwall goafs during the study periods.

It was also noted that one standard inertisation strategy was not sufficient for the entire length of the panel or for all longwalls. The inertisation strategies need to be modified with major changes in mining conditions such as ventilation system, geological structures, caving patterns, longwall retreat rates and goaf gas drainage.

The project studies also greatly improved the fundamental understanding of the various parameters and strategies on goaf inertisation process and greatly enhanced the safety of underground coal mines.

(e) Proactive inertisation guidelines It is recommended that proactive inertisation may be introduced in the following scenarios to reduce the risk of heatings development in the longwall goafs:

• Longwall panels extraction in the areas/seams with very high spontaneous combustion potential.

• Extraction in areas with severe geological disturbances, where it is expected that longwall retreat will be very slow for a period of weeks/months.

• Longwall stoppage for prolonged periods – e.g. due to tailgate/face collapse or other reasons.

• Abnormal CO readings behind the face or steady rise in CO readings at a number of seals in the active longwall goaf.

• When working longwall face retreating past suspected heating areas in the adjacent sealed goafs (proactive inertisation in adjacent sealed goaf).

• Abnormal CO readings or other indications of heatings in adjacent sealed goafs (proactive inertisation in adjacent sealed goaf).


BOI.036.001.0106

90

The recommended guidelines for proactive inertisation strategy are:

• inert gas should be injected into the goaf at 200 to 400 m behind the face, or inbye side of a suspected heating location in the goaf.

• inert gas flow rate of around 0.5 m3/s is recommended for most cases (flow rate may need to be increased or decreased based on field conditions - for example, higher inert gas flow rates are required in the case of heatings in open goafs).

• inert gas should be injected on intake side of the goaf in most cases to reduce oxygen ingress into the goaf. (different strategy may be required in case of heatings/ suspected heatings depending on their location and intensity).

• inert gas should be injected through goaf holes (preferable option, if goaf holes are available) - or - through cut-through seals on the intake side of the goaf.

• inert gas may need to be injected on both sides of the goaf if some heating is suspected on return side of the goaf or along major fault areas.

• onsite nitrogen generation units or boiler gas are recommended for proactive inertisation into longwall goafs.

• inert gas injection to be continued until face resumes normal production in case of prolonged stoppages or until face has retreated for more than 300 to 500 m past the suspected heating location, in case of heatings.

• ventilation system in and around the panel also should be designed to minimise oxygen ingress into the longwall goaf for effective inertisation.

A good understanding of the goaf gas flow patterns, the selection of the appropriate injection location points and inert gas flow rates, as well as a good management of longwall face ventilation and pressure drops across the goaf and minimisation of oxygen ingress are all critical to the success of proactive inertisation in the longwall goafs. It is to be noted that most of the above strategies are suitable for proactive inertisation to reduce oxygen ingress into the goaf. The specific inertisation strategies for control of any heatings/suspected heatings may differ from the above generic strategies and depends on the heating intensity, size, location, ventilation system and other mine site conditions. It is also to be noted here that just introducing inert gas into the goaf does not ensure prevention of heating incidents in the entire length of panels or all longwall panels. A number of parameters such as major changes in panel ventilation system, geological structures, goaf caving conditions, longwall retreat rate and goaf gas drainage can make a specific inertisation strategy ineffective and the inertisation strategies need to be modified based on the changed conditions at the field sites.


BOI.036.001.0107

91

6.2 FUTURE RESEARCH It is envisaged that the following areas should be further investigated to improve the practical applications of proactive inertisation strategies in longwall goafs under different geological, mining and operational conditions:

Foam injection technologies for improved goaf inertisation – Further research is needed to develop cost-effective and stable/durable high expansion foam technology and demonstrate its applications to reduce oxygen ingress into the longwall goafs and to improve the effectiveness goaf inertisation in combination with inert gases.

•

•

•

•

•

Investigation of CO and H2 gas flow patterns in longwall goafs – There is an urgent need to develop a fundamental understanding of the heatings, CO and H2 gas levels and their flow patterns and accumulations in longwall goafs. It is important to conduct these modelling investigations under a range of operating conditions with different mine geometries in order to assist in assessment of small area intense heatings and large area low-level oxidation incidents.

Impact of gas drainage on goaf inertisation – In case of highly gassy coal mines, the proper management of goaf gas drainage operations is critical to the success of goaf inertisation process. There is a need for detailed assessment of the effect of goaf gas drainage at different locations and at different drainage rates with respect to inert gas injection locations.

Impact of ventilation system in and around the panel – The control of air ingress into the goaf through proper management of ventilation system is critical to the success of goaf inertisation process. There is a need for detailed assessment of various levels of ventilation flow rates and pressure drops, and the effect of seals leakage to and from surrounding roadways and its impact on inert gas injection at different locations behind the face. Studies should also include the effect of seals leakage between adjacent goafs.

Mapping of sponcom liable zones – further investigations into the mapping of sponcom liable zones according to goaf oxygen levels, geological structures and intrinsic coal propensity to sponcom would be useful for early detection of spontaneous combustion/heatings and for targeted goaf inertisation operations.

- - - - -


BOI.036.001.0108

93

REFERENCES

1. Aziz, N, Srinivasa, R.B. and Baafi, E., Application of Computational Fluid Dynamics

Codes to Develop Effective Gas/Dust Control Measures in Underground Coal Mines, The Australian Coal Journal, No42, 1993, p19-27.

2. Balusu, R, Deguchi, G, Holland, R, Moreby, R, Xue S, Wendt M and Mallett, C,

2001. Goaf gas flow mechanics and development of gas and sponcom control strategies at a highly gassy coal mine, Australia-Japan Technology Exchange Workshop, 3-4 December, Hunter Valley, Australia, 18 pp.

3. Balusu R., Patrick Humphries, Paul Harrington, Michael Wendt and Sheng Xue,

Optimum Inertisation Strategies, Proceedings of the Queensland Mining Industry Health & Safety Conference 2002, 4 - 7 August, Townsville, Australia, pp 133 - 144.

4. Beamish, B and Jabouri, I, 2005. Factors affecting hot spot development in bulk coal

and associated gas evolution, Proceedings of the Coal 2005 – 6th Australasian Coal Operators’ Conference, 26–28 April, Brisbane, Australia, pp 187 – 193.

5. Bell, S, Cliff, D, Harrison, P and Hester C, 1998. Recent developments in coal mine

inertisation in Australia, Proceedings of the Coal 1998 – 1st Australasian Coal Operators’ Conference, 18-20 February, Wollongong, Australia, pp 701 – 717.

6. Cliff, D, Clarkson, F, Davis, R and Bennett, T, 2000. The implications of large scale

tests for the detection and monitoring of spontaneous combustion in underground coal, Proceedings of the Queensland Mining Industry Health and Safety Conference, Townsville, Australia, pp 419 – 428.

7. Cliff, D 2005. The ability of current gas monitoring techniques to adequately detect

spontaneous combustion, Proceedings of the Coal 2005 – 6th Australasian Coal Operators’ Conference, 26–28 April, Brisbane, Australia, pp 219 – 223.

8. Creedy, D P. and Clarke, R D C., Minimising Firedamp Risks on High Production

Coalfaces: A Computational Modelling Approach, Proceedings of an International Symposium: Safety, Hygiene and Health in Mining, Harrogate 1992, Doncaster: The Institution of Mining Engineers, pp192-203.

9. Currie, J A N., Measurement and Control of Diesel Particulate Emissions in

Underground Coal Mines, The Mining Engineer, September 1994, Vol 154 No396, pp58-62.

10. Fletcher, C A J, Jancar, T., Matthews, B., Guzman, M M de and Tu, J Y.,

Computational Simulation for Mineral Processing, Proceedings of the APCOM XXV 1995, Brisbane, Australia, 9-14 July 1995.

11. Gallagher, B., Southland Heating and Mine Fire, International Longwall News,

February 13, 2004.


BOI.036.001.0109

94

12. Healey, P., (1995), ‘1991 Ulan Heating” in Proceedings of the Department of Mineral Resources Spontaneous Combustion Seminar, Mudgee, November 6th to 8th. t

13. Lee, C J., Computational Modelling of Mine Fires, The Mining Engineer, July 1994,

pp17-21. 14. Lynn, K.P. (1987), ‘Warden’s Inquiry - Report on an accident at Moura No.4

Underground Mine on Wednesday, 16th July, 1986' Queensland Government Press. 15. Moloney, K.W, Hargreaves, D.M. and Lowndes, I.S., Computational fluid dynamics

(CFD) simulations in underground auxiliary ventilated headings, Proceedings of the APCOM98, Published by The Institution of Mining and Metallurgy, 19 to 23rd April 1998, London, UK. pp721-731.ISBN 1 870706 36 6.

16. Ren, T X. and Edwards, J S, Three-Dimensional CFD Modelling of Methane Flow

Through Permeable Strata Around A Longwall Face, Trans. Institution of Mining and Metallurgy, Mining Industry Section A, Vol 109, January-April 2000

17. Sullivan, P and Heerden, J.V., The simulation of environmental conditions in

continuous miner developments using Computational Fluid Dynamics, Journal of the Mine Ventilation Society of South Africa, January 1993, pp2-11.

18. Tauziede C, Mouilleau, Y. and Bouet, R., Modelling of Gas Flows in the Goaf of

Retreating Faces. Paper Presented at the 25th International Conference on Safety in Mines Research Institutes at Pretoria, SA. 1993.

19. Tauziede, C, Pokryszka, Z, Carrau, A and Saraux, E, 1997. Modelling of gas

circulation in the goaf of retreat faces, Proceedings of the 6th International Mine Ventilation Congress, 17-22 May, Pittsburgh, USA, pp 243 – 246.

- - - - -


BOI.036.001.0110

c12020 final report - coalminesinquiry.qld.gov.au...tb tb tb tb 17.9 2.9 1.6 76.2 3.2 16.7 24.5 49.5...

Documents