ACARP PROJECT C19010 PUBLISHED 06/07/2015

EMERGENCY RESPONSE: MINE ENTRY DATA MANAGEMENT

Darren Brady, Geoff Nugent, David Cliff, Steve Tonegato, Peter Mason & Seamus Devlin MINES RESCUE SERVICE

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.

ACARP Project Number C19010 Emergency Response: Mine Entry Data Management - Extension

Final Report

Darren Brady Geoff Nugent David Cliff Steve Tonegato Peter Mason Seamus Devlin

DATE OF ISSUE: 6th July 2015

Table of Contents Table of Contents ...... 2

Acknowledgments ...... 4

1. Abstract ...... 5

2. Executive Summary ...... 5

4. Methodology ...... 15

5. Industry Risk Assessment ...... 16

6. Results and Discussion ...... 22

6.1 Current legislative and standards requirements relevant to the research ...... 22 6.2 Recommendations from disaster investigations ...... 36 6.3 Case Studies ...... 43 6.4 Conditions likely to be Experienced During and After an Event ...... 52 6.5 Current status of existing systems ...... 54 6.6 Status in Other Countries ...... 65 6.7 Other Industries ...... 67 6.8 Discussions with Manufacturers and Suppliers ...... 68 6.9 Options for Increasing Survivability ...... 69 6.9.1 Introduction ...... 69 6.9.2 Cable/Tube ...... 69 6.9.2.1 Trenching (Floor) ...... 70 6.9.2.2 Recessing (Roof and Rib)...... 71 6.9.2.3 Cable Positioning/Location ...... 71 6.9.2.4 Armoured Cable/Tube ...... 72 6.9.2.5 Cable/Tube Shielding ...... 74 6.9.2.6 Fire Resistant Cable/Tube ...... 75 6.9.2.7 Fixing/Installation Cable and Tube ...... 76 6.9.2.8 Redundant Cable ...... 79 6.9.2.9 Fibre Optic Cable ...... 80 6.9.3 Shielding ...... 80 6.9.4 Location ...... 82 6.9.5 Fixing ...... 82 6.9.6 Recessing ...... 82 6.9.7 Redundancy ...... 83 6.9.8 Contingent Systems ...... 84 6.9.9 Summary ...... 84

7. Conclusions ...... 85

8. Recommendations ...... 87

Page 2 of 154 9. Bibliography ...... 89

Appendix A: Case Studies ...... 93

A1 , 19th November 2010 ...... 94 A2 Upper Big Branch, USA 5th April 2010 ...... 96 A3 Sago, USA 2nd January 2006 ...... 104 A4 Willow Creek, USA, 31st July 2000 ...... 109 A5 Endeavour Colliery, NSW 28th June 1995 ...... 111 A6 Moura No.2 Mine, Qld 7th August 1994 ...... 113 A7 Moura No.4 Mine, Qld 16th July 1986 ...... 126 A8 Appin Colliery, NSW 24th July 1979 ...... 131 A9 West Wallsend Colliery, NSW, 8th January 1979 ...... 137 A10 Roof Fall in Drift Mine A ...... 138 A11 Roof Fall in Drift Mine B ...... 140 A12 Crandall Canyon, USA, August 6th 2007 ...... 141 A13 Alma Mine No. 1, Aracoma Coal Co., USA, 19th January 2006 ...... 142 A14 #3 Mine Fairfax Mining Co., Inc., USA, September 16th 2002 ...... 143 A15 No. 5 Mine, Jim Walter Resources, Inc., USA, 23rd September 2001 ...... 144 A16 Darby Mine No. 1, Kentucky Darby LLC, USA, May 20th 2007 ...... 145 A17 Truck Fire in Gold Mine ...... 148

Appendix B: Conditions likely to be experienced during and after an event . 149

B1 Explosions ...... 150 B2 Fires ...... 151 B3 Roof Falls, Outbursts and Pillar Failures ...... 152 B4 Inundation ...... 153

Appendix C: Copy of mine survey ...... 154

Page 3 of 154 ACARP C19010 Extension

Acknowledgments The assistance of, and contributions from the following are greatly appreciated:  The respondents from the underground survey  Queensland Mines Rescue Service  Mines Rescue Pty Ltd  AmpControl  Austdac  Northern Lights Technology  Mine Site Technology  Ministry of Business, Innovation and Employment (NZ)  NSW Trade & Investment  Department of Natural Resources and Mines (QLD)  Simtars  MISHC  BMA Broadmeadow Mine  Mine Safety Institute of

Page 4 of 154 ACARP C19010 Extension

1. Abstract

Through a risk management process ACARP Project C19010 (Emergency Response: Mine Entry Data Management) identified critical information required to make informed, risk based decisions on whether mines rescue teams could enter or remain in a mine when responding to an incident. Key areas identified as providing information were monitoring and communication systems. The project also developed a proof of concept software tool (MRAS- Mine Re-entry Assessment System) to assist make informed and considered, risked based decisions founded on predetermined relevant and reliable information.

A question commonly raised throughout the original project was “will the systems providing this required information remain operational once an incident occurs?” Other associated questions raised were; what type and magnitude of incident could render existing systems non-operational in an emergency? What contingencies do operations have in place or available to them to counter this risk? And what can be done to “harden” or protect these systems.

These concerns are not new with the group of experts formed following the Moura No. 2 disaster to address Mines Rescue Strategy Development (Task Group 4), raising many of the same issues. The problem is not unique to the Australian industry; it is a common concern of emergency responders around the world.

To assist address these issues an extension to ACARP Project C19010 was sought and successful, allowing a scoping study to research and identify existing and future strategies, systems and hardware which have the potential to support and provide the information requirements of decision makers during or after an incident at an underground coal mine.

This report delivers the findings of this project and outlines what further research could be undertaken to build on the successful outcomes of this project.

2. Executive Summary

This report is a summary of the work completed as part of the extension to ACARP Project C19010 Emergency Response: Mine Entry Data Management. The original project identified critical information required following an event, to make informed, risk based decisions on whether mines rescue teams could enter or remain in an underground coal mine. The project also developed a proof of concept software tool to assist making informed and considered, risked based decisions founded on predetermined relevant and reliable information. This tool, MRAS (Mine Re-entry Assessment System) is now used by Queensland Mines Rescue when assessing their response to an incident. It was recognised by many (within the research group and external) that some of the data identified as critical may not be available following an event due to either the data source or communication hardware or transmission media being compromised by the event.

The aim of the extension project was to research and identify existing and future strategies, systems and hardware to provide the information required. An important aspect of this project was to identify what type or magnitude of incident could compromise data and communication. Often when “survivability” of these systems is mentioned the immediate event that comes to mind is an explosion. There are however other events capable of compromising essential data or information (including communication) at times when its availability can be critical. To address this, an Industry Risk Assessment was conducted to assess and analyse the potential incidents that could impact on effective communications and mine monitoring systems. A copy of the report generated from this risk assessment is attached to this report.

Existing legislation coupled with guidelines and standards in both New South Wales and Queensland set requirements for the effective communication systems and monitoring during and following an incident. These requirements are summarised in Table 2. New mining legislation was introduced in New South Wales after this report was initially written but the significance is such that the report was updated to

Page 5 of 154 ACARP C19010 Extension include the changes. The new legislation requires mine operators to ensure that arrangements are developed for monitoring following a fire or explosion with consideration given to the design of the post incident monitoring system to increase the likelihood of it being able to continue operating after an incident. Particularly with the introduction of this new legislation a means of quantifying survivability is required.

The requirement for these systems following an event is even more strongly called for in recommendations and reports from previous underground coal mine disasters. These recommendations have been summarised in Table 3. Although the need for post incident communication and monitoring has long been identified there is little guidance material (other than use of boreholes) available on how to achieve this.

When looking at how industries outside underground address these same problems it became very obvious that the environment communications and monitoring systems are required to work under in an underground coal mine are very much different to other industries using similar systems. In particular the confinement of an underground coal mine and the issues that presents during certain events. The requirement for electrical protection systems to cover such expansive areas also adds complication; particularly in relation to independent power supplies to keep communication and monitoring systems operating following the loss of underground power.

Review of past incidents revealed that not all communication and data was lost after an event. There are multiple examples of telephones being used post incident by workers underground advising the surface of the incident and the surface advising workers underground. There are also examples of gas monitoring systems, both real time and tube bundle operating following an event. From this it can be taken that an event doesn’t necessarily mean all is lost. There are however just as many if not more, examples of when communications and monitoring have been totally lost. Also identified are cases where the hardware is still capable of operation, but transmission to the surface has been compromised. Case studies are covered in detail in Appendix A and Table 4 in the main body of the report provides a summary.

Even with all of the previous recommendations and requirements, the underground coal mining industry is not in the position it should be to ensure ongoing reliable information/data during and post incident. In essence mines are set up for “peace time” communications and monitoring. This concern was raised at multiple forums the research team have presented at during the project with little objection or disagreement. These forums and conferences presented the research team with the opportunity to disseminate learnings from the project and make mine personnel aware of the implications of substandard installations. It is too late once an event has occurred to think about improvements.

According to a NSW Department of Industry and Investment report on emergency management systems in NSW underground coal mines (Macpherson 2010) loss of major controls (e.g. communication, monitoring, and transport) were not considered in withdrawal conditions. A reason for the absence of robust systems was attributed to the lack of an initial risk assessment process that meant critical controls were not recognised and included in withdrawal conditions. By employing more robust risk management processes and using findings from this project, underground coal mines should be able to significantly improve the chances of survivability of communication and gas monitoring systems.

Based on the findings of this project it would appear that an off the shelf solution to the problem is not just around the corner. It is noted that due to the possible magnitude and scale of possible events underground that survivability of systems under all conditions could never be guaranteed. There are however simple but effective solutions and practices that could be easily adopted and introduced that could have significant influence on the ongoing availability of communications and data during and after an event. An additional extension to this project was approved by ACARP which allowed for small scale testing using Simtars propagation tube to test the effectiveness of some of the theoretical suggestions. The results of this testing are covered separately in Part B of this report, although findings have influenced the content of Table 1 which outlines options available for increasing the chance of communication and gas monitoring equipment surviving an event. It seen that using these options in combination rather than isolation will add to the probability of survivability.

Page 6 of 154 ACARP C19010 Extension

An issue not covered in this project but highlighted in several case studies is where communication and monitoring hardware is setup on the surface. Mines need to be mindful of where this hardware is located and accessed in relation to exclusion zones. During the response to the Moura No. 2 explosion and the Sago explosion, the main administration areas had to be evacuated due to the concentration of carbon monoxide present. Following sealing Huntly West mine, access to the control room was limited due to it being within the “blast zone” and when the mine did explode surface buildings were damaged. Positioning of monitoring hardware near fans and in box cuts need to be covered in the risk management process for emergency response.

Recommendations and legislation identify the need for systems to survive but little guidance on how to achieve it is available. This may be the reason that mines are not in the position they should be, and they only find out that they are lacking when needed most. Recommendation 7 from Group 4 (1996) was the formation of an industry group to coordinate the advancement of capabilities to alert, to communicate with, and to assess the status of underground persons during a mine emergency. There is a definite need for the reformation of this group and it would be the ideal group to establish a guideline/standard for increasing the probability of survivability. The work from this project could provide the technical basis of any standard developed. It would be appropriate for the group to also include gas monitoring. The Department of Natural Resources and Mines through the Chief Inspector of Coal Mines has agreed to establish such a group to progress the original recommendations and by default continue and progress the work identified in this project.

Page 7 of 154 ACARP C19010 Extension

Table 1: Summary of options for improving survivability of communication and gas monitoring systems

Option Applicable to Commercially Research Advantages Disadvantages Comments available required Trenching Cables-power Yes No research  Greatest level of overall  Installation process Not a new option and limited and data (trenching required but protection  Hardware and connection uptake to date would indicate transmission tool) practical  Proven and favoured technique outside of trench still vulnerable not favoured. Would require Tube bundle demonstration/trial in other industries  Flooding of trench (components a change in attitude in sample lines may be  Everyday protection against need to be water proof) advantages over advantageous vehicle/ machinery interactions  RF and leaky feeder signal disadvantages  Minimised exposure to: strength if buried -elevated temperatures during  Access to buried components fires Floor movement implications, -stretching during roof falls or including those introduced by rib spall trenching -overpressures  Positioning in relation to -flying debris during explosions roadway position and traffic  Equipment used for drainage trenching already available  Can be used just for high risk areas Recessing (roof  Cables-power Yes No  High level of protection  Installation process Roof recessing offers more and rib) and data (cutting tool)  Everyday protection against Access to any cut out through advantages to rib due to transmission vehicle/ machinery interactions mesh issues with installation across  Tube bundle  Protection against  Effect on roof stability cut throughs. May be sample lines overpressures  Dependent on roof/rib favourable for mines using  Protection against flying conditions shotcrete for support rather debris during explosions  Compromised in roof/rib falls than mesh Recessing (roof Communication Yes No research  Everyday protection against  Dependent on roof/rib Communication hardware and rib) hardware (cutting tool) required, but vehicle/ machinery interactions conditions fitted into recess so flush with practical  Protection against  Compromised in roof/rib falls rib. Still requires adequate demonstration/trial overpressures fixing such as to rib bolt may be  Protection against flying advantageous debris during explosions

Page 8 of 154 ACARP C19010 Extension

Option Applicable to Commercially Research Advantages Disadvantages Comments available required Positioning/ Cables-power N/A Not required for  Easily implemented  No additional protection barrier Requires planning (including location and data improvements to  If installed along roof or upper offered other than trying to risk evaluation) and transmission be implemented rib exposure to overpressure can minimise exposure installation standards and Tube bundle but large scale be reduced  No protection against debris documentation sample lines testing may be of  Can be located out of way of minor benefit normal operational interactions  Can be adjusted for identified risk e.g. roof area likely to reach elevated temperatures during fire so where fire the major risk may be preferential to run along rib. Positioning/ Hardware N/A No  Easily implemented  No additional protection barrier Requires planning (including location  Minimal additional cost offered other than trying to risk evaluation) and  Can reduce exposure to flying minimise exposure installation standards and debris  May require additional documentation Foster-Miller  Can reduce exposure to total signage/alert devices to indicate (2009) recommended blast pressure location (particularly important mounting equipment in cut for communication devices that throughs rather than main may be located in cut throughs roadways as a means of rather than main escape travel protection against debris and roads overpressures (reduced by at least 1/3 of the total blast pressure). Need to consider which rib in cut through less likely to be exposed. Also reported that locations greater than 300m from methane explosion should not experience significant blast pressure or debris. Route Cables-power N/A No  Easily implemented  No additional protection barrier Requires planning (including and data  Minimal additional cost offered other than trying to risk evaluation) and transmission (dependent on approach taken) minimise exposure installation standards and Tube bundle  Can be adjusted dependent on  May incur additional cost, documentation sample lines identified risk (e.g. high risk particularly if boreholes used to areas such as belt roads avoided) limit underground runs  Running cables and tubes up boreholes minimises exposure to over pressures, flying debris and falls underground

Page 9 of 154 ACARP C19010 Extension

Option Applicable to Commercially Research Advantages Disadvantages Comments available required Armouring Cables-power Yes No  Used in other heavy industries  Joins and connection to Connection hardware needs to and data  Protection against over hardware can be more complex be as strong as the cable transmission pressure, crush forces, abrasion,  More expensive Tube bundle kinking and debris  Heavier, making installation sample lines  Can just be used in identified harder high risk areas  Using multicore tube decreases installation times Shielding Cables-power Yes No research  Used in other industries  Installation process The same and data required but  Wide range of conduit options  Access attention/requirements to transmission practical  Protection against over  Additional cost fixing/attachment needed for Tube bundle demonstration/trial pressure, crush forces, abrasion,  Heavier assembly conduits/cable trays etc. as sample lines may be kinking and debris  connection to components may required for cables/tubes. advantageous  Can be used just in high risk be more complex areas such crossing cut throughs  Area between component and shielding still vulnerable Shielding Hardware No, but easily Small scale test  Protection against over  Increased installation Shielding requires to be manufactured research pressures and flying debris  Possible access issues secured adequately promising, mine possible site trials required  Can be retro fitted  No recertification of electrical equipment required Fire Resistant  Cables-power Definitely for Not for cable but  Used in other industries  More expensive AS/NZS 3013:2005 – and data cables. Tube possibly research  Cables capable of withstanding Electrical installations- transmission available but into tube product over 1000oC available Classification of the fire and  Tube bundle not typical most suited to tube  Could be used in high risk mechanical performance of sample lines tube bundle bundle areas such as belt roads wiring system elements exists tube. advantageous.

Page 10 of 154 ACARP C19010 Extension

Option Applicable to Commercially Research Advantages Disadvantages Comments available required Fixing/Installation Tube and cable Yes Small scale testing  Easily introduced  May increase installation times Recommendation is for (refer to Part B has shown that  Pulled tight reduces the chance  Doesn’t offer complete installation with no sags and of this report) standards use of damage from flying debris protection against flying debris fixed every 1m. Hooks not significantly  Testing shows it makes a  Needs to be maintained suitable, cable ties OK. influence difference Where identified that some survivability “additional slack may be required or slack is present due to operational reasons, the slack must be secured adequately at multiple points and protected against stress beyond the minimum bend angle (wrapped around a protecting device ideal). Connections and joins need to be protected as well Fixing/Installation Hardware Yes ? standards could  Easily introduced  Doesn’t offer complete Use of clips and rope looped (refer to Part B be improved  Secure hardware can also protection against flying debris over rib bolts appears of this report) without research reduce damage to transmission common- unlikely to survive. media Hooks are not suitable Using backing plates secured to threaded bolts will significantly increase survivability Redundancy Tube and cable Yes No  Using a ring network would  No guarantee that redundant Redundant transmission allow data/ communication tube/cable would survive media runs should preferably transmission from both inbye and  Extra installation effort run in different roadways to outbye breakage points  Leak testing requirements for the primary  Spare tube advantageous if additional tubes sampling coal fire and tar blocks  Additional hardware required primary sample tube other end of ring network Redundancy Wireless Yes Maybe to show  “Self-healing”  System needs to be designed Used in US to meet transmission survivability adequately and single paths requirements of the MINER avoided Act  Require underground power Redundancy Hardware Yes No  Increases chance of  Doesn’t offer any physical Redundant communication – survivability protection legislative requirement.  Can cover different scenarios  The one event may destroy both Design needs to be evaluated with different techniques systems using a risk management  Additional expense process to cover what could go wrong where.

Page 11 of 154 ACARP C19010 Extension

Option Applicable to Commercially Research Advantages Disadvantages Comments available required Contingent Hardware and Yes Improvements  Won’t be destroyed by event  Typically slow to be deployed Contingent systems need to Systems transmission required so  Measuring ranges etc. can be  Often rely on boreholes which if be included in emergency media possibly matched to the event not predrilled will take response plans and resourced Can be pre-planned and “ready considerable time and as seen in adequately with limitations to go” case studies have often missed noted and understood  High risk areas can be intended location causing identified and boreholes significant delays preinstalled Housekeeping Tube, cable and N/A No  Easily implemented  Only assists in prevention of Noted by persons spoken to hardware  Can be used with other damage from flying debris that good housekeeping techniques  No actual protection standards minimised the  Added advantages-not just amount of flying debris survivability during explosions. Also advantage as far as escape

Page 12 of 154 ACARP C19010 Extension

3. Introduction

In an effort to ensure that emergency mine re-entry processes were developed using risk management methodologies, Queensland Mines Rescue Service (QMRS) facilitated a comprehensive risk assessment involving all key stake holders including New South Wales Mines Rescue (NSWMR). The risk assessment took four days to complete and resulted in identifying what information was required to make a risk based decision on the deployment of rescue teams underground. A task group was formed that converted the risk assessment into mine re-entry guidelines from which the initial ACARP funded project C19010 “Emergency Response: Mine Entry Data Management” developed a decision support system. The system known as MRAS (Mine Re-entry Assessment System) incorporates check lists of the information determined to be required as well as the capability of accessing information known prior to the event. Making use of the this tool with underpinning risk management logic, assists decision makers make informed, quality decisions in a timely manner. MRAS is now used by QMRS when evaluating their response to an incident.

Workshops were specifically developed and delivered in Queensland and New South Wales as part of the information dissemination process of the initial project. Presentations at local and international conferences were also made. A common question and one that was raised many times by the project team themselves was “will the systems providing the critical information survive the event?” This question led to others such as; what can mines do to protect these systems? What type and magnitude of event will compromise systems? What contingencies are in place or available to cover such events?

One of the main findings of the initial project was that mining operations give little consideration to the locality and survivability of information and communication systems critical to obtaining the required information post an incident of any type or size. The recommendations made in the initial project report included the following;

There is an urgent need to assess and determine the general capability and capacity of existing critical mine monitoring and communication systems to provide reliable and relevant information during emergencies of varied types and size. The research should include;

 The general status of Australian underground coal mines environmental monitoring and communications systems capabilities and capacities to provide adequate information after an incident.  What structural design specifications and strategic positioning considerations (including contingences) for environmental monitoring and communications systems would be considered best practice for emergency response?  The current status of existing systems available and suitable for Australian underground coal mines relevant to the scope.  The current status of any research and development being carried out or pending applicable to the scope.  What further specific research is needed to assist industry in implementing the functional specification developed in the original C19010 project and recommendations of how such research could best be achieved?

This is by no means the first time that recommendations of this type have been made. The report on the Warden’s Inquiry into the Moura No.2 Mine disaster (Windridge 1995) stated “The loss of telephone communication with the 5 South crew at the time of the first explosion left no means of ascertaining the status of those persons without some form of entry to the mine. There appears a need to examine explosion resistant means of communication”.

Task Group 4 that formed following the Warden’s Inquiry into the Moura No. 2 explosion to look at Mines Rescue Strategy Development determined that with regards to existing practice, knowledge of

Page 13 of 154 ACARP C19010 Extension conditions underground after an incident would be insufficient for accurate assessment of the mine environment (Task Group No. 4, 1996). The task group also identified “The highest priority need is for a communications system which would survive an incident and provide ongoing two way communications between escaping or trapped miners and rescue personnel on the surface”.

With regards to gas monitoring, Task Group 4 recommended that fixed tube bundles and gas chromatographs should be made available at all mines as the primary method of measuring post incident mine atmospheric conditions, something that many Australian mines have achieved. The recommendation went further stating that tube bundle systems should include;

 protection of tubes from damage,  locations of sampling points designed for both normal and post incident atmospheric conditions,  modularisation using boreholes to minimise delay in transmission and analysis as well as to make the system as robust as possible,  techniques for verifying tube integrity which could be routinely applied post incident.

Addressing these additional requirements is where Australian mines have failed in implementing tube bundle systems with increased likelihood of survivability for use post incident.

For real time (telemetric) systems Task Group 4 recommended “Research into the development of robust telemetric sensors for gas analysis and other environmental parameters, over the ranges existing after incidents, should be prioritised.”

Task Group 4 also highlighted that both pre-installed and post-incident boreholes should be considered when developing Aided Rescue Management Plans.

Even with these and subsequent recommendations and requirements (discussed in more detail within this report), the underground coal mining industry is not in the position it should be to ensure ongoing reliable information/data during and post incident. In 2010 the NSW Department of Industry and Investment released a report (Macpherson 2010) on emergency management systems in NSW underground coal mines. It was noted in this report that loss of major controls (e.g. communication, monitoring, and transport) were not considered in withdrawal conditions. When looking at communication the report stated “there should be an emphasis on mines moving to the best systems with consideration to “survivability” from foreseeable events (explosion or fire)”. A reason for the absence of robust systems was attributed to the lack of an initial risk assessment process that meant critical controls were not recognised and included in withdrawal conditions.

Most of the research team was involved in the response that followed the explosion at Pike River, where many of these same problems limited the ability to make timely informed decisions. With this knowledge and knowing that the information identified in ACARP project C19010 as critical for decision makers was possibly not going to be available during an incident in Australia, the project team made a successful application to ACARP for an extension to the project to address the research recommended. The aim of the extension being to research and identify existing and future strategies, systems and hardware which have the potential to support and provide the information requirements of decision makers during or after an incident at an underground coal mine.

Page 14 of 154 ACARP C19010 Extension 4. Methodology

Essentially this project is a scoping study aimed at providing the Australian underground coal mining industry information to put it in the best possible position to make informed decisions relating to ways available to ensure critical information is available post event. The project aimed to answer the following questions:

 What is the general status of Australian underground coal mine monitoring and communication systems to provide adequate information after an incident?  What structural design specifications and strategic positioning considerations (including contingences) for monitoring and communications systems would be considered best practice for emergency response?  What is the current status of existing systems available and suitable for Australian underground coal mines?  What is the current status of any research and development being carried out or pending applicable to this topic?  How can operations effectively sustain the systems which could provide the information required once an incident occurs?  What level and type of incident today could render our existing systems obsolete in an emergency?  What contingencies do operations have in place or available to them to counter this risk?  What further specific research is needed to assist industry in implementing the functional specification developed in the original C19010 project and recommendations of how such research could best be achieved?

The project research team developed and adopted the following approach to complete this project and help answer the questions posed above.

1. Identify and collate current and available information from recent incidents relevant to the research questions. 2. Identify current legislative and standards requirements relevant to the research questions. 3. Identify and collate existing recommendations from relevant incident investigations and Level 1 exercises (post Moura No. 2). 4. Conduct Industry Risk Assessment to assess and analyse the potential incidents which could impact effective communications and mine monitoring systems. 5. Review the current status at existing mines in NSW and Queensland based on the research questions and highlight best practice where possible. (Industry workshop for operations to present existing standards based on research questions in NSW and QLD) 6. Investigate and identify strategies and systems utilised inside and outside the Australian coal mining industry which may have potential for direct implementation or be improved to implement in Australia (identify industries with similar issues and analyse controls for further investigation). 7. Identify and assess studies and research already conducted or in progress relevant to the research questions and evaluate their potential effectiveness for industry. This should include the international mining industry, and where applicable, industries external to the coal industry. 8. Industry knowledge transfer; Information transfer via industry workshops, seminars, conference presentations and the development of a report with recommended specifications that can also be provided to the operations, service providers and manufacturers. 9. Provide recommendations on specific research required to assist industry achieve the recommended specifications.

Page 15 of 154 ACARP C19010 Extension 5. Industry Risk Assessment

One of the initial tasks the project team completed was an industry risk assessment to assess and analyse the potential incidents that could impact effective communications and mine monitoring systems supplying data and information. This risk assessment defined critical data/information sources (Figure 1), the types of sensors/hardware collecting data (Figure 2) and the transmission medium of that data/information (Figure 3). A detailed hazard analysis was then conducted to identify the incidents that could compromise the capture, supply and transmission of the critical data/information to the post-incident management team (Figure 4).

These threats were analysed for their consequences and potential magnitude. The current preventative and mitigation controls employed to maintain access to this data and information were also identified. It became very apparent during this process that most of the systems set up to collect and transmit critical data/information were done so in a manner that addressed the day to day requirements for this information with little consideration given to expected measurements or increasing the chance of survivability during or following an event. It appears systems are scoped, designed and installed for “peace time mining”.

Where possible the industry risk assessment team identified areas requiring improvement. A separate report detailing the risk assessment is included as an attachment to this report.

An issue that was identified as requiring urgent investigation was the need for independent intrinsically safe power supplies for systems providing critical data, such as gas monitoring and communications, when underground power is lost. The use of battery backup to address this problem is only useful to overcome short duration power supply issues. It is not adequate for an ongoing response that may last more than several days. Recent events in Australia (albeit lower consequence) and at Pike River in New Zealand have involved responses and withdrawal to the surface best measured in weeks rather than days, and would far exceed battery backup capacity to keep systems operational.

As a direct result of the identification of this problem the project team organised for manufacturers, certification experts and electrical inspectors to meet to discuss the issue and attempt to and come up with a solution to provide power for ongoing provision of critical data from such systems during extended periods of underground power loss. The outcomes from these meetings were positive with some suppliers already looking at the problem and investigating alternative solutions. Essentially underground operations could work with suppliers to provide solutions to this problem now. Ways of extending the time these systems not far from being commercially available include gas sensors with much lower power consumption requirements and the ability to turn sensors on and off from the surface so as to only draw power when measurements were required.

Another area that became obvious during the risk assessment was the need for “hardening” of sensors and other hardware, including transmission media to protect against both over pressure and flying debris. It is recognised that hardening of the sensors by the manufacturers although important is not going to be achieved in the short term (confirmed by manufacturers). For this reason the project team has recommended further investigation into designing retro fitted physical protection to critical hardware. There is also a need to investigate the most appropriate means of installation of hardware and transmission media (including tube bundle sampling tubes). The positioning of the hardware also needs to be considered as part of the installation. Communication hardware such as telephones may be better placed in cut throughs rather than main headings to provide protection.

It is normal practice following an incident such as an explosion underground to ring all underground phones and regularly communicate over the intercom systems (DAC) to try and get a response from workers underground. One of the problems frequently faced at these times when no reply is received, is knowing whether the communications are still operational underground. When this issue was brought up with suppliers they indicated that there are already ways that this can be determined.

Page 16 of 154 ACARP C19010 Extension

It was also identified that a desirable feature on intercom type systems is an override button, allowing the surface operator to activate the talk function at the unit underground allowing them to hear what was happening in that area.

The need for systems with inbuilt component redundancy, to increase the probability that the system could survive an explosion, was also identified. It was determined probable that borehole based systems would be required and if information was to be available in a timely manner these bore holes would need to be pre drilled with a risk based approach utilised to determine required locations.

Page 17 of 154 ACARP C19010 Extension

Figure 1: Critical Data by Type

Page 18 of 154 ACARP C19010 Extension

Figure 2: Critical Sensors/Data Sources

Page 19 of 154 ACARP C19010 Extension

Figure 3: Data Requirements

Page 20 of 154 ACARP C19010 Extension

Figure 4: Data Source Threats

Page 21 of 154 ACARP C19010 Extension 6. Results and Discussion

6.1 Current legislative and standards requirements relevant to the research

As part of the scoping study undertaken by the research and proposed as part of the work program the research team identified relevant requirements of existing legislation, standards and guidelines. Table 2 identifies these requirements for Queensland and New South Wales relating to the ongoing provision of monitoring and communications following an incident.

Table 2: Communication and monitoring requirements from legislation and standards

Document C/M/B* Clause/ Section Topic Requirement Coal Mining B 6 Objects of Act (a) to protect the safety and health of Safety and persons at coal mines and persons who may Health Act be affected by coal mining operations; and 1999 (Qld) (b) to require that the risk of injury or illness to any person resulting from coal mining operations be at an acceptable level; B 7 How to achieve (b) providing for safety and health objects of the Act management systems at coal mines to manage risk effectively; (j) providing for a satisfactory level of preparedness for emergencies at coal mines Coal Mining C 33 Plan of (2) For an underground mine, the site senior Safety and communication executive must also ensure the mine has a Health system plan of its communication system showing Regulation the location of each fixed communication 2001 (Qld) device at the mine. (4) The plan must be updated as soon as practicable, but not later than 8 days after— (a) for a plan of the communication system—the installation or removal of a communication device; B 35 Safety and health (a) identifying, by risk assessment, potential management emergency situations; system for (b) minimising risks associated with managing potential emergency situations; emergencies (d) carrying out emergency exercises, including testing the effectiveness of emergency management procedures and the readiness and fitness of equipment for use in an emergency; B 149 Principal hazard (a) emergency response management plans for at least the following- C 155 Fire officer (e) testing, and reporting on, the condition responsibilities of the mine’s communication system. B 168 Considerations for (a) the location of devices for assisting self- self-escape risk escape; assessment (b) the number of devices, including self- rescuers, to be distributed throughout the mine; (f) communication equipment and ways of using the equipment; C 176 Telephonic (1) The site senior executive must ensure communication the underground mine’s telephonic communication system complies with this

Page 22 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement section and has an adequate failsafe, or backup, power supply for the system. (3) The system must provide for effective telephonic communication to and from the following places at the mine— (a) each entrance underground, and on the surface, through which a person may enter into a shaft or other excavation used for ingress to or egress from the mine workings; (b) each underground battery charging station; (c) each underground workshop; (d) each underground crib room; (e) a place close to each switchgear used to isolate or control sections of the underground main electricity distribution system; (f) a place close to each underground conveyor belt drivehead; (g) a place close to each underground loading or transfer point on the conveyor belt system; (h) each emergency refuge chamber; (i) a place in each inspection district in the mine not otherwise mentioned in paragraphs (a) to (h) M 222 Gas monitoring (3) The gas monitoring system must also system provide for— (a) an alternative electricity supply to ensure the system continues to function if the normal electricity supply fails; M 253 Withdrawal of For section 273 of the Act, a part of an persons in case of underground mine is taken to be dangerous danger caused by if the part is affected by the failure or non- failure or non- operation of the gas monitoring system and operation of gas the mine does not have— monitoring system (a) a standard operating procedure for using portable gas detectors; or (b) sufficient portable gas detectors to continually monitor the part to the extent necessary to achieve an acceptable level of risk. B Schedule 5 Matters to be 8 the proper functioning of communication covered in and monitoring systems inspections Work B 3 Objects of the Act (b) to protect workers at mines and other Health and persons against harm to their health and Safety safety through the elimination or (Mines) Act minimisation of risks arising from work or 2013 from specific types of substances or plant, (NSW) (c) to ensure that effective provisions for emergencies are developed and maintained at mines,

Page 23 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement Work B 23 Principal mining (2) The mine operator must conduct, in Health and hazards and relation to each principal mining hazard Safety conduct of risk identified, a risk assessment that involves a (Mines) assessments comprehensive and systematic investigation Regulation and analysis of all aspects of risk to health 2014 and safety associated with the principal (NSW) mining hazard. (3) The mine operator, in conducting a risk assessment under subclause (2), must: (a) use investigation and analysis methods that are appropriate to the principal mining hazard being considered, and (b) consider the principal mining hazard individually and also cumulatively with other hazards at the mine. B 24 Principal mining (2) A principal mining hazard management hazard management plan must: plan (a) provide for the management of all aspects of risk control in relation to the principal mining hazard, (3) A principal mining hazard management plan must: (c) describe the analysis methods used in identifying the principal mining hazard to which the plan relates, and (e) describe the investigation and analysis methods used in determining the control measures to be implemented, and (f) describe all control measures to be implemented to manage risks to health and safety associated with the principal mining hazard, and (h) refer to any design principles, engineering standards and technical standards relied on for control measures for the principal mining hazard, and (i) set out the reasons for adopting or rejecting each control measure considered. (4) The mine operator of a mine must consider the following when preparing a principal mining hazard management plan for a principal mining hazard at the mine: (b) any other matter relevant to managing the risks associated with the principal mining hazard at the mine. C 32 Communication (2) (f) (vii) in the case of an underground plan mine, the location of each fixed communication device at the mine C 51 Communication The mine operator of an underground mine systems must maintain a system for effective communication throughout the mine and between the surface and locations at the underground mine. B 57 Air quality (2) The mine operator of an underground mine must immediately notify any affected workers or other persons at the mine of the relevant circumstances referred to in subclause (1). M 73 Gas monitoring (1) The mine operator of an underground coal mine must ensure that: (b) a gas content monitoring system is put

Page 24 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement into place that: (i) identifies the locations at which the gas content of air is to be monitored (d) an accurate plan of all gas content monitoring plant for the mine is maintained that specifies the locations at which air is monitored (i) detection heads of gas content monitoring plant are positioned to maximise the likelihood of detecting the gas being monitored and producing accurate readings, and (h) all gas content monitoring plant is calibrated and maintained, and (j) gas monitoring plant has an alternative power supply to ensure, so far as is reasonably practicable, that the plant continues to function if the normal power supply fails M 77 Post incident (1) The mine operator of an underground monitoring coal mine must ensure that arrangements are developed and implemented in accordance with this clause for the monitoring, so far as is reasonably practicable, of the atmosphere of the mine following an explosion or fire that leads to the withdrawal of persons from, and the cutting of the supply of power to, all or part of the mine. (2) In developing and implementing the arrangements the mine operator must ensure that consideration is given to the following: (a) the optimum locations for monitoring points, (b) the gases to be monitored, (c) the design of the post incident monitoring system to increase the likelihood of it being able to continue operating after an incident, (d) the need for and availability of external resources, (e) the regular testing of the arrangements. B 88 Duty to prepare (1) The mine operator of a mine must emergency plan prepare an emergency plan for the mine in accordance with this Subdivision. (2) In addition to the matters required by clause 43 (1) of the WHS Regulations, the emergency plan must: (a) address all aspects of emergency response, including by ensuring: (i) the establishment of a system that enables all persons at the mine to be promptly located, and (ii) that a record of all persons who are underground at a mine (other than an opal mine) at any given time and each person’s likely location is accurately maintained and is readily available in an emergency, and (iii) the provision of adequate rescue equipment

Page 25 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement B 92 Resources for The mine operator of a mine must ensure emergency plan that: (a) all resources, including rescue equipment, specified in the emergency plan for the mine are provided in accordance with the plan, and (b) all resources required for the effective implementation of the emergency plan are provided, and (c) all plant and equipment, including communications systems and rescue equipment, specified in the emergency plan is regularly inspected and maintained in good working order. B 97 Safe escape and (1) The mine operator of an underground refuge mine must provide adequate means of communicating with all affected persons when the emergency plan for the mine is implemented. Example. An alarm system. (2) The mine operator of an underground mine must ensure, so far as is reasonably practicable, that the communication systems for the underground mine enables communication to be established: (a) between persons underground in different parts of the mine, and (b) between persons underground and persons at the surface in the case of an emergency, and (c) across strategic locations at the mine, being places critical for communicating with persons in an emergency (such as refuge chambers, caches, refill stations, change-over stations and escape routes), and (d) from places unaffected by hazards associated with an emergency to those that are affected. (3) The mine operator of an underground mine must ensure that any power operated communication equipment used as part of a communication system for the mine (including the power supplied to that equipment) incorporates a fail safe or back up power supply for the critical parts of the system. (4) The mine operator of an underground coal mine must ensure that any power operated communication equipment used as part of a communication system that is installed underground (unless installed in a drift or shaft being driven from the surface in material other than coal) must be suitable for use in a place where the concentration of methane in the general body of the air is greater than 2% by volume. (6) The mine operator of an underground coal mine must, in conjunction with providing an adequate means of escape, ensure that an overall emergency escape to the surface strategy is developed for the

Page 26 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement mine that takes into account the following: (d) the provision of water and communications at refill stations and change-over stations, and (e) procedures, so far as is reasonably practicable, for rehydration and communication in an irrespirable atmosphere, and (f) provisions for monitoring the respirable air both within and outside a change-over station, C 102 Competent persons The mine operator of an underground mine at surface must ensure that at any time that persons are underground: (a) at least one person at the surface (the surface contact) is readily available to be contacted by those persons underground, B Schedule 7 Matters to be 1 Site and hazard detail included in (8) The infrastructure likely to be affected emergency plan for by an emergency. a mine 4 Resources and equipment (4) For an underground mine, a means of communication between the surface of the mine and any underground area of the mine where persons are located, that is effective, so far as is reasonably practicable, even if there is no electrical connection between the surface and the relevant underground area. Coal Mine B 47 Emergency (2) An emergency management system Health and management must adequately address, but is not limited Safety Act system to addressing, the following matters: 2002 (b) the description of the measures to be No. 129 taken to prevent or limit the harmful (NSW) consequences of incidents associated with (Repealed) each of the identified risks, including measures to identify the location of people who may be at risk, (c) the identification of the equipment, facilities and communication systems necessary to control or limit the consequences of those incidents and the arrangements for ensuring that they are readily available, Coal Mine C 16 Information and The information and communication Health and communication arrangements for a coal operation must Safety arrangements make provision for the following: Regulation …. 2006 (l) communications from the surface part of (NSW) the coal operation to people at the following (Repealed) locations underground by means of an intrinsically safe voice communication system (explosion protection category Ex ia): (i) every underground entrance to a shaft or outlet used for providing means of ingress and egress to people working at the coal operation, (ii) every place underground where plant is regularly serviced or charged, (iii) a place within each underground

Page 27 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement production area at which the telephone is reasonably accessible and not likely to be damaged, (iv) a place in proximity to underground switch-gear used for the purpose of isolating sections of the main underground electricity distribution system, (v) a place in proximity to every underground conveyor belt drive head and transfer or loading point. M 18 Monitoring (1) The monitoring arrangements for an arrangements underground mine must make provision for the following: (h) the strategic positioning of detection heads of gas monitoring plant to maximise the likelihood of detecting the gas being monitored while producing accurate readings or activation of alarms (k)gas monitoring arrangements that: (i) specify the locations at which the gas content of air is to be monitored, B 35 Contents of major For the purposes of section 36 of the Act, a hazard management major hazard management plan in relation plan: fire and to a major hazard comprising hazards explosion arising from fire and explosion must make management plan provision for the following matters: (a) regular assessment of the fire and explosion risk at the coal operation, (b) implementation of control measures to effectively manage risks identified, (c) the means by which the requirements of Subdivision 1 of Division 1 of Part 4 relating to fire and explosion risk are to be implemented at the coal operation. B 45 Contents of For the purposes of section 47 (2) (h) of the emergency Act, an emergency management system for management a coal operation must adequately address system the following matters in addition to those specified in section 47 of the Act (b) fire and emergency provisions for the underground parts of the coal operation, including the following: (ix) the appointment of competent persons to be on duty on the surface part of the coal operation whenever anyone is in the underground parts of the coal operation, with effective means of communication to people in the underground parts of the coal operation MDG 1020 B Purpose and Capability and All mineworkers must be provided with the Guidelines scope resources to capability and resources to facilitate escape for facilitate escape from their place of work to a place of underground safety. emergency Capability means: escape • provision of the most rapid, efficient and systems and safe means of escape available in each the circumstance that might be encountered provision of underground self rescuers • an effective communications system (NSW) • an early warning system

Page 28 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement M 2. Emergency Early warning Main risks escape system – • Monitoring system does not survive event elements and • Alarm system fails/damaged in event considerations Main risk considerations • Monitoring system adequately designed, maintained and calibrated • Detection points positioned and alarms initiated in appropriate locations • Monitoring sensor specification • Integrity and protection of the system during an event • Contingencies in the event of a failure of the primary monitoring system C Communication Required outcome: Effective communication to all persons required to work or travel underground on the paths of egress from each part of the mine. An adequate communications system makes possible the co-ordination of all other systems and provides the key to early notification of an emergency event and coordination of response.

When specifying the design, construction and installation of hardware associated with the emergency communication apparatus, the integrity of the system needs to be assured during any event that causes an emergency. Essential communication systems should not be left exposed in the mine and vulnerable to easy or casual damage. The Risk Assessment and Risk Management Systems must aim to preserve the functionality of the system through a catastrophic event. System Hardware: As a minimum requirement, fixed communication systems to the surface control shall be provided at the following locations:- (1) Underground • Working Places • Crib Rooms • Air splits at the entrance of panels • First withdrawal response muster areas and places of safety • Subsequent, higher level withdrawal response muster areas and places of safety - main headings - pit bottom (2) Surface • Surface control • Incident control centres • Control centres for the removal and restoration of power • Control centres for gas monitoring, for example, the tube bundle hut • Control centres for the starting and stopping of fans These locations shall be provided with fixed

Page 29 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement communication means to enable contact with other areas of the mine and surface, independent of underground power. Multiple redundancy: The fixed communication systems shall be augmented by a minimum of one secondary communication system. At least one of these will be independent of the underground power. Main risks: • All underground personnel not notified of the need to escape and details of the incident/ safest route of escape • Escaping persons cannot locate or use communication system • Personnel do not respond to the incident control centre • Communication systems do not survive the incident • Communication system is not reliable Main risk considerations • Communication system operates in the absence of underground power • Contingency plans are in place should the communication system fail MDG 1003 M Gas Monitoring Analysis of The mine must establish and maintain a Windblast explosive gases in robust system of gas monitoring capable of Guideline the Hazardous zone providing real time gas analysis of (NSW) explosive gases in the Hazardous zone – Windblast before, during and after a windblast. MDG 1004 C Escape/First Communication The mine must develop and implement Outburst Response processes to ensure the safety of all Rescue personnel in areas affected by an outburst. These are to cover both escape (self rescue) and assisted, first response rescue of affected persons and must include, but may not be limited to the following: … • notifications and communications;

Recognised B Appendix 1 – Emergency Emergency Principal Hazard Management Standard 08 Definitions And Principal Hazard Plan shall- (Qld) Abbreviations Management Plan • Include, but not be limited to, organisational structures for emergency response, planning, activities, responsibilities, communications, practices, risks identified, audits and reviews. C Place of Safety A designated place where persons will assemble without being in any danger from the hazard that triggered the evacuation. The place of safety – • Must have an effective means of communication with the surface control B Appendix 2 – Effectiveness of the • Emergency communications Guidelines For Mine Control • Gas monitoring The Room

Page 30 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement C Organisation, With Respect to • Communications Management “SELF ESCAPE” And Conduct Facilities and Of A Level 1 Strategies Mine Emergency Exercise B Appendix 3 – Plan Element 4.Familiarity with mine egress routes, Mine ventilation systems, communication Emergency systems, emergency equipment Response Plan 8. Communication procedures – Audit / • Hardware Review Process 12. Emergency response equipment and installations e.g. surface airlocks / emergency seals Draft Code B 2.2 Contents of the The mine operator should include of Practice emergency plan additional information based on the Emergency outcome of the site specific risk Response at management processes. Risk management Australian processes will assist the mine operator to: Mines (Safe • determine the hazards produced as a result Work of the emergency that workers self escaping Australia) and self rescuing may face and what responses are necessary to effect their escape or rescue • determine what responses are necessary to ensure all people escape safely, such as first response, self-escape, aided escape or aided rescue, and • determine what other resources may be needed in order to effectively manage emergency situations. The emergency plan should address all aspects of primary emergency response. For emergency responses to have any chance of functioning properly, a well thought out and practiced emergency plan should exist together with enough infrastructure and resources to ensure that the plan is operable. The emergency plan should also, so far as is reasonably practicable provide for: • strategically positioned and well protected fire and or gas monitoring equipment • communications to all relevant persons • locating and accounting for persons at all times B 3 Site And Hazard • the infrastructure likely to be affected by a Details major incident. C 5.2 Communication - Communication systems must be effective underground mines during an emergency. In order to achieve this the mine operator must ensure that the communication system: • incorporates an adequate fail safe, or backup, power supply for the system • where electrical components are installed underground in a coal mine are suitable for use in an explosion risk zone that may contain greater than 2% methane unless the components are installed in a drift or shaft being driven from the surface in material

Page 31 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement other than coal • provides for effective telecommunication to and from - the entrance underground, and on the surface, of the mine through which a person may enter into a shaft or other excavation used for ingress or egress from the mine workings - each underground battery charging station, workshop and crib room - a place close to each switch gear used to isolate or control sections of the underground main electricity distribution system - a place close to each underground conveyor belt drive head - a place close to each underground loading or transfer point on the conveyor belt system - each emergency refuge chamber M 6 Resources and Schedule 9.4 clause 4 of the WHS Equipment Regulations requires the following resources and equipment to be included in the emergency plan: • on-site emergency resources, including first aid equipment, facilities, services and personnel, emergency equipment and personnel, gas detectors, wind velocity detectors, sand, lime, neutralising agents, absorbents, spill bins, and decontamination equipment. C 7.6 Place of Safety A place of safety is a designated place where persons can assemble without being in any danger from the hazard that triggered the evacuation. The place of safety: • must have an effective means of communication. C 7.9 Monitoring of the Each mine should have a system to location of persons monitor: • persons entering and exiting the mine, and • the general location of persons while underground. The system should provide an ability to check that all affected persons have moved to the required place of safety. The system should • provide adequate detail of location of all people underground, and • identify places of safety with communication to surface control. B 7.11 Emergency The emergency plan should provide for evacuation - evacuation from the mine when conditions general of potential or imminent emergency requiring escape from a mine are identified. The appropriate alarm must be communicated to persons who may be endangered. When developing a monitoring system that will trigger evacuation, the mine operator should consider: • monitoring system adequately designed, maintained and calibrated

Page 32 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement • detection points positioned and alarms initiated in appropriate locations • integrity and protection of the system during an event • contingencies in the event of a failure of the primary monitoring system C 7.12 Emergency The communication systems should be telephone augmented by a minimum of one secondary procedures communication system. At least one of these should be independent of the mine power. C 7.15 Refuge procedures • occupants of refuge chambers have means of communication to surface C 7.16 Boreholes The mine operator should consider the use procedures of boreholes in communications and air supply to refuge stations and/or respirable air change-over stations, and in the recovery of personnel from underground workings. When considering this option the mine operator should ensure: • a suitable rig is available within appropriate timeframe or borehole is pre drilled • where a borehole is part of planned rescue strategy the surface location is available, secure, surveyed, cleared, consolidated and provided with all weather access • where a borehole is part of planned rescue strategy the underground target site is surveyed, suitably supported, cleared and marked, and • where a borehole is part of planned rescue strategy the depth, stratigraphy, hole stability and drill-ability should be known. M 7.19 Sealing and The procedure for sealing a mine should inertisation ensure: • monitoring the atmosphere behind the seal from the safe position B 7.21 Fire fighting – When developing procedures for fighting underground mines fires in underground mines, the mine operator should consider: • atmospheric monitoring sites, stations and sampling lines • the location of communication lines and telephones Draft NSW B 4.4 Identify, assess and • identifying critical infrastructure, such as Emergency control communications, power or water, and its planning for emergencies ability to continue functioning in an mines For emergency. public C 6.2 Notifying people at the plan must include the provision of comment the mine and communications systems for how people emergency services are to be notified of an emergency (including those underground and after a loss of power). This may include audible and visual alarms, and public address systems. Redundant (back-up) communication systems are required in underground mines, between the surface and any underground area where people are. C 6.3.3 Communication The emergency plan should address the

Page 33 of 154 ACARP C19010 Extension

Document C/M/B* Clause/ Section Topic Requirement during evacuation communication systems and procedures so it provides: (a) Consistent, timely and clear directions and/or information to control any panic and facilitate an effective, coordinated and orderly evacuation or withdrawal. (c) Ways to positively check that the information has been received, understood and acknowledged by all the people who are affected. For example, the ability to check that all affected people have moved to a safe place or assembly area. B 6.5 Re-entry to any part The mine operator should develop of a mine – during procedures for workers to safely re-enter or after an the mine, or parts of it, after an emergency emergency has occurred or contained (for example, during a rescue). In developing re-entry procedures, the following should be addressed: • obtaining sufficient information on all the hazards and conditions surrounding the emergency situation. This includes the mine and environmental conditions that may have changed as a result of the emergency, thereby introducing new hazards and increased risks (for underground coal mines, see WHS (Mines) Regulations clause 77 Post Monitoring and clause 85(9) Inspection program for more details) • a risk-assessment of all hazards and conditions, which considers: (c) is there sufficient monitoring of hazards that have developed as a result of the emergency? (e) have inspections or remote monitoring been considered before re-entry (for underground coal mines, see WHS (Mines) Regulations references listed above for the dot point in this section of the code)? (i) is there adequate back-up systems? e.g. communication (l) is there adequate communication between key persons and the control centre? * C –Communications M-Monitoring B-Both

Since this project began New South Wales has repealed its legislation and replaced it with new legislation. Both the repealed and replacement legislation are included in Table 2 as they have both applied at stages during this project. The introduction of the new Regulation arguably brought with it the most significant legislative requirement applicable to this project.

The Work Health and Safety (Mines) Regulation 2014 requires the operator of an underground coal mine in NSW to ensure that arrangements are developed and implemented for the monitoring of the atmosphere of the mine following an explosion or fire that leads to the withdrawal of persons from, and the cutting of the supply of power to, all or part of the mine. It includes the need to consider requirements for optimum locations of monitoring points and the design of the post event monitoring system to increase the likelihood of continued operation after an event.

It is likely that any improvements that this requirement brings to the survivability of gas monitoring systems and or post event monitoring options in NSW would flow into standard Queensland practices

Page 34 of 154 ACARP C19010 Extension and improve standards there as well. It is yet to be seen how the mines in NSW address this issue. The approach taken by the mines will be interesting as will that of the Regulator, with regards to what they consider to be acceptable. A review of the legislative and standard/guideline requirements even prior to the new requirements led to the reasonable interpretation that ongoing communications and atmospheric monitoring in both New South Wales and Queensland must be provided for during an incident. There is recognition that the primary system may not survive in which case there should be a backup system. Queensland legislation identifies and classifies part of a mine to be dangerous if that part is affected by the failure or non-operation of the gas monitoring system without a means for backup monitoring.

The for the Mine Tragedy when reviewing communications and legislative requirements reported “In New Zealand there are no express legal requirements for the provision of communications systems in underground mines, either during an emergency or in normal operating conditions. There is guidance in the MinEx code for suitable means of communication to be provided and maintained in specified areas.” In comparison with other jurisdictions they reported “Queensland, by way of contrast, does have express legal requirements for the provision of a telephonic communication system in underground coal mines, including that it have an adequate back- up power supply. It also specifies where the communication devices must be located in the mine. Queensland’s Recognised Standard 08: Conduct of mine emergency exercises includes the requirement for mines to have an effective means of communication with surface control, and specifies their locations. The also requires mines to establish and maintain ‘communication systems to enable assistance escape and rescue operations to be launched.”

This supports the views of the authors that the legislative and standards/guidelines in Australia call for the provision of emergency communications. Although the Royal Commission referred only to Queensland, the legislative requirements in New South Wales are possibly even clearer in setting these requirements. The Coal Mine Health and Safety Act 2002 (NSW) required an emergency management system to adequately address the identification of the equipment, facilities and communication systems necessary to control or limit the consequences of incidents and the arrangement for ensuring that they are readily available. The Regulation (2006) required the emergency management system to adequately address the appointment of competent persons to be on duty on the surface whenever anyone is underground, with effective means of communication to people in the underground parts of the coal operation. The new legislation essentially calls for the same requirements. MDG1020 clearly defines the requirement for systems, both communication and monitoring to aid the escape of workers following an incident.

MDG 1020 is a very useful document when looking at system requirements and has identified issues that must be considered such as systems not surviving the incident. MDG1003 calls for “robust” monitoring systems cable of measuring methane prior, during and after a wind blast event. Implementation of such systems mine wide would increase the chance of survivability following another type of event, not just a wind blast.

Although many electrical standards exist that cover installations they do not cover actual physical installation or level of protection and have therefore not been included.

The legislation and standards/guidelines set expectations but do not offer much information or guidance on how to meet the expectations. A “recognised” or official guideline outlining best practice methods and ways of achieving these expectations would be beneficial to industry.

Page 35 of 154 ACARP C19010 Extension 6.2 Recommendations from disaster investigations

The disasters included below are not exhaustive but are sufficient to demonstrate the repeated calls and need for reliable, effective communication and monitoring data following an incident. Most outline what is required but don’t explain how to achieve it.

Table 3: Disaster recommendations

Investigation Gas Reference Recommendation Monitoring or Comms Box Flat GM 5 (a) That detector equipment for immediate assessment of gases be available at mines. (b) That qualified persons and equipment for rapid analysis of gas samples be available in times of emergency. (c) That there be regular atmospheric sampling, analysis and recording of return airways and sealed areas. (d) That provision be made for atmospheric sampling from fan ducts and sealed areas. Kianga GM 4 All mines have available at short notice the means of analysing the air samples obtained while dealing with an out-break of fire below ground. This end may be accomplished by either mobile laboratories or laboratories established in each mining locality. Moura No. 4 Both 5 Emergency procedures should include detailed instructions concerning identification of the occurrence, evacuation of employees, GM Other Matters f The continued use of the tube bundle system even though some of the tube installation suffered damage could have provided invaluable information concerning the conditions at the time of and immediately after the explosion. The pre explosion information quickly established the absence of a spontaneous combustion development before the explosion and it has demonstrated the need for such equipment to be put into general use in the Queensland coalfields. GM 6 It is recommended that each mines rescue station in Queensland be equipped with sufficient gas analysis and other equipment to enable it to accurately and expeditiously determine the explosibility of a mine air sample. GM 10 It is recommended that an approved continuous monitoring system, capable of automatically determining the composition of the mine atmosphere at pre-determined points in the airway system be required at all underground coal mines in Queensland. Features of this system should include as a minimum standard:— (i) the ability to accurately determine the concentration of methane and carbon monoxide and to record the results at the surface; (ii) monitoring points in all return airways where methane content has exceeded 0.5% (iii) monitoring points in return airways from waste or goaf areas where spontaneous combustion may develop; (iv) monitoring points on main conveyor roadways which are not frequently travelled by men; (v) a continuous recording system at the surface which provides an historical record of gas levels at each monitoring point; (vi) an auxiliary power supply which enables continuity of operation in the event of a power interruption; (vii) the positioning of monitoring points should be determined by the Manager approved by the Inspector and recorded in a book kept for the purpose. Moura No. 2 Comms Emergency Development and provision of portable equipment capable of rapid Warden’s Escape deployment to mine sites to bore a large diameter hole from the Inquiry Facilities surface to reach miners trapped below ground. This would be a means of quickly establishing communication, providing life support and a possible route for emergency recovery of personnel. GM Mine Surface The plan [surface area plan] should also indicate the location of any Facilities surface boreholes that may facilitate the monitoring of the underground atmosphere.

Page 36 of 154 ACARP C19010 Extension

Investigation Gas Reference Recommendation Monitoring or Comms Comms Remote Sensing There appears a need to examine explosion resistant means of and Exploration communication or other means of ascertaining the status of persons remaining underground after an explosion or other catastrophic event at an underground coal mine. Moura No. 2 Both Executive Aided rescue management plans should address: Task Group 4 Summary and • means to establish and monitor the status of persons underground; Report Recommend- • means to establish, monitor and assess conditions in the mine; ations Monitoring, communications systems and other new rescue technology will provide surface personnel with the capability to safely deploy aided rescue measures to rescue those unable to self-escape. Comms Self Escape of Where caches of SCSR’s are provided to enable persons to escape, the Persons caches shall be stored in a changeover/communication station which Underground is serviced and maintained in an operable condition. Consideration should also be given to equipping changeover stations with communication facilities, capable of surviving an incident, to facilitate escape coordination. Changeover/communication stations should be: • constructed to resist low intensity explosions • provided with robust communications facilities to the surface; Both Aided Rescue The components of an Aided Rescue Plan are: • means to establish and monitor the status of persons underground • means to establish, monitor and assess conditions in the mine The inability to locate and communicate with persons affected by an underground incident presents a major hurdle to aided rescue and incident control. It also severely disadvantages those attempting to effect self escape from the mine. Comms Requirements- All persons underground must be immediately alerted to the fact that Warning an incident has occurred and instructed as to the location, and status of the incident and of any immediate response required, e.g. control, standby, evacuate. In some circumstances, it may not be immediately obvious that anything of significance has occurred. Normal communications would in all likelihood have been cut. Emergency warning signals should be fail-safe, minewide and unambiguous. Comms Requirements- The highest priority need is for a communications system which Maintain would survive an incident and provide ongoing two way Communications communications between escaping or trapped miners and rescue personnel on the surface. The system must be compatible with the type of self rescue breathing apparatus to be used and the likely escape of refuge options available to survivors. As power to the mine is likely to be interrupted during an incident, self-contained battery powered backup should be integral to the system. Whilst voice is the highest priority for transfer, systems which can also transmit data and video signals should be encouraged to assist the rescue process. The minimum coverage requirement is for regular nodal communication opportunities along escape routes. Continuous and mine-wide post incident coverage is a desirable, but non-essential objective. Comms Requirements- The location and tracking of all persons (and most vehicles) in Location underground mines is a high priority requirement. Effective two way voice communication will contribute to this requirement but more efficient electronic systems should be pursued. The location and data logging of persons and vehicles immediately prior to an incident should be a priority goal for mine-wide communications networks, even if there is some likelihood that some of the network might not survive a major fire or explosion. The location of individuals need not be exact, but within logical cells (e.g. face areas, specific ventilation splits, transfer points etc.). Comms Requirements- Communications between escaping or trapped miners and people on Condition the surface will afford the opportunity to assess physical and emotional status and allow intervention to be applied.

Page 37 of 154 ACARP C19010 Extension

Investigation Gas Reference Recommendation Monitoring or Comms Comms Requirements- Communications and status systems designs must recognise the harsh Safety and conditions in which operating units will be deployed. Issues impacting Operational on the deployment of new devices include: Issues • range and frequency of repeater stations, aerial requirements • capability to deliver other service functions (eg. monitoring role • reliability and maintenance requirements • diagnostics, fail-safe provisions • battery life • size and shape of terminals worn by persons underground • affordability vs. coverage of workforce vs. capability of units Comms Table 7.1 • develop alert/alarm systems Personnel status • investigate “reverse PED” concept recommend - • develop robust telephone nodes ations • test various escapeway MF radio systems (including vehicle units) • reinforce mine-wide networks • communications between escapees • develop distress beacon Comms Recommendation An industry committee/forum (technical experts and operators) should 7 be established to coordinate the advancement of capabilities to alert, to communicate with, and to assess the status of underground persons during a mine emergency. GM Conditions in Knowledge of conditions in a mine following an incident is essential the Mine in planning any rescue effort. Information systems must be provided to support implementation of the most appropriate rescue measures. GM Recommendation Fixed tube bundles and gas chromatographs should be made available 8 at all mines as the primary method of measuring post incident mine atmospheric conditions. Tube bundle monitoring systems should include: • protection of tubes from damage; • locations of sampling points designed for both normal and post incident atmospheric conditions; • modularisation using boreholes to minimise delay in transmission and analysis as well as to make the system as robust as possible; • techniques for verifying tube integrity which could be routinely applied post incident. GM Recommendation Research into the development of robust telemetric sensors for gas 9 analysis and other environmental parameters, over than (sic) ranges existing after incidents, should be prioritised Both Borehole When planning drilling programs, consideration should be given to monitoring future needs for information, environmental analysis and system systems security for both routine and emergency applications. Key borehole locations could be sited and prepared for fast emergency drilling if required. Refuge chambers may require borehole connection to surface to supply air, monitoring, communications and small supplies. Both Recommendation Both pre-installed and post-incident boreholes should be considered 11 when developing Aided Rescue Management Plans. Guidelines include: • prepare and maintain emergency drilling sites and accesses; • maintaining a list of drillers who could readily provide emergency services; • protocols required to prevent secondary explosions when drilling boreholes in an emergency.

Page 38 of 154 ACARP C19010 Extension

Investigation Gas Reference Recommendation Monitoring or Comms Miner Act Comms Two-Way General Considerations - An alternative to a fully wireless (USA Communication communications system used to meet the requirements of the MINER following System 1. Act for post-accident communication either can be a system used for Sago) day-to-day operations or a stored system used in the event of an accident. Examples of currently available technologies that may be capable of best approximating a fully wireless communications system include, but are not limited to, leaky feeder, wireless or wired node-based systems, and medium frequency systems. Any alternative system generally should: a. Have an untethered device that miners can use to communicate with the surface. The untethered device should be readily accessible to each group of miners working or traveling together and to any individual miner working or traveling alone. b. Provide communication in the form of two-way voice and/or two- way text messages. If used, pre-programmed text messages should be capable of providing information to the surface necessary to determine the status of miners and the conditions in the mine, as well as providing the necessary emergency response information to miners. c. Provide an audible, visual, and/or vibrating alarm that is activated by an incoming signal. The alarm should be distinguishable from the surrounding environment. d. Be capable of sending an emergency message to each of the untethered devices. e. Be installed to prevent interference with blasting circuits and other electrical systems. Comms Two-Way Coverage Area Communication a. The system must provide coverage throughout each working section System 2. in a mine. b. The system also generally should provide continuous coverage along the escapeways and a coverage zone both inby and outby strategic areas of the mine. Strategic areas are those areas where miners are normally required to work or likely to congregate in an emergency and can include belt drives and transfer points, power centers, loading points, refuge alternatives, SCSR caches and other areas identified by the District Manager. While a coverage zone of 200 feet inby and 200 feet outby strategic areas normally should be adequate, the District Manager may require longer or shorter distances given circumstances specific to the mine. c. The District Manager may approve alternative coverage areas to those areas identified in 2(b), such as adjacent entries, for reasons such as radio frequency interference or other factors that may reduce the coverage area at the identified strategic areas. d. Miners should follow an established check-in/check-out procedure or an equivalent procedure when assigned to work in bleeders or other remote areas of the mine that are not provided with communications coverage. e. Communications for refuge alternatives must be provided as required under 30 C.F.R. §75.1600-3.

Page 39 of 154 ACARP C19010 Extension

Investigation Gas Reference Recommendation Monitoring or Comms Comms Two-Way 4. Standby Power for Underground Components and Devices Communication a. Stationary components (infrastructure) generally should be System 4. equipped with a standby power source capable of providing sufficient power to facilitate evacuation and rescue in the event the line power fails or is cut off. In many mining situations, at least 24 hours of standby power based on a 5% transmit time, 5% receive time, and 90% idle time duty cycle (denoted as 5/5/90) should be adequate, but mine- specific conditions may warrant more or less standby power capability. The system should display whether it is operating on-line or with standby power and give an indication of the state of charge of standby power. b. Untethered devices, such as hand-held radios, generally should provide sufficient power to facilitate evacuation and rescue following an accident. In many mining situations, at least 4 hours of operation in addition to the normal shift duration (12-hour minimum total duration) based on a 5/5/90 duty cycle should be adequate, but mine- specific conditions may warrant more or less capability. This total operation time can be achieved via spare portable devices or cached batteries if the device is approved for battery replacement in the hazardous area. Comms Two-Way Surface Considerations Communication a. The surface portion of the communication system generally should System 5. include a line-powered surface component with a standby power source to ensure continued operation in the event the line power is interrupted. Comms Two-Way Survivability Communication a. The post-accident communication system generally should provide System 6. redundant signal pathways to the surface component. The system should display pathway interruptions and system malfunctions. i. Redundancy means that the system can maintain communications with the surface when a single pathway is disrupted. Disruption can include major events in an entry or component failure. ii. Redundancy can be achieved by multiple systems installed in multiple entries, or one system with multiple pathways to the surface; provided that a failure in one system or pathway does not affect the other system or pathway. b. If system components must be installed in areas vulnerable to damage (such as in front of seals), protection against forces that could cause damage should be provided. Another Comms Guidelines 2. Line powered devices must be equipped with a standby power Program being source to allow continued operation in the event the line power is lost Policy Letter administered by during an emergency. The standby power source must be capable of (P11-V-11) the Approval & providing additional operating capacity (24 hours minimum) based on was released Certification a 5% transmit time, 5% receive time and 90% idle time, denoted as by MSHA in Center 5/5/90, duty cycle. April 2011to 3. Untethered communication devices, such as hand-held radios, and establish individually worn/carried tracking devices, such as tracking tags, must approval provide at least 4 hours of operation in addition to the normal shift guidelines for duration (a minimum of 12 hours of operation) based on a 5/5/90 duty communicatio cycle. Additionally, these individually-worn/carried tracking devices n and tracking must provide a low power warning. devices 5. The cable supplying power to the system and all cables between communication and tracking components must be MSHA-approved as flame-resistant or enclosed in MSHA-approved, flame-resistant hose conduit. These cables must be protected from mechanical damage by position, MSHA-approved, flame-resistant hose conduit, metal tubing or troughs. Cables worn by the miner are exempt from these requirements.

Page 40 of 154 ACARP C19010 Extension

Investigation Gas Reference Recommendation Monitoring or Comms Pike River Comms Communication 29. The provision of adequate communication devices, capable of Royal and personnel surviving emergency events such as explosions, ensures that workers Commission location system underground can raise the alarm with the surface and with other Report workers, and can advise surface personnel of the status of workers underground and their planned escape route. A communication system allows those on the surface to guide workers underground towards the best way out of the mine. This is especially important if a particular route could lead workers into danger. Communication and personnel location devices are also beneficial when workers cannot get out of the mine on their own. In such situations, surface rescue teams can be directed straight to the survivors without having to undertake a time- consuming search of the mine. 34. All underground mines should have an adequate communications system that allows effective contact between miners underground and the surface during an emergency. New Zealand should keep abreast of the development of effective personnel location systems. As reliable and suitable systems become available, mine operators should also be required to install these. GM Understanding 39. An underground coal mine’s emergency management system the atmospheric should consider how the atmospheric conditions of the mine will be conditions monitored and understood after an emergency such as a fire or explosion. This means installing a tube bundle system as well as the real-time monitoring system. Both Recommendation To support effective emergency management, operators of 16 underground coal mines should be required to have modern equipment and facilities. • Operators should be required to have equipment and facilities to support self-rescue by workers during an emergency. • Operators should be required to include, in their emergency management plans, provisions for continued monitoring of underground atmospheric conditions during an emergency. Level 1 • The maintenance of gas monitoring after an event should be Exercises completed to include the redundancy of sensors and tubes/borehole back-up for sampling of key areas and communications. • All mines to consider physical separation of critical mines services to avoid loss of all services in a fire, explosion, fall of roof etc. i.e. install separate communication phone, DAC and separate gas monitoring. • Mine consider location of emergency services (e.g. at bottom of downcast shaft) i.e. water, compressed air, communications, gas monitoring etc. • Review managing risks associated with key infrastructure running through the single mine entry (as with the intake drift): use of fire suppression systems or other effective fire fighting technologies in drift, redundancy in communications/gas monitoring systems. • Investigate and implement the use of a borehole to get the dupline and phone underground i.e. separate the 2 services so both systems cannot be so easily destroyed. • Separate the mines services, i.e. separate communication-phone and DAC, separate gas monitoring-real time and tube bundle gas monitoring. • Each mine demonstrates where they have redundancy in the provision of mine services. •Mine design recognised standard for segregation, mines services, etc.

As evidenced above, there are common calls for improved communications and determination of atmospheric and physical conditions underground following investigations of disasters. With the introduction of risk management principles it is unlikely that mines rescue teams will enter a mine without supporting information/data that it is safe to do so. While technology has improved in some areas, the level of survivability envisaged by Moura No. 2 Task Group 4 as likely to emerge has not eventuated in more than 15 years since it was called for. Similar issues were seen after the implementation of the MINER Act in the USA where the available technology did not meet the demands of the legislation, although there have been more developments to meet these demands than the ones from Task Group 4. The lack of increased survivability or hardening, particularly for gas

Page 41 of 154 ACARP C19010 Extension sensors indicates that an immediate hardware solution is unlikely and that developments in providing more robust systems are more likely to come from improved installation standards, positioning and possibly shielding.

Although the Pike River explosion occurred in New Zealand the Royal Commission was undoubtedly influenced by Australian practices and legislation with much evidence and expert opinion provided by Australians and the appointed Commissioner with mining experience was the then Commissioner for Mine Safety and Health for Queensland and Deputy Director –General of the Safety and Health Division of the Queensland Department of Natural Resources and Mines. Multiple Australian organisations were also involved in the response. It is also the most recent inquiry into a disaster in Australasia. For these reasons the findings and recommendations of the Royal Commission are rated as highly significant to the situation in Australia. Like inquiries into explosions previously, recommendations from the Pike River Royal Commission included the need for operators to include provision for continued monitoring of underground atmosphere during an emergency. The Royal Commission also states “All underground mines should have an adequate communications system that allows effective contact between miners underground and the surface during an emergency.”

Additionally the Royal Commission reported that assessment of survivability should begin as soon as possible and that this assessment is an essential requirement for the future. Once established that there are no survivors actions may be required to prevent explosions. In making an assessment on the survivability of workers underground at the time of the explosion at Pike River, factors included were the absence of contact from anyone underground following the explosion, other than one of the survivors, and the toxicity of the gas concentrations present. Both of these factors require effective communication and monitoring systems for assessment following an incident.

Many of the Level 1 Exercises held in Queensland have identified the need for information post event but few of the recommendations relate directly to the survivability of communication and monitoring systems and again little advice is provided on achieving identified outcomes of ongoing operation.

Page 42 of 154 ACARP C19010 Extension 6.3 Case Studies

Table 4 is a summary of observations relating to case studies where communication and monitoring system status has been reported on or where estimations of over pressures or flame fronts have been made. Full commentary on these case studies is included in Appendix A: Case Studies.

Table 4: Case study summary

Mine Topic Observation Pike River Communications Phone 2000m inbye portal exposed to gas velocities well in excess of 108- 2010 252km/h, remained operational and utilised by survivor after first explosion Explosions (good installation standard). Phone was fixed to rib by way of rib bolt (NZ) through the phone’s bracket and secured by a nut. The cable was firmly anchored below the phone so that it could not stress the connector. The cable was then run to the surface predominately in the centre of the roadway (roof) clipped to a catenary wire. DAC still operational following initial explosion. Gas Monitoring Real time gas monitoring system no longer reporting to the surface even though sensors fitted with uninterrupted power supply units. No indication of any abnormal gas results immediately prior to explosion to indicate what could or had happened. Initial response to the was hindered by the lack of available information on the conditions within the mine. No means on site to collect samples. Hand held gas detectors initially used to measure the atmosphere in the fan housing. Stomach pump from ambulance used to collect samples. Samples collected sent via helicopter to mines rescue station for gas chromatograph analysis. External gas chromatograph brought to site for analysis (operational ~20 hours after initial explosion). Due to access difficulties to sample locations, samples continued to be delivered by helicopter. Temporary “plastic” sample lines installed at vent shaft destroyed requiring replacement following each subsequent explosion. No monitoring from vent shaft possible following fourth explosion until flames extinguished. No pre-existing representative boreholes available for sampling. Issues with first (critical) borehole drilled not intersecting workings. Inbuilt pumps on portable equipment not capable of drawing representative samples. Second explosion occurred when typical coal fire products decreasing at vent shaft. Overland tube bundle system installed post explosions with analyser hardware in main administration area. Low density polyethylene sampling tube (typical tube bundle tube) had to be replaced with copper tube because it melted (at least 120oC). Ongoing issues were experienced with tar clogging sample lines. Up to 11.8% carbon monoxide and 7.3% hydrogen measured (GC required). Methane sensor located in the surface évasé at the vent shaft was still in place after 4 explosions. External wiring gland was missing and there was evidence of heat and mechanical damage. When power was supplied to the sensor and methane test mixes applied it was observed that the sensor was still capable of measuring methane although the concentrations were not accurate (results generated prior to explosion also questioned). Physical Initial explosion generated gas velocities between 108-252km/h. Conditions Ventilation control devices destroyed. Sustained temperatures greater than 120oC around underground fire.

Page 43 of 154 ACARP C19010 Extension

Mine Topic Observation Upper Big Communications Communications destroyed. Branch Gas Monitoring Gas monitoring was initially conducted at fans and portals. 2010 Rescue teams withdrawn on several occasions due to explosive gas mixtures Explosion and possible ignition sources. (USA) Concentrations of carbon monoxide encountered exceeded measuring range of hand held gas detectors. Portable gas chromatograph onsite 10 hours after explosion. Mobile laboratory onsite 24 hours after explosion. Drilling of three boreholes planned for representative sampling, first intersected workings 37 hours after explosion. Problems encountered in intersecting workings with borehole. Physical Pressure piling occurred and maximum reflected over pressures up to 105psi Conditions eventuated. Flame zone contained a volume of about 880,000m3. Damage to conveyor belts varied according to the strength of explosion forces and the gauge of the C-channel used in their construction. The greatest damage occurred where they were impacted broadside by explosion forces. For the most part, severe damage was confined to crosscut intersections, and damage quickly lessened away from the intersections where the coal pillars provided shelter from wind forces. Many samples collected returned evidence of the coking to the coal indicating temperatures of at least ~ 370º C widespread. When objects are exposed to flame for a sufficient duration of time, heat is transferred and produces coking. However, even within the area affected by the flame, coking of the coal does not occur at all locations. Sago Communications The pager phone cable was found to be cut or pulled apart, especially where 2006 it traversed crosscuts, exposing it to the apparent forces from the explosion. Explosion The signal line of the trolley phone system was severely damaged in the area (USA) affected by the explosion. There were three communications systems. The paging phone system failed for the 2nd Left Parallel crew that perished. 2-way radios did not fail but distance and line-of-sight limitations would have likely prevented the 2nd Left Parallel crew from using them to communicate with others in the mine. The third means of communication the trolleyphone system was not functional the day of the disaster. Barricaded miners did not try to call out because all of the communication devices were damaged. The explosion damaged wiring and several pager phones. Pager phone communication to 2 North Mains and 2nd Left Parallel was not possible. Gas Monitoring First indication on the surface that something unusual was occurring underground was a carbon monoxide sensor alarming with a measured concentration of 51ppm. Post explosion monitoring data from some carbon monoxide sensors was not available on the surface even though these sensors could be heard alarming underground on the mine phone; indicating the sensor was working but communication with surface compromised. The Atmospheric Monitoring System (AMS) was equipped with a battery backup that maintained power to the system when there was a loss of mine power. The system would remain energised until it was manually disconnected. That was not done at this time, and the AMS remained energised until discovered by mine rescue teams during exploration forcing the withdrawal of teams (AMS not considered IS). Most of MSHA’s incident response equipment and man power was responding to a fire at West Elk mine. Gas concentration measurements were made at mine openings. Variations in reading observed dependent on analyser used. Carbon monoxide off scale on hand held detectors. Sampling required mine rescue team members wearing full apparatus to monitor the gases exiting the mine with results reported to the command centre.

Page 44 of 154 ACARP C19010 Extension

Mine Topic Observation Logs of manual carbon monoxide measurements made during the response were found to be inaccurate and MSHA had to provide training on use of hand held gas detectors. Gas chromatograph onsite and operational 8.5 hours after explosion. MSHA was set-up and monitoring the mine atmosphere exiting No. 1 Drift Opening ~13 hours after the explosion. Communications Preparations were started to drill a borehole into the 2nd Left Parallel section and Gas for sampling and communications purposes. Monitoring Re-entry was delayed due to risk of borehole holing through initiating an explosion. First borehole intersected workings nearly 24 hours after explosion. Physical 4000 m3 of methane consumed by the explosion generating pressures in Conditions excess of 93psi but quickly dissipated in all directions to much lower values. Ventilation control devices destroyed. Willow Communications A Personal Emergency Device (PED) system was in use at the mine, and was Creek instrumental in alerting miners working in active and remote areas of the 2000 mine to evacuate following the initial explosion. These miners all safely Explosions exited the mine. (USA) The communication system remained operational during the emergency with miners able to communicate with the surface using the pager phone on the section. Gas Monitoring Communication failures with many of the sensors surrounding the D-3 section following the first explosion observed on surface, although some remained operational. Elevated carbon monoxide readings occurred in the bleeder entries and in the D-3 No. 1 headgate entry. The monitors near the headgate bleeder connector regulators experienced a communication failure. Data from some operational carbon monoxide off scale (greater than 50 ppm). Useful data collected from others. Gases monitored at mine return portals with hand held gas detectors. Physical Four explosions over 31 minutes. Conditions As little as 1.4m3 of methane diluted to 6.5% was involved in the initial explosion generating pressures estimated at approximately 5psi near the origin, 3psi exiting the headgate into the bleeder entries,2psi reaching the tailgate bleeder regulators and 0.5psi across the longwall face. Forces generated during the second explosion were of a lower overall magnitude than those of the first explosion. The third explosion was the most powerful. As with the first explosion, obstructions prevented the full thrust of the explosion from propagating outbye along the maingate fringe of the goaf. No miners underground recalled the fourth explosion; it was identified by a spike on the fan chart. Endeavour Communications Communications remained operational - surface advised by telephone that 1995 the explosion had occurred. The phone in the crib room was heard ringing Explosion by two disorientated miners that allowed them to establish where they were (NSW) in low visibility. Gas Monitoring Tube bundle system continued to provide results post explosion. Closest sampling point to the 300 Panel was located some 5 km outbye, additional sampling locations identified as required. Boreholes drilled and sampling and analysis conducted with government mobile laboratory. Physical ~ 6 m3 of methane mixed with air to about 6 -7% composition (total volume Conditions of 100m3) generating a peak flame speed of 122 m/s and pressures of up to 4psi in the 300 Panel and about 0.5psi outbye. Significant damage to all brattice stoppings within the panel and also to plasterboard stoppings (in poor condition prior to explosion) in 8 West- 2 km from the 300 Panel. The overpressure was sufficient to raise dust in the 8 West headings. Roof fall leading to the explosion may have generated similar overpressures.

Page 45 of 154 ACARP C19010 Extension

Mine Topic Observation Some of the workers were knocked down by the effects of the explosion. A shuttlecar driver had his helmet blown off and the bracket holding his cap lamp to the helmet was torn off.

Moura No.2 Communications During a telephone conversation between the undermanager (on the surface) Mine and a section electrician (5 South) the phone was cut off simultaneous to the 1994 explosion. There were no survivors from this area. Explosions A worker in 1NW was able to phone the surface from the crib room phone. (Qld) Gas Monitoring The tubes were unprotected but installation was reportedly performed to a high standard. Tube bundle used to collect samples for gas chromatograph analysis. Analysis of the pressure drop within the tube bundle gas sampling lines coupled with the results indicated that some tubes were intact in their original locations whereas others had clearly been damaged and were reporting from completely different points in the mine. This prompted the drilling of four boreholes to ensure gas samples were available from known locations underground. Difficulties were experienced during early borehole sampling operations: • air ingress • methane ingress from overlying coal seams • setting the borehole sampling tube into the pit roadway • Borehole G-512B did not intersect the intended roadway • a borehole attempted for video evidence proved unsuccessful Although not all tubes were sampling from intended locations, results from the system and subsequent gas chromatograph analysis of samples collected via the tube bundle system provided valuable information on the status of the underground environment. This included identification of potential secondary explosion (which eventuated) so rescue teams were not deployed. Carbon monoxide results for some locations were over the maximum measurable concentration for the tube bundle analyser (1000ppm). Of the 10 tubes operating normally prior to the explosion, five were deemed to be unaffected by the explosion. Those closest to the explosion were the ones damaged. During the post incident investigations conducted onsite it was identified that the oxygen paramagnetic analyser may have been damaged/influenced by the second explosion with significant errors associated with measurement that were not apparent prior to or following the initial explosion. One tube was severely pinched and testing confirmed that under these conditions (restricted flow) a sample could be contaminated by gas from another location. Ways of determining whether tube was compromised based on results identified during post incident investigations. Physical The volume of methane estimated to be less than 70 m3. Conditions The pressure wave the crew in 5 South experienced would have been ~4 psi, (first explosion) with the original explosion estimated to only be about 8 psi. Some ventilation controls within 5 South were damaged by the explosion but not elsewhere (other than seals). No damage was done to either fan in the initial explosion, although an explosion relief door was blown about four metres away from the fan housing. Moura No.4 Communications Five men in the 3 South area were advised by telephone to make their way to Mine the surface. 1986 The other three survivors were out of contact but made their way to the Explosion surface. (Qld) Gas Monitoring The tube bundle system was powered by the main electricity supply to the mine which was isolated following damage to the ventilation fan, so stopped working when power was lost. Some of the tube installation had suffered damage (no details available).

Page 46 of 154 ACARP C19010 Extension

Mine Topic Observation Initial measurements using the Mine Rescue analysers from the main fan drift returned results over range for carbon monoxide (greater than 5000ppm), 1.7% methane and 18% oxygen. Gas chromatograph brought to site, results available 11 hours after explosion. Borehole drilled and completed 17 hours after explosion. A decision was made to suspend further recovery until a more reliable sample point could be established at 25 ct 1 Heading. During underground inspection concern arose about the accuracy of sample results received up to that time because a thick bluish smoke and a “fire stink” were detected indicating the existence of an active fire (not identified by monitoring). Physical There was significant damage within the panel but the energy of the Conditions explosion had largely dissipated by the time it had reached the exit to the panel. Over pressures of up to ~13psi estimated. “Taj Mahal” was completely destroyed. Extensive damage to the ventilation system. Substantial damage to the fan ducting, the internal baffles blown 25 metres. Smouldering floor coal and burnt out props. Appin Communications White Panel deputy contacted the surface for assistance from the crib room. 1979 Telephone lines in LW6 went dead. Explosion Gas Monitoring Signs of tube bundle sampling lines burnt. (NSW) Drager tube readings of carbon monoxide at the entry to A Heading returned off scale readings (greater than 3000 ppm). Government mobile laboratory used to monitor the return gases issuing from the main fan évasé. Mine’s own tube bundle was monitoring three underground points. Although more dilute mobile laboratory identified changes more quickly than the mine tube bundle system due to length of tubes and draw times. Physical Electrical engineer 800m away from explosion received 2nd and 3rd degree Conditions burns to his hands. Power lost. Damage to the mine extended well beyond K Panel. The initial explosion created secondary fires. At the intersection with K Panel there was some marlin (bolt rope) burning. Signs of burning on the high tension cable. There was burning in B heading of K Panel around Blue Panel. Some stoppings down. Overcast at the intersection of A Heading of Red Panel and B Heading of K Panel destroyed. The volume of methane that exploded would have been much less than 400 m3 and probably less than 150 m3. Evidence that coal dust was the major component of the explosion. Inanimate objects had been thrown around, but not for any distance. Stone dust barriers had been completely demolished. Much debris – cables and old belt. Steel vent tubes in ribbons, blown down the heading. Fan lying blasted a substantial distance from its original site and overturned. West Communications A telephone was found on the surface having been propelled from Wallsend underground. 1979 Gas Monitoring Mine didn’t have its own tube bundle so NSW Government mobile Explosion laboratory used (located at the upcast shaft). (NSW) Two nylon sampling lines lowered down the shaft. One line had a dust filter fitted which slowly became blocked with dust and the second line had to be used. A light haze became visible at the fan évasé, and the carbon monoxide readings increased suddenly. Within minutes dense black smoke was issuing from the shaft, quickly blocking the nylon sampling lines and delaying the arrival of the gas to the analysers.

Page 47 of 154 ACARP C19010 Extension

Mine Topic Observation The increasing heat in the upcast shaft melted the sample lines and sealed them, preventing further gas analysis, and blocked all sample lines. Physical The explosion was extensive and violent. Conditions Fan elbow blown ~100m. Objects from underground were thrown several hundred feet vertically from the upcast shaft, having travelled underground, negotiated a right-hand bend. A wheelbarrow, a stretcher, an oil drum and a telephone were all found on the surface having been propelled from underground. Underground heavy equipment was moved over substantial distances and some broken. Transformers were moved up to 6m and personnel cars blown off the rails and into the ribs. Timber roof supports were scattered, steel "W" straps were buckled, and conveyor belting was disrupted. Damage extended beyond pit bottom at least a distance of 1800m. The damage in the pit was not uniform but was similar over large areas. From the main north heading over several hundreds of metres the damage did not increase or decrease greatly in severity, indicating not a local but an extensive explosion. Some areas suffered much less damage than others. Mine A Communications Power and communications compromised by fall. Roof Fall Gas Monitoring The tube bundle installed was in two bundles of five core tube (encapsulated in drift in an outer protective sheath). 6 of the 8 tube bundle lines in service that were run in the drift survived. The two that did not survive were squashed and no sample could be drawn through them. The extra resistance on the other tubes did introduce extra leakage for at least one of the tubes. Additional leakage easily identified for low oxygen sample locations not so easy for general body locations. Monitored and logged vacuum pressure invaluable in assessment of tubes. Mine B Communications The fall severed the DAC system line and the 100 paired telephone Roof Fall communications line. in drift Initial contact to the control room was by DAC from the outbye side of the fall. No communications or services were usable on the immediate inbye side of the fall. No telephone communications were possible. Another DAC line running down the belt drift connecting with the other line at the bottom of the drifts. Throughout the incident the DAC was relied on for communications. Gas Monitoring Real time and tube bundle gas monitoring systems remained operational throughout as both are run down a purpose borehole rather than either drift. Crandall Communications Two independent phone systems located in separate entries were run, Canyon however phone communication with the affected section was lost. 2007 Immediately after the outburst, attempts were made to reach the South Coal Barrier section by pager phone with no response. Outburst PEDs were sent to the miner with a receiver on the section to call the surface (USA) without response. It cannot be established if systems had either failed and/or the miners had perished or were otherwise unable to respond. Did not appear to damage the PED loop antenna as the PED was used successfully to communicate with surviving miners and rescuers. Gas Monitoring The real time gas monitoring system reported communication failure from sensors throughout the affected section. In total, seven boreholes were drilled from the surface to the mine workings. Issues reported associated with collecting samples from boreholes including holes initially being blocked. Physical Ventilation structures damaged. Conditions

Page 48 of 154 ACARP C19010 Extension

Mine Topic Observation Alma Mine Communications At least initially, the phones worked after the fire began. Miners at the site of No. 1 the fire travelled 33m to a mine phone to call outside and direct the 2006 Dispatcher/AMS operator to order No. 2 Section (inbye the fire) to evacuate. Fire The Dispatcher called the section on the mine's pager phone. There was no (USA) response. He then activated the signal light on the section pager phone. Again, no response. He then used the AMS computer to remotely stop the North East Mains No. 1 belt. At this point, 5:39 pm, almost 28 minutes after the first sensor alarmed, miners on No. 2 Section used the mine's pager phone to call outside to find out why their belt had stopped. At that point they were told about the fire and were ordered to evacuate. Eventually mine pager phone communication with the longwall face failed due to the fire. The pager mine phone system utilised two wires within the AMS cable bundle. The AMS cable bundle for the carbon monoxide sensors along the longwall belt was installed in the longwall belt entry The cable bundle containing AMS and mine phone circuits were destroyed by the fire. Gas Monitoring The AMS system operated as designed after the fire began, 5:14 pm was when the first AMS CO sensor in the belt entry alarmed and at 5:16 a second sensor alarmed. Eventually (about 45 mins after fire started) AMS communication with carbon monoxide sensors in the No. 9 Headgate longwall belt entry failed due to the fire. Existing borehole used for gas analysis (exhausting). Up to 1700ppm carbon monoxide coming from hole. Carbon monoxide concentrations off scale on mines rescue detectors. Jim Walter Communications The mine communication system was a two-wire pager phone (cable was No. 5 PVC coated, 18.5 gauge, copper covered steel conductor wires). 2001 Phone cable was supported from the roof in the centre of entry. Roof fall The phone closest to the affected area was inoperable after the first followed by explosion. Explosions The phone wire was found under the roof fall. It was damaged by the fall, (USA) explosive forces, and heat from the explosion. It is not known if the damage was from the 1st or 2nd explosion but the phone was inoperable after the first explosion. The phone itself was recovered by the rescue team and functional when tested later by investigators. A phone farther outbye the section was operable and was used to call the surface but the call was interrupted by problems with the phone system. Gas Monitoring The mine had a monitoring system that monitored carbon monoxide. Five minutes after the first explosion, the system showed communication failures for three monitors caused by damage from the forces of the first explosion. Physical Over pressures up to 12 psi were experienced. Conditions Darby Mine Communications Communications were damaged in the area that three miners who died No. 1 attempting to escape were found. 2006 Gas Monitoring Post explosion, gas measurements were made at the fan and mine entrances. Explosion No means of carbon monoxide measurement until MSHA arrived. (USA) Carbon monoxide off scale (>500ppm) sample sent away for analysis came back as 6,162ppm. During rescue operations carbon monoxide concentrations were measured between 80ppm and off scale. Physical Force of the explosion completely destroyed the seals. Conditions Flame limited to the area behind the seals and did not extend into the mine. The personnel carrier that transported two of the victims to the seals where the explosion was initiated was blown outbye nearly 80 metres. Estimated that the minimum pressure exerted on the vehicle was 22 psi. #3 Mine Communications Surface alerted to fire by call from underground and reporting of smoke.

Page 49 of 154 ACARP C19010 Extension

Mine Topic Observation Fairfax Surface attempted to contact the section crews on the mine phone, section Mining Co. crews could hear his call, but their replies were not received on the surface 2002 because fire had already damaged the underground mine phone transmission Fire lines. (USA) One miner was able to communicate with the other working section on mine phone. The communication wiring going to the two active sections passed by the #2- 3NW belt drive (fire site). The wiring included two-way radio antenna wire and the mine phone wires. Insulation on the wires was destroyed and communications were disrupted to the working sections. Gas Monitoring The mine had no gas monitoring system instead relied on a heat detection system for the identification of fires. Physical The flames extended inbye from the #2-3NW Belt head roller for a distance Conditions of 9m, to the takeup unit, and across the full width and height of the entry (4.6m x 2.2m). The heat of the fire melted the plastic light covers on the belt starter box (1.2m from the drive area), charred the cap pieces on the roof bolts (a distance of 9m), and charred wood posts alongside the belt. In addition, heat from the fire melted fire suppression and water supply hoses, communication lines, and electrical cables. Testing indicated that the melting point of the fire suppression hose was 165-172oC. The insulation on the fire sensor wire was melted off; which, when tested, showed a melting point of 185-196oC. Truck Fire Communications A truck fire not far from the portal resulted in the loss of communications. in Gold Communications lost to workers in refuge chamber. Mine Review of incident resulted in running an additional phone line (fire resistant low smoke cable) via a completely separate route to the refuge chamber.

There are enough reported examples where data and communications remained operational after incidents to know that it is possible for systems to survive. There are also many cases where systems are destroyed or are not capable of measuring the concentrations present. Hardware survivability will depend on amongst other things the nature, location and magnitude of the event. As evidenced in the case studies above the magnitude of explosions can vary significantly with some being rather isolated and generating overpressures less than 5psi (e.g. Endeavour) and others can be wide spread generating overpressures greater than 100psi (Upper Big Branch).

The size of the area damaged by the explosion(s) varies enormously; usually larger explosions involve coal dust whereas smaller explosions involve mainly methane. Many of the explosions were quite small and the damage very minor away from the immediate blast zone. Ventilation was often interrupted because the ventilation structures in place had essentially no overpressure resistance. There is no doubt that there is the potential for total destruction close to and within a blast zone. In a number of cases explosions propagated preferentially down one roadway (eg. Moura No. 4) and thus redundant systems may well have survived.

It is obvious from the case studies that the location and positioning of the hardware and routes used to transmit data and samples to the surface also play a major role in the likelihood of ongoing operation. The installation and maintenance standards are also crucial. These are also the factors that the mine can plan and control.

In a number of cases ventilation was able to be re-established to large areas of the mine relatively easily simply by reactivating the mine fan. In other cases it needed the replacement of a few stoppings. In many cases the damage is confined to an active area or panel and the rest of the mine suffers little damage, yet power to the underground is lost or disconnected. Even in areas where damage is significant some structures seem to survive, for example the pipe ranges at Huntly West mine survived a major methane explosion even close to the source of the explosion. Similarly video footage taken in the drift at Pike River indicates that even after four explosions the pipes used for

Page 50 of 154 ACARP C19010 Extension transporting the coal slurry out of the mine remained intact (however evidence of considerable damage in places to the belt structure).

Where equipment has survived it has at times considerably enhanced the capacity of miners to self- escape (Moura No. 2 – the outbye crews were able to use the mine phones to assist in escaping the mine). It thus seems that where equipment is installed to minimise the potential for damaged from flying debris, the chance of survival if outside the blast zone is considerably increased. There are many examples also of miners in sections not affected by the event being alerted to the need to withdraw via communications initiated from the surface. Although it cannot be said with certainty, it is possible that if communication systems had remained operational that following certain events those that were unable to escape and perished may have been able seek and receive advice on the best means escape route (Darby and Sago).

At the Jim Walter No. 5 incident where there were multiple explosions, investigators suggested that the underground supervisor from No. 4 Section who phoned the surface after the first explosion may have ordered a total mine shutdown if the phone call had not been interrupted because of problems resulting from the explosion. If this had happened it is possible that subsequent explosions may have been prevented. The phone wire was found under a roof fall and was damaged by the fall, explosive forces, and heat from the explosion. However, the phone itself was recovered by the rescue team and found to be functional when tested later by investigators. This indicates that, in some circumstances, equipment cabling may be more prone to damage than the components attached to them. Cables are exposed to much greater lengths of mine openings and therefore may be more likely to be subjected to damaging forces. The fact that once repowered the methane sensor that had been located in the ventilation shaft at Pike River was able to detect methane shows that the sensors themselves may be more resilient than previously thought.

In some cases even when equipment is damaged it can still offer valuable information relating to what is happening in the mine atmosphere post explosion. This is particularly the case with tube bundle as evidenced at Moura No.2 where although it was noted that some tubes were damaged and others weren’t, the information was sufficient to determine that conditions were not safe enough to allow deployment of rescue teams. Post incident investigation also showed that there are ways to determine whether tube integrity has been compromised and some ability to determine where tubes may be sampling from. Using this information during the incident response may increase the value of the available information.

There are cases where rescue/re-entry was undertaken with only an imperfect idea of what was happening in the underground environment (for example at Sago and Upper Big Branch mines). This is generally based on a lack of available data. An extreme example of this is the response to an explosion at Raspadskaya mine in Western Siberia, where a second explosion occurred only hours after the first and rescue teams were killed. In a number of cases explosions were initially reported as wind-blasts from goaf falls, thus misleading responders/rescuers and underplaying the danger of re- entry (for example Moura No. 4).

If provisions are not in place for continued gas monitoring and backup monitoring plans are based on drilling boreholes from the surface, the time taken to establish adequate gas monitoring might prevent any rescue of miners trapped underground from taking place. Although all deaths are believed to have occurred at the time of the initial explosions, the time taken at Moura No.2 and Pike River, to obtain gas samples by means of new boreholes were significant in a situation where response time is critical. The lack of data relating to the underground atmosphere delayed decision making on whether it was safe to re-enter the mine. It is difficult to speculate whether the availability of gas data immediately after the explosion would have provided information that resulted in a different decision as to whether or not to enter the mine but under different circumstances it may be the difference between getting to a survivor in time or not. An extreme example of the problems associated with delays from boreholes is at Sago, where rescue teams were actually kept out of the mine for a borehole to intersect workings (due to risk of this initiating another explosion) delaying rescue efforts by hours and when teams did reach victims, only one was still alive, others had succumbed to carbon monoxide poisoning. It is not

Page 51 of 154 ACARP C19010 Extension known whether the delay was significant with regards to survival of victims, but does highlight the extreme importance of time.

At Pike River, West Wallsend and although not detailed above Blair Athol, problems have been experienced with tars from coal fires clogging sample lines. This is a difficult problem to overcome if the sample lines are preinstalled underground as often the only effective solution is replacement of the lines.

The Pike River Royal Commission report details the need for the incident team to make an assessment on the survivability of persons unaccounted for. As in the case of Crandall Canyon without robust communication systems the failure of underground personnel to initiate or respond to communications with the surface could be as a result of either failure of the communication system or miners have not survived the event (or are unable to get to communications). Being able to determine the status of the underground environment is also critical in making survivability determinations. Post event investigations indicated that the atmospheric conditions inbye the outburst were oxygen deficient and would not have been able to support life. If this was known with certainty the rescue efforts which led to further deaths may have instead been deemed recovery operations and the approach taken may have been more conservative and further lives not lost.

Although not included in detail in this section, all of the research team have been involved in events, details of which are not published or released into the public domain, in which no lives were lost but following evacuation from the mine, regulators have enforced the need for additional monitoring locations prior to re-entry to the mine was allowed. In these cases the time spent on the surface can be measured in days if not weeks. By having established representative means of sampling this time could be reduced significantly depending on the status of the underground environment. The importance and issues associated with maintaining power to keep real-time monitoring operational have also been apparent.

Upper Big Branch may appear as a good example of rescue operations proceeding without reliance on gas monitoring or communications surviving an explosion; instead relying on post explosion implemented systems/means. It must however be noted that although very little is mentioned elsewhere in the reports on this explosion, McAteer et al (2011) report that “there was one large movement of air in the Headgate 22 section. Investigators were unable to determine what caused the event. Although there were no injuries or deaths, the potential existed for disastrous consequences. There were large numbers of mine rescue teams underground, but they were not backed up by an equal number of teams on the surface.” There were multiple occasions when rescue teams where underground and explosive (or close to) gas mixtures and evidence of ongoing fires were identified and rescue teams had to be withdrawn. Withdrawal generally took hours, leaving rescue teams exposed to the risk of an explosion.

6.4 Conditions likely to be Experienced During and After an Event

Table 5 provides a summary of the conditions and consequences of the most common disaster events. These events are discussed in more detail in Appendix B: Conditions likely to be experienced during and after an event.

Page 52 of 154 ACARP C19010 Extension

Table 5: Likely conditions and consequences of events

Event Likely Conditions Consequences Explosion • Typically overpressures up to a • Displacement-Projectiles maximum of 30psi in non-sealed • Residual fires areas but with pressure piling and reflected pressure waves can increase (up to 105psi was estimated at Upper Big Branch) • Higher pressures possible in sealed areas due to confinement (Sago estimated as >93psi but up to 640psi possible) • Exposure to temperatures greater than 1000oC for milliseconds

Fires • Surface temperature of burning • Temperature limits for typical plastic coal can reach 900°C materials used for wire and cable • Sustained temperature- can heat insulation, or plastic conduit, are in surrounding atmosphere the 120-200°C range over several • For large but localised fires roof hours temperatures can exceed 500ºC • Typical polyethylene tube bundle • For smaller fires roof temperatures sample tube melts can exceed 260ºC • Hardware melts • Lead melts (327.5oC) • Zinc melts (419.5 °C) • Tin melts (231.9 °C) • Copper OK melts >1000oC

Roof Falls • Windblast • Physical damage including VCDs ‐ Peak wind blast velocity-123m/s • Displacement-Projectiles ‐ Maximum air flow distance- 197m • Hardware and transmission media ‐ Peak over pressure- 35kPa (5.1psi) buried/compromised ‐ Impulse-89kPa.s (12.9psi.s) • Sensors requiring oxygen for • Low oxygen environment operation may not function as expected Outbursts/Pillar • Low oxygen environment • Sensors requiring oxygen for failures operation may not function as expected • Physical damage including VCDs • Hardware and transmission media buried/compromised

Inundation • Flooding • Hardware and transmission media • Pressures of over 100 psi for severe submerged inundations with deep water • Sensors unlikely to operate • Communications may be compromised • Displacement-Projectiles • Wireless communications unlikely to operate through flooded areas • Flooded roadways block RF signals

Page 53 of 154 ACARP C19010 Extension 6.5 Current status of existing systems

Mines in both New South Wales and Queensland were sent a survey (Appendix C: Copy of mine survey) aimed to determine the communication and monitoring systems deployed in their mine and if protection of same was considered. Response to the survey was much better in NSW with 20 out of the 30 mines sent the survey replying. Only four mines in Queensland responded. The information gathered from this survey was not as informative as hoped. There appears to have been some confusion over what protection was being questioned. Those surveyed took it to mean anything from electronic noise suppression to FRAS covers on tube bundles.

After reviewing the responses and determining that the information was limited as a result of the format of the survey the research team took the opportunity to assess underground installations and personally question mine workers.

The general conclusion that the research team came to was that monitoring and communications systems are installed with day to day operational requirements considered without consideration of maximising survivability during or following an incident or event. This doesn’t necessarily mean that installation standards are poor they just don’t reflect what may be required to ensure they are operable following an event. This is supported by the comments included in the report by Macpherson (2010) that loss of such systems appeared not to be considered in emergency response plans.

Considerable variation in the standards applied at different mines is obvious, variation within the same mine also occurs. The lack of a standard or guideline to follow may contribute to the standard observe. Many mines questioned didn’t have documented procedures for tube bundle sample lines. Installation practices are biased towards convenience and speed.

To demonstrate the variation and at times lack of consideration to survivability photos from underground installations have been included, with some discussion on how exposure to an event may affect survivability.

Figure 5 shows an installation where real-time sensors are connected to a pogo stick and little support/securing of the cables is apparent. This may be fine during normal operations however it is not expected that such an installation would remain in place if exposed to an overpressure event. If the pogo stick was dislodged and fell, the sensors if still working would be measuring the atmospheric conditions on the floor which may not be representative of the true atmosphere (combustion products tend to be found near the roof). The sensors may not operate correctly if they fell in mud or water.

Page 54 of 154 ACARP C19010 Extension

Figure 5: Real-time sensors installed on pogo stick

Similarly Figure 6 shows suboptimal real-time sensors installations that although are adequate during normal operations, are unlikely to withstand an overpressure event , during which they are likely to be propelled away from installation site and possibly separated from cable. There is no evidence of cable support/fixing in these installation either. Figure 7 demonstrates a better installation standard, although additional cable fixing particularly close to the sensor may further improve survivability.

Figure 6: Real-time sensors installed using rope

Page 55 of 154 ACARP C19010 Extension

Figure 7: Better installation standard

Page 56 of 154 ACARP C19010 Extension

Figure 8 shows examples of communication hardware unlikely to survive an overpressure event based on the method of fixing the hardware to the rib and lack of securing the connecting cable. In several of the photos the hardware can be seen hung by rope over a rib bolt. A more robust means of installation could be achieved by fixing the hardware to a backing plate with a hole through which the rib bolt could be inserted and a nut used to tighten the backing plate to the rib. This was the method used to secure the phone that survived and was used by one of the survivors close to pit bottom at Pike River.

Figure 8: Communication hardware unlikely to survive overpressure event

Many mines make use of self draining water traps at seal sites. Two examples of these are given in Figure 9 (the self draining catch pot is not in frame for the example on the right). These self draining water traps rely on the container being full and the resistance between the trap and the sampling point being less than the head of water that would need to be lifted if sampling was to be through the drain line. In the water cooler set up the water that drains out of the water trap down the drain line and just overflows out of the top of the cooler. The example on the right has a tube fitted inside that terminates near the bottom of the catch pot and exits out the side at the top. The drain tube ends below the exit at the top. Although the example on the left is functional (and a much cheaper alternative) if it was exposed to an over pressure it is likely that the water cooler would be displaced and the drain line no longer submerged. This would allow sampling from the drain line. This could result in elevated oxygen concentrations for a sealed area and decision makers may make an incorrect assumption that the seal was compromised (little change to vacuum pressures or flow rates would be observed to indicate damaged tube). The purchased setup on the right allows the whole system to be adequately secured and the drain tube is fixed to the panel at multiple locations. Typically the panels would be fixed to the rib, but in this case it was secured to a wooden crib/cog. Hard to make out in the photo but the panel is also secured at multiple locations along its length. The installation on the right is much more likely to survive an over pressure event than the one on the left.

Page 57 of 154 ACARP C19010 Extension

Figure 9: Tube bundle seal site water traps

Figure 10 shows an example of poor handling of tube bundle. This tube even during normal operation is likely to kink, but even more so if exposed to an over pressure. Ideally the excess tube should be looped and secured in multiple locations and then firmly secured in multiple locations to a fixed structure.

Figure 10: Example of tube likely to kink

Page 58 of 154 ACARP C19010 Extension

Figure 11 shows two angles and frames of the same location underground where different installation standards are observable in the one place. One of the installations is secured well against the roof with no sags. Another is secured regularly but allowed to sag somewhat between fix points and others are sagging too much. If exposed to overpressures sagging tubes tend to move more and also can be whipped into the roof and damaged.

Figure 11: Variations in installation standards and damage potential in same area

Page 59 of 154 ACARP C19010 Extension

The tube installed through the timber cog shown in Figure 12 is very prone to damage (even in normal operation). It is not advisable to run tubes through possible pinch points.

Figure 12: Tube likely to be damaged or crimped

The green tube joining the blue tube shown in Figure 13 would be susceptible to damage if exposed to an over pressure event. The tube is not secured and just allowed to fall vertically; it is also entwined in other infrastructure, increasing the chances of damage. In this case a better option would have been to run the tube over to the rib (secured to the roof) and then run to where it needs to be (again with regular securing points). Having the tube join (a point of possible weakness) hanging like that is also not advisable as any overpressure would pull on it possibly compromising it. The join would be better located in a horizontal run of the tube.

Figure 13: Tube (green joined to blue) susceptible to damage

Page 60 of 154 ACARP C19010 Extension

The tube installation shown in Figure 14 is of a high standard with the tube secured well to the roof and minimum sags. It must be noted that the service (assumed to be an inertisation line) is less likely to survive an over pressure event based on it being secured with hooks (circled in red).

Figure 14: Use of hooks and cable ties

Figure 15, Figure 16, Figure 17 and Figure 18 are examples of good tube/cable installation, pulled tight close to the roof with minimal sag and regular fixing points.

One thing that has been noted is that mines tend to afford significantly more protection to tube runs on the surface where they need to run either from the portal or boreholes to the analyser housing than they do underground. This is typically by way of trenching, running through conduit or in covered cable trays, armoured tube or any combination of these. The influencing factor may be issues with the UV stability of the tube.

Page 61 of 154 ACARP C19010 Extension

Figure 15: Good installation standard

Figure 16: Good installation standard

Page 62 of 154 ACARP C19010 Extension

Figure 17: Good installation standard

Page 63 of 154 ACARP C19010 Extension

Figure 18: Good installation standard

Page 64 of 154 ACARP C19010 Extension 6.6 Status in Other Countries

Communications with international research, statutory and mines rescue organisations by the project team revealed similar issues and concerns with the survivability of critical data and communication infrastructure during or following an event. There has been no evidence that any country has found a total solution. The approach of most is to make use of mine openings and bore holes for monitoring where possible. Apart from the use of armoured tube bundle tube in the UK (and even then really only for protection in areas where mechanical damage during normal operation was likely such as in shafts) international attention seems to have focused more on communications rather than gas monitoring. The difference in approach may be attributed to the fact that Australia leads the world in gas monitoring practices during normal operations.

The US has led the way with post event communication system research. Following the disaster at Sago the US introduced the MINER Act that called for personnel tracking and a post-accident communication system between underground personnel and surface personnel via a wireless two-way medium. However fully wireless systems were not available in time to meet the requirements and MSHA had to provide guidance on alternatives to fully wireless post-accident two-way communications. This guidance really didn’t explain how to achieve the objective rather just what had to be achieved. It must be noted that there are now complying communication systems, a list of which can be found at; http://www.msha.gov/techsupp/PEDLocating/CommoandTrackingMINERActCompliant.pdf.

The wireless systems required under the MINER act often make use of nodes that if configured with redundancy and survivability in mind can continue to operate if a node is destroyed as they are able to “re-route” the communication via an alternative pathway not using that node. Private correspondence with the project team have revealed that there are still some uncertainties as to the survivability of these systems as they exist but work is continuing to improve them.

The US Bureau of Mines engaged Arthur D. Little Inc. in the early 1980s to undertake a study and report on buried communication cable for underground mines (Lagace and Mouossa 1982). The objective was to determine the desirability/practicality of burying communication cables to protect them from the effects of fires and explosions. The likelihood of cable involvement in a mine fire or explosion, the extent of cable involvement, likelihood of loss of communication in terms of added risk to the miners were all examined.

It was reported that cable survival during explosions is a “moot point” as the miners in the area of the explosion are invariably killed, if not instantly within a few minutes, leaving no one to use the survivable communications systems. It also identified that if the explosion is methane only the cables appear to survive even in the explosion area, but violent dust explosions are “highly destructive to everyone and everything, including cables, in the explosion area and beyond. Although beyond the working section containing the explosion, the cables appear to survive.”

The report concluded that “the overall desirability/practicality of cable burial must be regarded at this time as highly questionable, because of its benefit limitations to special scenarios, its practical operational shortcomings and the availability of alternative cable protection options. Suggested alternatives are covered in the Options for increasing survivability section of this report.

The Alpha Foundation for the Improvement of Mine Safety and Health, Inc. (Foundation) was established as part of a Non-Prosecution Agreement (Agreement) entered in December 2011 by the United States Attorney's Office for the Southern District of West Virginia, the United States Department of Justice and Alpha Natural Resources, Inc. ("Alpha") and Alpha Appalachia Holdings, Inc. This Agreement was related to the explosion at Upper Big Branch Mine, an underground coal mine owned by Performance Coal Company at the time of the explosion and acquired by Alpha in June 2011. As part of that agreement Alpha has established a trust to fund projects designed to improve mine health and safety by providing US$48,000,000 into the trust.

Page 65 of 154 ACARP C19010 Extension

A research meeting was held in October 2012 to discuss priorities and issues in mine safety and health research with the broad research community, and to solicit recommendations for attracting participation and projects from outstanding researchers and organizations. The issue of survivability was mentioned by multiple attendees and organisations represented. Joseph Main, Assistant Secretary, MSHA stated “Very important work is also underway to tackle other technologies to improve mine safety and health that also needs more attention. This work includes: development of improved atmospheric monitoring systems that would provide instant information at strategic locations during normal mining operations, and be operational during a mine emergency following a mine event such as a mine fire, explosion, inundation or ground failure; development of improved communication systems, such as through-the-earth voice communications, and communication systems that will link mine rescue teams directly with the control and command center; electronic systems that will provide instant information on mine rescue explorations; improving seismic capabilities to better detect miners who may be trapped or lost underground; and building upon the latest breakthrough of through-the-earth voice technology are other issues that need more attention.”

The United Mine Workers of America’s representative identified the need for post –accident evaluation of the mine environment, including the potential use of AMS (Atmospheric Monitoring Systems) and tube bundle systems.

A session on disaster prevention and response was included and all three presenters referred to the need for system survivability. The need to develop methods to measure or quantify the survivability was also identified. This is an important matter if mines are going to rely on hardened systems as without some quantification of survivability what was thought to be hardware that would survive an incident may not actually survive.

The Office of Mine Safety and Health Research at NIOSH asked the Board on Human-Systems Integration at the National Research Council to appoint a committee to examine and report on the essential components of self-escape. The Committee on Mine Safety: Essential Components of Self- Escape (2013) considered environmental and human-systems factors as well as technologies to propose ways to improve self-escape and identify knowledge gaps where further research is needed. The report identified that although progress had been made on improving the robustness of communications during and following events, technology needs and gaps persist as per below.

Communication Technology Needs and Gaps

Integrated Primary and Secondary Communication Systems to Improve System Survivability

 Mine-specific modeling and simulation tools  Improved modeling of the communication links  Better understanding of secondary systems  Shared definition and quantitative measure of survivability  Develop mine specific modeling tools to be able to assess survivability for:

‐ Any mine configuration ‐ Any installed communication & tracking technology or combinations of technologies ‐ Various types of disasters in various locations within the mine ‐ Various location of miners

Primary Systems

 Similar to conventional radio handsets  Use small antennas  Wearable devices  Long battery life  Sufficient throughput for general operations

Page 66 of 154 ACARP C19010 Extension

Secondary Systems

 Unconventional radios  Unconventional signal propagation  Require large antennas (not wearable)  Typically one channel (very low throughput)  Likely more survivable

Following the MINER Act NIOSH commissioned Foster-Miller to develop a practical methodology to model and calculate survivability and reliability for communication, tracking, atmospheric monitoring and power systems. The project also included research and development into hardening and redundancy techniques for these systems. The information from the Foster-Miller report (2009) is included in the Options for increasing survivability section of this report.

The underground coal industry in the United Kingdom has a long history of using tube bundle systems. Information available indicates that they make more use of protected or armoured tube bundle lines (detailed in the Options for increasing survivability section of this report). Personal correspondence with UK personnel involved with tube bundle systems tends to indicate that the use had declined and was primarily used for protection against mechanical damage during normal operations in high traffic areas and those particularly prone to damage such as shafts. As reported by the Pike River Royal Commission the United Kingdom also requires mines to establish and maintain “communication systems to enable assistance escape and rescue operations to be launched” however how this is to be achieved is not detailed.

6.7 Other Industries

Contact was made with metaliferous mines and metaliferous inspectors but information provided demonstrated little attention was given to harden systems other than the overseas gold mine that introduced fire resistant cables for communications with refuge chambers (detailed in case studies). What is clear is that the underground coal mining industry is unique in relation to the hazards it faces and the physical conditions in which it is conducted. The combination of the hazards faced and the conditions that monitoring systems are installed are unique, with no other industries required to keep equipment and systems operating in and after such circumstances. These problems and a limited market to supply to are likely factors that have limited research and development in this area.

Of note is the use that other industries make of trenching for running cables, particularly telecommunication providers. Many organisations are making use of buried fibre optic cable for the transmission of data, often over long distances. It is typical for electrical, communication, gas and water supply networks to be run underground to provide protection and integrity to the system. Other heavy industries make use of armoured or shielded cables, but typically the distances these need to be run doesn’t compare with underground coal mines. Running cables through conduit of various dimensions and strengths for protection is standard in many industries. Cable trays can also be used as means of protection as well as to improve housekeeping.

Another industry with similar issues to underground coal mines is the petroleum and gas industry. It is clear that the actual setup and installation of gas monitoring systems at petroleum and gas sites is generally to a higher standard than those systems installed in underground coal mines even though the hardware is very similar. This could be due to the fact they do not have the same work environment as an underground coal mine. This is no excuse however for poor installations.

A significant amount of the gas monitoring equipment used in the petroleum industry is installed in areas that are open to the atmosphere, very different from the confinement of a coal mine, and distances involved are not as great. This affords the opportunity to use armoured or protected cabling,

Page 67 of 154 ACARP C19010 Extension or the use of cable trays or protective conduit. Use is made of fibreglass re-enforced plastic conduit in petro and chemical plants because of its strength and fire resistance.

Mounting of sensors can either be a simple mounting on a bracket or with a protective back plate on more robust installations. While underground coal mines could probably argue similar sensor mounting techniques, it is more the consideration to positioning and connection that sets apart the petroleum industry systems.

The petroleum industry doesn’t face the same issues in providing more than one power supply system to their gas monitoring installations if required. Their exclusion zones for land based installations are relatively close to the refinery infrastructures themselves, which makes auxiliary power to contingency monitoring systems much easier to accommodate. It is different on oil platforms where they are under the same explosion protection constraints as underground coal mines and also have their SIL/SIS to comply with, but distances involved and areas to be covered are much less, significantly reducing cost implications.

On the 4th May 1982 the HMS Sheffield was hit by an Exocet missile while on active duty during the Falklands War. The hit caused significant damage to the Sheffield and a fire broke out causing damage to fire control systems. As a result of the compromised fire detection and fighting systems reviews and research were initiated by the United Kingdom Defence Force to find alternatives to the aluminium cabling that had been employed on the Sheffield. The researchers of this project have been advised by a fire engineer involved in its development, that this work ultimately lead to the development of AS/NZS 3013:2005 –Electrical installations-Classification of the fire and mechanical performance of wiring system elements. This standard sets the requirements for emergency wiring systems, not only specifying the cable, but also the cable support system. The standard sets out a classification system for wiring system elements based on the system’s ability to maintain circuit integrity under fire conditions for a specified period and maintain circuit integrity against mechanical damage of specified severity. The standard is referenced in the Building Code of Australia. The application of cables conforming to this standard in buildings to ensure that fire safety systems are operational at the time most needed is obviously a different environment to underground coal mines with lengths of cable deployed significantly less in buildings. It does however offer a way of mines looking for a solution to quantify the “survivability” of such cables under fire conditions. Fire resistant cable was the solution chosen by an overseas gold mine that had experienced lost communications with refuge chambers during a fire when looking for an effective control to prevent a reoccurrence.

The Brisbane Airport Link Project made use of over 1,100km of a plastic cable that when exposed to heat forms a protective insulating coating capable of withstanding over 1000oC and allowing current to flow. The cable was designed to maintain the integrity and continuity of circuits for building safety systems such as emergency lighting, alarms, pumps and fans that are vital for safe evacuation and firefighting.

6.8 Discussions with Manufacturers and Suppliers

Meetings were held with major suppliers of gas monitoring and communications systems to discuss what possibilities there were for improvement relating to survivability of systems. What is apparent is that the underground coal mining industry is only a small market for these systems and research and development of products is limited by potential for return on investment. Although there are some improvements that could be seen in coming years, such improvements are commercially sensitive; they are also not likely to be available for some time.

The lack of any hardware solution likely to become available in the next few years necessitates the investigation of ways of protecting or shielding vulnerable hardware and applying retro fitting of same.

On a more promising note several of the manufacturers/suppliers indicated that they had solutions to the issue identified during the industry risk assessment relating to the provision of independent power

Page 68 of 154 ACARP C19010 Extension to monitoring and communication systems that would soon be commercially available, if not already available. Again some of the details remain commercially sensitive until products are released.

Some of the options available or on the horizon include increased capacity battery backup with an ability to turn on and off on an as needs basis from the surface which will extend the operating time frame that way. There is also work being done on developing lower power consumption gas detectors. Discussions also included making use of solar powered recharging systems connected to hardware underground via boreholes.

What all the manufacturer/suppliers were clear about was that regardless of the time that alternative power could be supplied, if the signal transmission medium (e.g. cables) was compromised then provision of power becomes irrelevant, they saw a need for guidance or standards for installation.

When asked about the influence shielding may have on underground gas detectors, it was suggested by suppliers that shielding may actually improve the operation of the sensors even during normal mining operations particularly in areas of high ventilation velocities.

Several of the suppliers when questioned about adding the functionality to determine if mine phones were still operating post event, responded that the capability already existed it was just never a feature anybody ever asked about. This means that a mine can establish whether a phone is ringing or not underground and thereby assisting in the determination of the status of underground personnel.

6.9 Options for Increasing Survivability

6.9.1 Introduction

Recommendations from the Moura No. 2 Task Group 4 were very clear in the need to develop communication and gas monitoring systems that could survive an incident. Little has been done since these recommendations were made to achieve this.

NIOSH commissioned a study by Foster-Miller (2009) on system reliability and environmental survivability including hardening and redundancy techniques. As part of the study interviews were conducted with mine personnel, mines rescue teams and government agencies. Of note interviewees considered hardening of communication equipment as having merit but thought it unlikely that any hardening technique could protect equipment close to the source a significant explosion. It was also noted in this report that while components could be protected using blast shields and other protective enclosures, much of the feedback focused on the challenge of protecting cables from damage. This is supported by similar findings during this project by the research team.

Many of the hardening techniques which received more support involved using equipment and methods already available in the mine (recessing, installing cable tautly and in a straight run, and using redundancy) as opposed to techniques which require extra equipment, maintenance and resources (such as trenching).

This section does not include results and findings from testing approved as part of this project which have been reported separately in Part B of this report.

A table summarising the techniques reviewed for increasing the survivability of communication and gas monitoring systems can be found in the Executive Summary of this report.

6.9.2 Cable/Tube

There is plenty of evidence that cabling and connection of the cable to the hardware may be more prone to damage than the components they connect (e.g. Pike River and Sago gas monitoring systems) and although the end of line equipment may be functional, transmission to the surface is compromised.

Page 69 of 154 ACARP C19010 Extension

The importance of maintaining transmission to and from the surface warrants reviewing hardening techniques specific to cables. Current practice would indicate that little consideration is given to protection of cable in an event. This may be as a result of a perception that what is on the end of the line will no longer be operational so why protect the cable, but this perception must now be questioned in light of specific examples already mentioned. Much of the referenced material relates to the protection of cable but is often equally applicable to tube bundle lines.

6.9.2.1 Trenching (Floor) Trenching is used extensively in other industries, particularly in the provision of utility services. However as reported above a study into trenching by Lagace & Moussa (1982) concluded that “the overall desirability/practicality of cable burial must be regarded at this time as highly questionable, because of its benefit limitations to special scenarios, its practical operational shortcomings and the availability of alternative cable protection options. Alternative solutions to trenching cables suggested were;  Cable loop-back through parallel tunnels or other passages.  Repositioning of cables away from immediate roof area where fires tend to spread more rapidly to a more favourable rib location in the tunnel.  Use of heavier duty or better insulated communication cable.  Covering the cable on the floor of the tunnel instead of burying it.

It must be noted that the assessment of trenching by Lagace and Moussa was done in the US in 1982 and the pros at that time may not reflect what they are in the current Australian industry. It may now be a case that the pros outweigh the cons.

Trenching may well be the preferred hardening option for areas in which an extremely high level of protection is required e.g. risk of severe roof falls or heavy blast damage is thought to be highest. Other less vulnerable areas of the communication network may make use of other means of protection. Foster-Miller have extensive experience in trenching techniques in the utilities industry (gas and electrical) and recommend use of the smallest practical trench dimensions width and depth (Figure 19). This reduces cost, improves access and minimizes floor disturbance.

Manufacturers were asked by Foster-Miller whether they would recommend recessing equipment into the rib or the roof. Most recommended burying the power cables in the floor because it is more attractive because it is always stable whereas rib and roof can be subject to spalling, especially close to the face (the lack of consideration to floor heave needs to be noted).

Figure 19: Typical Trench Configuration Proposed by Foster-Miller

Page 70 of 154 ACARP C19010 Extension

Einicke et al (2011) reported on research into providing communication networks for underground coal mines that could operate following an explosion, major fire, roof collapse or water inundation without relying on internal batteries or underground power. They proposed the use of trenched power over fibre technology (fibre optic cable discussed further on) buried to a depth of 0.6m.

Trenching arguably offers the greatest level of protection. Advantages of trenching include isolation/minimisation of exposure to elevated temperatures during fires, avoidance of damage from stretching during roof falls or rib spall, maximum protection from debris during explosions. Some trenching is already done for drainage purposes so equipment is available.

A major disadvantage of trenching is that cables still need to connect to hardware outside the trench and this may be the most vulnerable part of the cabling. Other issues include flooding of trenches so all components need to be water proof and corrosion resistant, issues with RF and leaky feeder signal strength if buried, connections and test ports require access, the added complications of the installation process itself, a channel needs to be cut, backfilling required, issues where floor movement problematic eliminating some of the benefits, possibility of introducing floor stability problems and the actual positioning of the trench with consideration to above ground connections and traffic (close to rib preferable).

6.9.2.2 Recessing (Roof and Rib) Similar to trenching in the floor, recessing in the roof and rib provides a high level of protection against explosion forces and flying debris. Everyday protection against physical damage from machinery, vehicles and equipment is also afforded by this technique. As mentioned in the section on trenching many manufactures prefer the option of trenching to recessing in rib or roof due to spalling issues. Recessing would however require additional expenditure and installation would be more difficult. Recessing cables in the roof would involve something similar to trenching, with the trench cut into the roof not the floor. An example of rib mounted cable proposed by Foster-Miller is seen in Figure 20. Although well protected in the slot, there are issues with getting the cables across intersections, a place identified as where cables are vulnerable to damage. Other issues anticipated with this technique would be access to any cut out in the roof through roof mesh and the effect it might have on roof stability. The applicability of this technique would depend on roof and/or rib conditions.

Figure 20: Example of Recessing Cable in Rib as Proposed by Foster-Miller

6.9.2.3 Cable/Tube Positioning/Location Manufacturers consulted by Foster-Miller on the best means to harden systems recommended that cables be installed on the upper rib area rather than the roof. This was because they considered the upper rib area more protected, and that there is greater chance for accidental damage from vehicles on

Page 71 of 154 ACARP C19010 Extension the roof. Suggested mounting location from the consultation process for leaky feeder cable is shown in Figure 21.

Figure 21: Suggested Leaky Feeder Cable Mounting Location

The pressure in a blast wave is lower at the roof or the rib as compared to the centre of the roadway. Due to the difference in pressure levels at the surface of the roof or rib versus a few inches away, the location of the installed cable even if encased in conduit will make a difference in the ability to survive a pressure blast wave. Therefore, cable or the conduit it is installed in should be fixed tight to the roof or rib to increase the likelihood of surviving a pressure wave.

Foster-Miller identified that in the Wilberg mine fire a cable secured to the mine roof was severely burned while an adjacent cable that accidentally fell to the mine floor when its attachment strap melted was not. They then note that the locations of cables and other communications and atmospheric monitoring system components with respect to elevation in the mine openings will be an important consideration for hardening.

A significant influence on the decision of where to locate cable runs is damage caused during normal operation and expected low consequence events such as rib spall. It is also important to consider the likely events to occur in the roadways the cables are run, for example a fire may be more likely in a belt road so this would not be the best roadway to run all communication and monitoring transmission cables/tubes. For critical communication or monitoring during emergencies it may be preferable to run the transmission cables from the component directly to the surface via a borehole, minimizing the distance it is vulnerable underground (as recommended by Task Group 4). A risk management process should be applied to the determination of the route used to connect sample or communication transmissions to the surface.

6.9.2.4 Armoured Cable/Tube Armoured cable is commercially available and used in many heavy industries as a means of hardening, and can withstand crush forces to 17-50 psi. These cables are also able to withstand abrasion and debris damage better than regular cable. When using armoured cables, adequate support and stronger connections at electrical boxes (to prevent pullouts) are critical. Connection hardware for the armoured cable must be as strong as the cable.

Armoured/hardened tube bundle lines are also available commercially. The use of armoured tube in the UK was detailed by the National Coal Board (1977). The description given by the National Coal Board of the core bundles indicates that even the unarmoured bundles were different to those currently used in Australia. The individual tubes in a multicore tube were laid helically like strands in a rope, holding the tubes together and helping to distribute stresses around curves. Where more than one layer of tubes is contained in the core the direction of twist is alternated between layers. A wrapping of Mylar or similar tape was used to shroud inner tubes to prevent penetration by PVC plasticizers from the outer sheathing.

This is in contrast to the tubes used in Australia where individual tubes in a core are run parallel (no twisting) and no shroud exists between the tubes and the outer PVC sheathing. It can be assumed that Page 72 of 154 ACARP C19010 Extension the configuration used in the UK added to the strength of the tube. Even the sheathed core tube used in Australia provides additional protection, particularly against kinking and where possible should be used.

Generally it was multicore tubing that was armoured and this consisted of galvanised steel wire laid in a direction opposite to that of the final layer of tubes as seen in Figure 22. A PVC jacket was applied over the armouring. Armoured tube was twice as expensive as unarmoured, and experience had shown that it need only be used in the shaft to avoid damage from impacts from equipment travelling in the shaft or in roadways where the risk of mechanical damage was high. Use was also influenced by the expected life of the system, the speed with which it was to be installed and the availability of suitable sizes and lengths.

The use of the armoured tube was more as a means of protection for day to day operations rather than to increase the survivability of the tube following a major incident. An ex UK National Coal Board Scientific Control unit employee advised of a case in the UK where wood from ventilation doors shredded by an explosion impacted and damaged a tube bundle sample line. It was believed by those involved, that the use of armoured tube would have offered sufficient protection to tube to prevent the damage and subsequent leakage (this didn’t prompt the subsequent use of armoured tubing in underground roadways).

Discussions with ex UK National Coal Board Scientific Control unit employees indicates that use of the armoured tube is in decline. The additional cost was given as a main reason. Working with the armoured tube can also be a lot more difficult. Joining tubes is made more difficult as the steel protection needs to be cut carefully with a hacksaw without damaging the tubes. The twelve core tube used was considered very heavy to work with. The single and two core armoured tube is reportedly easy to run out and work with manually, but after that the tubes become increasingly heavy and difficult to handle. Using armoured tubing over unarmoured increases the weight by a factor of approximately four. This is very significant for cores with larger numbers of tubes where even unarmoured tube weighs around 100 kg for every 100 metres.

Figure 22: Armoured tube (from National Coal Board 1977)

An example of armoured tube available in Australia and used in to run sample tubes on the surface is shown in Figure 23. Tubes that are suitable but not necessarily specific for use with tube bundle systems are available with various forms of armour/protection, Figure 24 and Figure 25 being examples. The plastic tubing industry has developed significantly since the introduction of tube bundles, and many options are possible. Essentially Australian mines have been using the same basic tube used for over forty years.

It may not be necessary to use armoured cable or tube for the full installation, however using risk management principals high risk areas could be identified and armoured cable/tube used for those areas. Page 73 of 154 ACARP C19010 Extension

Figure 23: Armoured tube used on surface

Figure 24: Protected Tubes-Supplier Collex

Figure 25: Protected Tube- Supplier Parker

6.9.2.5 Cable/Tube Shielding A method used in many industries for substantially hardening cabling is by enclosing it in conduit. There are many conduit options currently available with varying composition, physical properties and Page 74 of 154 ACARP C19010 Extension dimensions which can be chosen based on the needs. As mentioned in section 6.7 of this report, petro and chemical plants make use of fibreglass re-enforced plastic conduit because of its strength and fire resistance (also applicable to underground coal mines subject to FRAS compliance).

When used as a means of protection for cabling it is critical to ensure that the conduit itself is fixed to the roof or rib at sufficient spacing with suitably strong connections. It is possible to install the conduit with a breakaway standoff that would allow it to be struck, pulled down, and even buried to a degree but compliance loops and strong pipe connections could preserve the integrity of the cable itself.

Some obvious advantages of using conduit to protect cable are that it is immediately available and offers improved resistance to crush, abrasion, and debris damage. The associated disadvantages are that it is more costly, less flexible, and heavier than regular cable and it makes connecting to components more complex.

An additional layer of protection is the application of grout/shotcrete cable in conduit to the roof or rib. This improves resistance to overpressures and debris. Foster-Miller reported that a combination of shotcrete layers over 2” PVC conduit has resisted explosive overpressures to 190 psi.

Review of explosions has shown that cables (and conduit) are more susceptible to damage from blast forces and debris where they cross a roadway (intersections). In particular, cables in intersections are often subjected to blast waves across the entire length of cable and are frequently damaged. The use of blast shields an example of which was proposed by Foster-Miller and shown in Figure 26 to protect cables and conduits at these points Using blast shields for cables at passage intersections addresses a common cable failure point.

Figure 26: Cable Shield Proposed by Foster-Miller

Similar to the application of conduit, cable trays may be used to over a means of protection against smaller debris and wind drag from exposure to over pressures. An advantage of cable trays over conduit is the relative ease of making connections to components and accessing cables/tubes inside. It is envisaged that trays used would be enclosed with the face secured but able to be opened.

6.9.2.6 Fire Resistant Cable/Tube Fire resistant cable is another option that is already commercially available and used in other industries. As mentioned in Section 6.7 of this report there is an Australian Standard covering fire resistant cable- AS/NZS 3013:2005 –Electrical installations-Classification of the fire and mechanical performance of wiring system elements. This standard sets the requirements for emergency wiring systems, not only specifying the cable, but also the cable support system. The standard sets out a classification system for wiring system elements based on the system’s ability to maintain circuit integrity under fire conditions for a specified period and maintain circuit integrity against mechanical Page 75 of 154 ACARP C19010 Extension damage of specified severity. A limitation to the use of this cable is cost but again application should be done on a risk management basis, the gold mine that lost communications with the underground refuge chamber during a truck fire considered the additional cost worthwhile to ensure future survivability of communications.

Cable manufacturer Olex and CSIRO (CSIRO website) developed a plastic cable coating that transforms into a fireproof ceramic during a fire. The cable is able to withstand temperatures above 1000oC. A “ceramifiable” polymer cable coating material combining the properties of plastic and ceramics is used. As the polymer melts the ceramic forms a solid protective insulating layer enabling current to keep flowing. The cable was designed to maintain the integrity and continuity of circuits for building safety systems. Over 1,100 km of this product (Nexans Olex’s Alsecure®) has been supplied to the Brisbane Airport Link project for use in the project’s road tunnels.

Although not standard tube bundle sample line, flame resistant vacuum hose is available commercially, an example of which is given in Figure 27. As mentioned already the formulation used for tube bundle tube has essentially remained unchanged for forty years, during which time plastic technology has advanced.

Figure 27: Flame Retardant Vacuum Hose

6.9.2.7 Fixing/Installation Cable and Tube Cables are often simply secured by a tie or hook from a convenient point such as a roof bolt or in a similar manner on the rib. Slack is often left in the cable to prevent over tensioning from movements, including repositioning of connected equipment. Slack also increases the survivability of the cable in the event of small rock falls and other incidental contact with the cabling. Some manufacturers recommend leaving slack (5-10%). This does however leave droops between the secure points (Figure 28), increasing the chance of damage from being whipped if exposed to an over pressure even as low as a few psi. Using a “hanging” style of installation makes tubes and cable quite vulnerable to blast and debris damage. This was noted by MSHA when investigating explosions at Darby, Sago, Southmountain No. 3 Mine and Jim Walters No. 5 Mine.

Page 76 of 154 ACARP C19010 Extension

Figure 28: Hanging Tube Bundle Lines

Consideration has to be given not only to the damage of cables where they run connecting hardware to the surface but also where they connect to the hardware itself. Unless addressed it is possible that when exposed to some means of force the cable pulls out from the component eliminating any means of transmission (power or signal). Some components have a superior design for connection as shown in Figure 29. The example on the left of Figure 29 is more susceptible to losing transmission as it is only the terminals holding the cable in place, whereas the example on the right makes use of a saddle to hold the cable in place prior to the connection with the terminals. The saddle utilises the strength of the complete cable rather than just the conductor as is the case with the first example.

Figure 29: Cable Connections with Components (Foster-Miller 2009)

By allowing a loop or slack in the cable immediately prior to the connection the extra length may allow the cable to move without pulling the cable from the terminals. Any excess cable would need to be managed so as not to increase the potential for damage such as by debris or excessive loops that are exposed to wind drag effect of pressure wave.

Page 77 of 154 ACARP C19010 Extension

Finite element analysis undertaken by Foster-Miller on leaky feeder cable suspended from J hooks shows that drag loading along the cable will result in the cable being pulled out from electronic enclosures at very low pressures with cable spans 6m and higher. Cables hung every 1.5m feet can however withstand a 10 psi pressure wave. Cable was analysed, with the potential failure mechanism being pull-out from the electronics boxes. The “droop” between supports had a significant effect on the tensile load, with larger amounts of droop leading generally to lower tensile forces. The amount of droop between supports is recommended to be not less than 12 inches. Therefore, for this analysis the amount of droop between supports was 12 inches for all span distances. The span between hangers was varied to assess the effect of putting more hangers within the mine to support the cable.

Further analysis was made comparing unarmoured cable and conduit-covered cable with the following results:

 At 2 psi overpressure, the unarmoured cable fails in tension with 6m spans between supports, but survives with 3m spans. Conduit armoured cable survives with 6m spans if the diameter is 5cm and the material failure stress is above 18 ksi.  At 4 psi overpressure, the unarmoured cable fails in tension with 6m spans between supports, but survives with 3m spans. Conduit armoured cable survives with 6m spans if the diameter is 5cm and the material failure stress is above 35 ksi.  At 6 psi overpressure, the unarmoured cable fails in tension with 3m spans between support, but survives with 7.5 foot spans. Conduit armoured cable survives with 10 foot spans if the diameter is 1.5 inches and the material failure stress is above 18 ksi, or if the diameter is 2 inches and the failure stress is above 13 ksi.  At 8 psi overpressure, the unarmoured cable fails in tension with 10 foot spans between support, but survives with 2.3m spans. Conduit - armoured cable survives with 3m spans if the diameter is 5cm and the material failure stress is above 18 ksi.  At 10 psi overpressure, the unarmoured cable fails in tension with 2.3m spans between support, but survives with 1.5m spans. Conduit armoured cable survives with 2.3m spans if the diameter is 3.8cm and the material failure stress is above 17 ksi, or if the diameter is 5cm and the material failure stress is above 13 ksi.

Analysis was also performed to assess the effect of using a reinforcing cable to suspend a leaky feeder cable. The analysis demonstrated negligible improvement with this approach.

Foster-Miller concluded that although fibre-reinforced polymeric materials with failure stresses above 13 to 18 ksi are available, they are generally expensive. Therefore, conduit-armouring the cables will help, but the available increase in tolerance for overpressure is higher through increasing the number of supports for the cable. The preferable method was to increase the number of cable supports, allowing slack between the cables (minimum 12 inch droop). Note that this does not consider protection offered to the cable against debris.

This work was based on analysis rather than testing and didn’t consider hanger configuration which could be a significant issue. Cables have a minimum acceptable bend radius to continue working which is not considered in the modelling. It is possible that the cable could be stressed beyond this minimum in the real situation. It was recommended by Foster-Miller that a saddle shaped insert be installed where the cables are supported so as not to stress the cable beyond the minimum bend radius. This work was also carried out on leaky feeder cable installed fixed ~15inches from the roof.

Interviews conducted prior to this analysis suggested that the taut cables in straight runs had the best chance of surviving an explosion. Testing performed as part of this project and reported in Part B of this report found better survivability for cables and particularly sample tubes when they were fixed at regular points (~1m) and puled taught with no sag.

The distances that cables run in underground coal mines means that cables must be joined. These joins can be points of weakness and some means of physical protection needs to be applied to stop

Page 78 of 154 ACARP C19010 Extension separation at connectors and loss of transmission. This is true for tube bundle tubes as well which may be pulled from joiners resulting in leaks and no longer sampling from original location.

Figure 30 shows the way multicore tubes were recommended to be joined by The National Coal Board (1977). They reported that joints were the most frequent source of leaks and best housed in purpose designed junction boxes providing a suitable degree of protection. When assembling the tubes in a junction box, bending them so that they met in line at the central bulkhead may set up considerable stresses in the plastic so the looping-in technique shown in Figure 30 spreads the stress over a comparatively long length, hence averting mechanical failure. Litton (1983) also recommended the use of junction boxes for connecting bundles containing four or more tubes. The tubes entering the junction box being held firmly in place by brackets so that no tension is exerted on the connectors themselves. For joins not utilising junction boxes, Litton (1983) recommended firmly securing the tube to either the roof or rib on either side of the connection so as not to exert any tension on the connection, and then wrapping the connection with heavy duty tape.

The National Coal Board (1977) recommended that junction boxes be sited at deliberately chosen points rather than where a given length happened to end; proximity to a telephone and/or permanent lighting being clearly advantageous. This would allow the junction box to be positioned in a low risk area.

Figure 30: Looping-in at junction box (from National Coal Board 1977

6.9.2.8 Redundant Cable The running of redundant cables can be an appropriate means of increasing the chance of transmission during or following an event. It may be that redundant transmission lines are more cost effective than using armoured or hardened cable or installation techniques such as trenching.

If utilising redundancy as a technique, designing a system layout with a focus on installing the cables in separate pathways substantially apart or isolated so as not to be affected by the same event, is a key factor in improving system survivability. This may not be possible for all events, at Crandall Canyon, two independent systems were located in separate entries to provide redundancy and comply with the MINER Act. However, both systems failed in the section affected by the coal outbursts due to the large area affected. The severity of an accident may affect the redundancy of the systems and likely events need to be considered in the design stage to determine if redundancy will offer sufficient protection.

A simple and relatively low cost redundant technique is the use of a loopback cable that exits the mine through a borehole or portal not used by the primary cable. If the cable is broken communication is possible from either side of the break as there is an alternate transmission path to the surface. This type of set up is also known as a “ring network”. A failure at one point does not isolate other points in the system. It makes transmission inbye the break possible. It is compatible with many transmission media including standard copper cable and fibre optic cable. Page 79 of 154 ACARP C19010 Extension

Redundant sample tubes are typically run inbye seals in case the main sampling tube becomes inoperable, however redundant sample tubes from the seal (or elsewhere) to the surface are rare. Having redundant tube obviously increases the chance of having monitoring available during an emergency, and provides a backup tube in the case the primary tube clogs with coal tar as observed in coal fires. Although beneficial it is unlikely that mines would introduce redundant sample tubes considering the effort required not only for the installation but the required integrity testing (required monthly according to AS2290.3 1990).

6.9.2.9 Fibre Optic Cable Fibre optic offers multiple functionality capability in communications and monitoring of underground conditions. It can be used to measure methane concentrations, temperature and as a transmission medium for data generated by other techniques such as real-time sensors and communications. It has also been used as a means of providing IS power to electrochemical sensors. Dubaniewicz and Chilton (1995) reported on a US Bureau of Mines project using electrochemical sensors (carbon monoxide, nitrogen dioxide and sulphur dioxide) and fibre optics. In this project not only was the fibre optic used for telemetric communication of results but also as the power supply for the sensors. Each remote sensing unit contained gas sensors, an optical to electrical power converter and telemetry circuitry. A high powered solid state diode laser provided the laser power. The optical to electrical power converters supplied enough electrical power to operate the sensor and telemetry circuitry. The system reportedly worked well during a two week trial.

Einicke et al (2011) reported on research into providing communication networks for underground coal mines that could operate following an explosion, major fire, roof collapse or water inundation without relying on internal batteries or underground power. They proposed the use of trenched power over fibre technology buried to a depth of 0.6m.

Fibre optic is commonly used for data transmission from underground and opportunities exist to enhance its current use. It appears to be a technique in which there is significant interest and area where research is being directed (possibly due to its multi functionality). Work has been conducted in Australia supported by ACARP (example, Aminossadati et al 2013 ACARP Project No C20014) and according to the NIOSH website (http://www.cdc.gov/niosh/mining/researchprogram/contracts/contract200-2012-52515.html) RSL Fiber Systems, LLC, was recently contracted to develop an advanced fibre-optic-based mine-wide methane and temperature monitoring system. The objective of the project being to develop and test a prototype mine-wide monitoring system, based on laser energy distributed through a fibre optic network to monitor in near real time and record methane concentrations and ambient temperatures. This information is envisaged as being able to provide an indication of potential or actual explosion or fire conditions in underground coal mines. Also, the system needed to be designed to be functional following a large-scale catastrophic event.

Other industries use fibre optic cable and armoured versions are available. For fibre optic systems it must be noted that if the fibre optic is severed during an event, any monitoring data inbye the severed fibre will not be available at the surface, however anything outbye will still be transmitted. In the worst case this could mean all monitoring data could be lost.

With the multi functionality that exists with fibre optic, it may make implementation of redundanent systems more financially attractive. Despite the potential the technique may offer, there is a lack of a commercially available approved product (Aminossadati et al 2013) which obviously presents an obstacle to the use of such systems.

6.9.3 Shielding

Not only are critical communication and monitoring components vulnerable to damage by overpressures but they are also susceptible to damage by impact from airborne debris. Shielding components may offer a solution to this problem. Because of where they need to be monitoring real time gas sensors are particularly vulnerable to blast and debris damage. Task Group 4 identified that there was a need for more robust sensors but this is unlikely to eventuate at least in the short term. Page 80 of 154 ACARP C19010 Extension

The project team proposes the use of shielding plate as shown in Figure 31 to offer the sensor protection (see test results in Part B of this report). Sensor manufacturers have indicated that the shield will not hamper the gas measurements and may in high velocity areas even assist in operation but testing is required to ensure it offers sufficient protection from over pressures.

Figure 31: Shielding Plate for Real-time Gas Sensors

Foster-Miller also proposed shielding as a means of hardening for critical components such as tracking stations. Figure 32 shows a standard non-protected installation for a tracking station. Figure 33 shows the Foster-Miller proposed shielding/enclosure that offers protection particularly against flying debris but also provides protection and diversion of over pressures.

An advantage of using shielding and enclosures is that they can be retrofitted to existing equipment, with no need for product development or recertification for electrical equipment.

Figure 32: Non-Protected Rib Mounted Tracking Station (Photo credit Foster-Miller)

Page 81 of 154 ACARP C19010 Extension

Figure 33: Foster-Miller Proposed Shielding

6.9.4 Location

One of the simplest techniques for increasing survivability is locating equipment (that can be), in areas less likely to be exposed or affected by an event. Foster-Miller (2009) recommend mounting equipment in cut throughs rather than main roadways as a means of protection from debris, and to reduce over pressure (at least 1/3 of the total blast pressure). Consideration should also be given to which rib in a cut through equipment is fixed to in relation to debris that may be propelled forward.

Where possible, critical equipment should be installed away from areas at risk for explosion such as working faces or seals. Foster-Miller reported that locations greater than 300m from methane explosions should not experience significant blast pressure or debris. Other areas to avoid include belt roads, wet locations, and areas subject to roof falls. If critical equipment must be located in these areas hardening and redundancy techniques should be used.

6.9.5 Fixing There is evidence that consideration of survivability is neglected when installing communication and monitoring components. Components are often hung in the most convenient way, on rib bolts, clipped to mesh or suspended by light roof hangers or wire ties, with no “hard mounting” of cable loops.

Pressures as low as 2 psi will propel unanchored devices away from the source of the pressure. Not only is this increasing the chance of damage to the component but the chance of damage to the transmission cable is increased as is the likelihood of pulling the cable from the component connectors. Fixing components directly to the roof or ribs with bolts designed to withstand 45 psi pullouts will substantially harden the components. Use of backing plates such as seen in Figure 33 will assist in the robust securing of components.

6.9.6 Recessing Foster-Miller report that recessing components in roof/rib protects offers protection by reducing total pressure by approximately 1/3. This also eliminates pressure impact on the side of components that Page 82 of 154 ACARP C19010 Extension can damage or dislodge them. The likelihood of impact from flying debris is also reduced. Recessing in the rib protects from roof falls. A roof mounted enclosure was proposed by Foster-Miller and is shown in Figure 34. A similar setup could be established in ribs. The application of this technique would however be very dependent on the properties of the coal with ribs and roof prone to spalling unlikely to be suitable.

Figure 34: Foster-Miller Proposed Roof Mounted Enclosure

6.9.7 Redundancy Review of past incidents highlights the importance of redundancy for communication and monitoring survivability. Equipment may fail despite hardening efforts depending on the magnitude of any event. Requirements for secondary or redundant communication systems feature in legislative requirements. For redundant systems to be effective following or during an event the system must have been designed considering what may go wrong and where. All redundant components should be located such that an explosion or other mine accident cannot damage or destroy both components.

The application of multiple technologies provides redundancy only if the same event doesn’t affect both systems and failure in one system does not affect the other. It is important to consider the transmission methods and pathways of the redundant systems. For example if DAC communications are going to provide backup for telephone communications transmission cables need to be run separately.

In some cases a tube bundle system may provide redundancy for real-time monitoring (and vice versa) so consideration should be given to the path that the tubes are run in comparison to the transmission cables for the real-time monitoring. If they are the same it is possible that both could be lost in the one event.

The application and availability of wireless mesh systems is increasing. Meshing technology offers inherently redundancy and by increasing the number of nodes that can communicate with each other, redundancy is increased.

In a system, such as the wireless mesh system, where there are many paths which could result in success, the “self-healing” layout of the nodes helps retain system coverage post-event. In a system which is linear and depends on a single path for data transfer, any disruption to a central portion of the system may have significant effects inbye that location.

Page 83 of 154 ACARP C19010 Extension 6.9.8 Contingent Systems

It is quite possible that after an event all monitoring or communication systems are compromised and contingent systems need to be introduced. This is more likely to be successful and operational faster if considered prior to any event. It is likely that at least initially these contingent systems will need to be instigated from the surface with the availability of boreholes becoming critical. As detailed in the Moura Task Group 4 report “Both pre-installed and post-incident boreholes should be considered when developing Aided Rescue Management Plans.” It is likely that tube bundle and even real-time gas monitoring systems could be established through boreholes, but it is also possible to establish communications through same. At Pike River Mines Rescue personnel lowered two way radios down the slimline shaft into what was referred to as the FAB, providing an opportunity for communication with the surface.

Another alternative is the use of unmanned devices (such as robots or drones) that typically carry surveillance and atmospheric monitoring equipment. Despite much effort and research this option has not proved to be reliable when used in a real event (several robots remain in the drift at Pike River).

6.9.9 Summary

Unfortunately there is no one means that will guarantee survivability for monitoring and communication systems for all events. There are however means that will increase the chance of system survival (or at least part of); much of it easily implemented. The best solution may be use of multiple techniques (related to risk zone) and multiple or redundant technologies. The attachment of both components and cabling/wiring and the protection of the cable can be at least as important as the ruggedness of the components themselves and as discussed above there are means available now for improvement.

Improvements are more likely to be achieved by improved installation techniques and retro fitting shielding, rather than the development of hardened more robust components by manufacturers, at least for the next few years.

It would appear that when installing communication systems not enough consideration is given to their survivability or robustness (other than for day to day operational reasons). By using risk management process and considering this in the design and implementation stages significant improvements are easily achievable.

The work done by Foster-Miller (2009) in this area is very useful but as included in their report a lot of the recommendations are based on modelling analysis and there is a need for explosion type testing to verify recommendations made. Small scale testing has been under taken as an approved extension to this project and results are included in Part B of this report.

Page 84 of 154 ACARP C19010 Extension

7. Conclusions

The parameters and physical constraints that underground coal mines operate under, set them apart from all other industries and introduces unique issues with ensuring continued effective communication and atmospheric monitoring during or following an incident. Those incidents don’t just include explosions and fires. Roof falls, outbursts and inundations need to be considered as well.

Existing legislation, standards and guidelines require mines to make provision for ongoing effective communication and monitoring during emergency situations, however most mines are probably not adequately prepared. Mines tend to be set up for “peace time” operation rather than for incidents, so there is a definite need for improvement in this area. The report prepared by Mine Safety Operations on emergency management systems in NSW (Macpherson, 2010) identified that risk assessments weren’t used in the development of underground emergency systems, contributing to the failure to make monitoring and communications systems more robust. Mines sites need to adopt effective risk management processes to ensure that emergency response plans are effective, that nominated controls will be available and that legislative requirements are complied with.

Once an incident or emergency situation occurs it is too late to start planning on how effective communications and data collection will be achieved. This needs to be covered in emergency response plans. It may be that backup systems are used; however this needs to be pre planned. Mines rescue teams won’t deploy underground without adequate knowledge of the atmospheric conditions underground.

The evaluation of survivability of workers unaccounted for underground needs to be addressed early, parallel to rescue efforts. Information from communication and data systems is critical in this evaluation, as is the knowledge that any information used was reliable and representative.

Hardware solutions for more robust gas monitoring are some time away if at all; underground coal mining offers a limited market through which investment in research may not be returned with the added requirement (and expense) for specific approval processes for equipment use. This may be a reason why recommendations of hardening of gas monitoring and communications systems recommended after Moura No. 2 have not eventuated. The reality is if mines want to improve the survivability of their systems they will need to look at other ways. Uptake of hardening strategies is more likely if it can be done with existing hardware and little change to current practices, maintenance requirements and resources and minimal disruption or compromise to productivity. Shielding, positioning and installation techniques are currently available and offer solutions able to be introduced in the short term.

There are some potential developments and commercial options for the provision of power to these systems that are likely to allow operation for longer periods than the current battery backup systems can cater for. The current power backup time frames are not representative of the times required for significant events.

Survivability will depend on amongst other things the nature, location and magnitude of the event. It may not be practical to provide protection to all equipment against all events, for example the over pressure at the site of an explosion may be 45psi or temperatures around a fire may be over 1000oC. However continued communications and monitoring inbye and outbye of the event, where conditions are not as extreme may provide critical information particularly for those escaping the mine.

Although required, little guidance or information is available to operations in achieving acceptable levels of robustness for monitoring and communication systems. Installation standards or guidelines are required to assist operations improve robustness of systems. There are enough example cases of

Page 85 of 154 ACARP C19010 Extension data and communication systems remaining operational after incidents to know that it is possible for systems to survive and worth the effort to protect them in case of such events. Simple improvements to installation practices including choice of locations are likely to increase the survivability of these systems. It appears that the cable or transmission medium is currently the most vulnerable component of systems and worthy of protection consideration. The suggestions made on methods for hardening made by Foster-Miller (2009) were in a large based on finite element analysis and they recommend that testing be conducted at a facility such as Lake Lynn Experimental Mine (no longer operational). Testing is required to assist develop best practices for installation and shielding of critical components. Small scale testing using the propagation tube at Simtars offered a means of achieving this testing at a reasonable cost and time frame. This testing highlighted that installation methods and could significantly improve the survivability of cables, tubes and hardware. There are advantages in pursuing further testing to include debris and the inclusion of cut through configurations but since the closure of Lake Lynn, no such facilities are known to exist. Results from larger scale testing are not required before the findings from the small scale testing are acted on.

If mines are going to rely on “hardening” hardware as a means of ensuring ongoing system availability, then a system of determining the level or degree of protection any hardening provides needs to be developed.

Regardless of how much it is hardened some equipment may fail so using systems with redundancy or contingencies will help to ensure that communication and data collection can be ongoing following a greater range of events.

Increasing the survivability or ensuring the availability of communications and data mine wide may be best addressed utilising different approaches for different risk zones. Working faces which represent the highest risk zones may need redundant or backup systems employed after an event whereas for other areas throughout the mine, hardening techniques may be sufficient. Like all areas, the approach should be risk management based.

There is a definite need for the forming of an industry group as per Recommendation 7 from Moura No. 2 Task Group 4 to coordinate the advancement of capabilities to alert, to communicate with, and to assess the status of underground persons during a mine emergency. It is pleasing that as a result of this project The Department of Natural Resources and Mines through the Chief Inspector of Coal Mines has agreed to establish such a group to progress the original recommendations and by default continue and progress the work identified in this project.

Page 86 of 154 ACARP C19010 Extension

8. Recommendations

1. Formation of an industry committee/forum as per Recommendation 7 from the Moura No. 2 Task Group 4. This group (technical experts and operators) was intended to coordinate the advancement of capabilities to alert, to communicate with, and to assess the status of underground persons during a mine emergency. This original group only meet a couple of times after forming and there is no evidence of any significant progress against the recommendation. The Department of Natural Resources and Mines through the Chief Inspector of Coal Mines has agreed to establish such a group to progress the original recommendations (and by default continue and progress the work identified in this project). It is recommended that equipment manufacturers be included in this group so that they are aware of the issues facing and requirements of the industry allowing them to provide short term solutions and where possible research and development for longer term solutions.

2. An industry best practice/guideline be developed (championed by the group formed in Recommendation 1) for the installation and protection of communications and monitoring equipment using technical information contained within this report.

3. Mines improve the probability of survival of their communication and gas monitoring systems by applying a risk management process covering at least the following:  what type and magnitude of events possible  areas of elevated risk  protection required  installation standards  positioning of hardware  routes for cable and tubes to surface  redundancy  contingencies

4. Ensure that a means of power is available to operations for communications and monitoring even during long periods of loss of underground power (championed by the group formed in Recommendation 1).

5. Develop a system of quantifying “survivability” (championed by the group formed in Recommendation 1).

6. Trial the blast shields shown to protect gas sensors in testing in an operating mine to evaluate whether normal operation is compromised.

7. Mine sites include (and resource) the determination of whether communications are fully operational following an event in “duty cards”. This functionality is available in many of the systems in use but mine sites typically are unaware of how to use it, but information available from suppliers.

8. Mine sites include (and resource) the determination of the status of gas monitoring following an event in “duty cards”. Parameters to be reviewed for tube bundle include; tube vacuum pressures, gas results (including comparison with real-time), comparison of known draw times with those for first observation of signs of event (as detailed in the Moura No. 2 case study).

9. Mine sites should consider the resources required (people with the necessary skills and knowledge) to make these evaluations and determinations to support the decision making process. It took an appreciable amount of time and knowledge and it is not envisaged that

Page 87 of 154 ACARP C19010 Extension

mine sites with the current level of preparedness would be in a position to make such evaluations and determinations within a timeframe that could be applied to results generated during response to the incident. This is an area needing improvement within the industry.

10. Investigate the possibility of a talk over ride button from surface to allow listening to what is happening underground without the need for someone underground to activate (championed by the group formed in Recommendation 1).

11. If facilities for large scale testing become available further testing should be conducted; however this should not delay implementation of findings from small scale testing.

Page 88 of 154 ACARP C19010 Extension

9. Bibliography Aminossadati, S, Bowen, W, Sheridan, E, Amanzadeh, M & Kizil, M 2013, ACARP Project No C20014: An Innovative Fibre Optic Based Methane Sensor, report sponsored by the Australian Coal Association Research Program, The University of Queensland, Brisbane.

Anderson, I, Urosek, JE & Stephan, C 1997, 'The recovery and investigation of the explosion at the Endeavour colliery', in R.V. Ramani (ed.), Proceedings of the 6th International Mine Ventilation Congress: May 17-22, 1997, Pittsburgh, Pennsylvania, USA.

Brune, JF 2013, 'Methane-air explosion hazard within coal mine gobs', Transactions of the Society for Mining, Metallurgy, and Exploration, vol. 334, pp. 376-390.

Brune, JF & Goertz, B 2013, Lessons Learned from Mine Disasters: New Technologies and Guidelines to Prevent Mine Disasters and Improve Safety, Research Report on Centers for Disease Control and Prevention Grant No. 1H750H009822-01, submitted to Wheeling Jesuit University, Colorado School of Mines, Golden, Colorado.

Committee on Mine Safety: Essential Components of Self-Escape, Board on Human-Systems Integration, Division of Behavioral and Social Sciences and Education, 2013. Improving Self-Escape from Underground Coal Mines. National Research Council of the National Acadamies.

Department of Mineral Resources New South Wales Outburst Mining Guideline MDG No: 1004 July, 1995.

Dubaniewicz, TH, Jr. & Chilton, JE 1995, Optically Powered Remote Gas Monitor, Report of Investigations 9558, US Bureau of Mines, Pittsburgh, Pennsylvania.

Einicke, G., Hainsworth, D., Munday, L., Haight, T., 2011. Optically –powered Underground Coal Mine Communications, 11th Underground Coal Operators' Conference, University of Wollongong & the Australasian Institute of Mining and Metallurgy, 2011, 189-196.

Ellis C, 1980. Explosion at West Wallsend No. 2 Colliery on 8th January 1979: Report on Certain Scientific Aspects, Chemical Laboratory Branch, Department of Mineral Resources, New South Wales. PO Box 76, Lidcombe, NW 2141.

Foster-Miller, Inc., 2009, NIOSH Contract Number 200-2008-26864: System Reliability and Environmental Sustainability, report No NSH-090164-1882, prepared for NIOSH, Foster-Miller, Inc., Waltham, Massachsetts.

Fowler J.C.W. and Hebblewhite B.K., 2003, ACARP Project No. C8017 Final Report, “Wind Blast and Methane Expulsion: Extension of Field Monitoring to Generalise the Results of Projects C7031 and C8017”

Fowler, J.C.W. & Sharma, P. 2000. ACARP Project C6030 Final Report, ‘Dynamics of Wind Blasts in Underground Coal Mines’

Gates, RA, Phillips, RL, Urosek, JE, Stephan, CR, Stoltz, RT, Swentosky, DJ, Harris, GW, O'Donnel, JR, Jr. & Dresch, RA 2007, Report of Investigation: Fatal Underground Coal Mine Explosion, January

Page 89 of 154 ACARP C19010 Extension

2, 2006, Sago Mine, Wolf Run Mining Company, Tallmansville, Upshur County, West Virginia, ID No. 46-08791, MSHA, Arlington, Virginia, 9 May 2007.

Gates, RA, Gauna, M, Morley, TA, O'Donnel, JR, Jr., Smith, GE, Watkins, TR, Weaver, CA & Zelanko, JC 2008, Report of Investigation: Underground Coal Mine, Fatal Underground Coal Burst Accidents, August 6 and 16, 2007, Grandall Canyon Mine, Genwal Resources Inc., Huntington, Emery County, Utah, ID No. 42-01715, CAI-2007-15-17, 19-24, MSHA, Arlington, Virginia.

Golledge, P, Humphreys, D, Cliff, D, Reinhardt, D, and Hester, C, 1995. Report of Investigation of various matters associated with explosions at Moura No. 2 underground mine on 7 and 9 August 1994. Report R005-94/95, SIMTARS, DME, January 1995.

Goran, AJ 1979, Explosion at West Wallsend No. 2 Colliery on 8th January 1979: Report of His Honour Judge A.J. Goran Q.C. Following an Inquiry by the Court of Coal Mines Regulation Established Under Section 31 of the Coal Mines Regulation Act 1912, As Amended, Government Printer, Sydney, Australia.

Goran, AJ 1980, Explosion at Appin Colliery on 24th July, 1979: Report of His Honour Judge A.J. Goran Q.C. Following an Inquiry by the Court of Coal Mines Regulation Act, 1912, as Amended, Government Printer, Sydney, Australia.

Griffin, KR 2013, Utilization and Implementation of Atmospheric Monitoring Systems in United States Underground Coal Mines and Application of Risk Assessment, Ph.D thesis, Virginia Polytechnic Institute and State University, Balcksburg, Virginia.

Green A.R. Upfold R.W. (1987) The explosion at Moura No. 4 Mine, Queensland: A case study, Proceedings of the 22nd International Conference of Safety in Mines Research Institutes, Beijing, China.

Kininmonth, RJ 1981, “Summary Of Investigations Into Appin Colliery Explosion”, in Ignitions, Explosions and Fires in Coal Mines Symposium, AusIMM, Wollongong, Australia, 12-15 May 1981.

Kininmonth, RJ 2010, “Appin Colliery explosion reassessed”, in Aziz, N (ed), 10th Underground Coal Operators' Conference, University of Wollongong & the Australasian Institute of Mining and Metallurgy, 2010, 299-311.

Lagace, R.L. & Moussa, N. A., 1982. Summary Data Report, Contract No. J0308037 United States Department of the Interior, Bureau of Mines “Initial Study of Buried Communication Cable for Underground Mines”.

Light, TE, Herndon, RC, Guley, AR, Jr., Cook, GL, Sr., Odum, MA, Bates, RM, Jr., Schroeder, ME, Campbell, CD & Pruitt, ME 2007, Report of Investigation: Fatal Underground Coal Mine Explosion, May 20, 2006, Darby Mine No. 1, Kentucky Darby LLC, Holmes Mill, Harlan County, Kentucky, ID No. 15-18185, CAI 2006 27-31, MSHA, Arlington, Virginia.

Litton, CD 1983, Design Criteria for Rapid-Response Pneumatic Monitoring Systems, Information Circular 8912, US Bureau of Mines, Pittsburgh, Pennsylvania.

Page 90 of 154 ACARP C19010 Extension

Lynn, KP 1987, Report of an Accident at Moura No4 Underground Mine on Wednesday, 16th July, 1986, Warden's Inquiry Conducted Pursuant to Section 74 of "The Coal Mining Act, 1925-1981", Government Printer, Queensland Government, Brisbane, Australia, viewed 7 October 2014, .

Mackenzie-Wood, AP & Ellis, CG 1981, 'Use of a mobile gas laboratory at mine fires', in Ignitions, Explosions and Fires in Coal Mines Symposium, AusIMM, Wollongong, Australia, 12-15 May 1981, pp. 15-1 - 15-11.

McKinney, R, Crocco W, Tortorea, J S, Wirth, G J, Weaver, CA, Urosek, JE., Beiter, D A, and Stephan, C R, (2001) United States Department Of Labor Mine Safety And Health Administration Coal Mine Safety And Health, Report Of Investigation Underground Coal Mine Explosions July 31 - August 1, 2000 Willow Creek Mine - Msha Id. No. 42-02113 Plateau Mining Corporation Helper, Carbon County, Utah.

McKinney, R, Crocco, W, Stricklin, KG, Murray, KA, Blankenship, ST, Davison, RD, Urosek, JE, Stephan, CR & Beiter, DA 2002, Report of Investigation: Fatal Underground Coal Mine Explosions, September 23, 2001, No. 5 Mine, Jim Walter Resources, Inc., Brookwood, Tuscalossa County, Alabama, ID No. 01-01322, CAI 2001-20 through 32, MSHA, Arlington, Virginia, 11 December 2002.

Macpherson, D., 2010. Report into Emergency Management Systems in NSW underground coal mines.

Mines Rescue Working Group and Mine Safety Operations Division of Industry and Investment NSW, 2010. Mine Design Guidelines MDG 1020 Guidelines for underground emergency escape systems and the provision of self rescuers. New South Wales Department of Primary Industries.

Mine Safety and Health Administration, United States Department of Labor, Report of Investigation Underground Coal Mine Explosions July 31 - August 1, 2000 Willow Creek Mine - MSHA Id. No. 42-02113 Plateau Mining Corporation Helper, Carbon County, Utah.

Mine Safety and Health Administration, United States Department of Labor, Report of Investigation Fatal Underground Coal Mine Explosion January 2, 2006 Sago Mine, Wolf Run Mining Company Tallmansville, Upshur County, West Virginia.

Mine Safety Operations Division, 2007. Mine Design Guidelines MDG1003 Windblast Guideline New South Wales Department of Primary Industries.

Murray, KA, Pogue, CW, Stahlhut, RS, Finnie, MG, Webb, AA, Burke, AL, Beiter, DA, Francart, WJ, Tjernlund, DM & Waggett, JN 2007, Report of Investigation: Fatal Underground Coal Mine Fire, Aracoma Alma Mine #1, Aracoma Coal Company, Inc., Stollings, Logan County, West Virginia, January 19, 2006, ID No. 46-08801, MSHA, Arlington, Virginia.

National Coal Board 1977, The Tube Bundle Technique for the Continuous Monitoring of Mine Air, National Coal Board, London.

NIOSH 2011. Tutorial on Wireless Communication and Electronic Tracking Part 1: Technology

Page 91 of 154 ACARP C19010 Extension

Overview. National Institute for Occupational Safety and Health (NIOSH) Mining Division. http://www.msha.gov/techsupp/PEDLocating/WirelessCommandTrack2009.pdf (accessed 28/12/2012)

NSW DMR 1996, Explosion at Endeavour Colliery, 28 June 1995, Summary of Investigation, MDG No. 1007, New South Wales Department of Mineral Resources, Minerals & Energy House, St Leonards, NSW, Australia.

Phillips, CA 2012, Report of Investigation into the Mine Explosion at the Upper Big Branch Mine, April 5, 2010, Boone/Raleigh Co., West Virginia, West Virginia Office of Miner's Health, Safety & Training, West Virginia, 23 February 2012.

Royal Commission on the Pike River Coal Mine Tragedy 2012, Final Report, , New Zealand, Reported to the Governor General on 30 October 2012.

Staunton J.H., 1998. Report of a Formal Investigation Under Section 98 of the Coal Mines Regulation Act, (Gretley).

Stephan, C R, Urosek, J E, and Giardino, D A , 1994. A Technical evaluation of information provided by the Department of Minerals and Energy, Brisbane, Queensland Australia concerning the Underground Coal Mine Explosions Moura No. 2 Mines, Moura, Queensland, Australia, August 7 and 9.

Stout, DL, Rinehart, JW & Snyder, MP 2003, Report of Investigation: Underground Coal Mine, Non- Injury Mine Fire Accident, September 16, 2002, #3 Mine, Fairfax Mining Co., Inc., Clarksburg, Harrison County, West Virginia, I.D. No. 46-08633, MSHA, Morgantown, West Virginia, 20 August 2003.

Task Group No. 4 1996, Moura No. 2 Inquiry Warden's Report Implementation – Mines Rescue Strategy Development (Recommendations 9 and 10) (C Rawlings - Chairman), Joint Coal Industry Committee from QLD and NSW, Australia.

West Virginia Office of Miners’ Health, Safety and Training, Upper Big Branch Mine Explosion, Report of Investigation presentation, February 23, 2012.

Windridge, F, 1995. Warden’s Inquiry. Report on an Accident at Moura No. 2 Underground Mine on Sunday, 7 August 1994.

Zipf, RK, Jr., Sapko, MJ & Brune, JF 2007, Explosion Pressure Design Criteria for New Seals in U.S. Coal Mines, Information Circular 9500, United States Department of Health and Human Services, Pittsburgh, Pennsylvania. http://www.csiro.au/en/Organisation-Structure/Flagships/Future-Manufacturing-Flagship/Sustainable- High-Performance-Materials/High-performance-composites/Ceramifiable-plastic-cables.aspx accessed 30/07/2013

Page 92 of 154 ACARP C19010 Extension

Appendix A: Case Studies

Page 93 of 154 ACARP C19010 Extension

A1 Pike River Mine, New Zealand 19th November 2010

Reference: Royal Commission Report on the Pike River Coal Mine Tragedy, 2012, and Authors’ own notes and personal correspondence.

Video evidence of an underground coal mine explosion is rare; however the video camera at the portal at the Pike River Mine in New Zealand captured footage of all four explosions in November 2010. The images from the camera provided useful information during response to this disaster including the direction of ventilation flow and during the investigation phase to determine what most likely had happened. The camera was in a position that meant it survived all four explosions.

Out of the 31 men underground at the time of the first explosion at Pike River only two survived. After spending some time unconscious one of the survivors regained consciousness and travelled a short distance inbye to a telephone, from which he contacted surface control. Calculated likely velocities of the gas exiting the portal reported by the Royal Commission are between 30 and 70m/s (108-252km/h). The phone used to contact control was almost 2000m inbye the portal and positioned only one metre from the corner of the main drift on the outbye rib of B1 cut through (Figure 35). This positioning would mean that the phone was exposed to most of the force travelling up the drift. This phone was fixed to a rib bolt through a hole in the phone’s bracket and held in place by a nut. Of importance is the fact that the cable was also firmly anchored directly below the phone so that it could not stress the connector. The cable then ran up the main drift predominantly in the centre of the roadway clipped to a catenary wire. This phone (and its connection with the surface) withstood the velocity and over pressure of the explosion and was operational as evidenced by its use to contact control. The standard of installation is the most likely reason the phone remained operational.

Figure 35: Location of Phone Used Post First Explosion

Like many disasters the initial response to the Pike River Mine disaster was hindered by the lack of available information on the conditions within the mine. Following the initial explosion the mine’s real time gas monitoring system was no longer reporting to the surface even though sensors had been fitted with uninterrupted power supply units. The Royal Commission reported “it is likely that the sensors or their wiring were damaged in the explosion”. There was no tube bundle system installed at Pike River. Pike did not have alternative equipment designed to obtain gas samples from the mine. There were no boreholes connecting the surface to the workings other than in the stone drive and an intake shaft. The only available sample location not in intake air was the ventilation shaft. As reported by the Royal Commission, in the early evening following the explosion, Pike employees flew (helicopter) to the main vent shaft and used hand held gas detectors to measure the atmosphere in the

Page 94 of 154 ACARP C19010 Extension fan housing. This was a hazardous activity. The mine was not equipped to collect samples and used a stomach pump borrowed from ambulance personnel to collect samples into bags from a tube lowered down the shaft. The significance of the recommendation from the Box Flat Inquiry “That provision be made for atmospheric sampling from fan ducts” is particularly relevant to this situation. Samples were sent via helicopter to a nearby mines rescue station for analysis by gas chromatography. An external gas chromatograph was brought to site and set up and operational approximately 22 hours after the initial explosion. Samples continued to be delivered to both the administration area and mines rescue station by helicopter due to remoteness and access difficulties to sample locations.

It must be noted that the temporary “plastic” sample lines were generally destroyed requiring replacement following each subsequent explosion. This meant that no data was available for some time following each explosion. Following the fourth explosion when an ongoing major gas fire was initiated at the vent shaft, sampling was not possible from this location until the flames were extinguished. There were also issues identified with portable analysers with inbuilt pumps not being able to draw samples through the tube that extended from the surface to the workings (~100m). This resulted in unreliable, non-representative data being generated for at least one sample point. The problem was rectified by using a larger capacity pump to draw the sample to the surface, from which the analyser drew its sample.

The Royal Commission reports that the samples from the vent shaft were unlikely to be representative of the mine. Issues with the available sample locations, lead to a decision to drill a borehole to provide an additional and more representative sample location. The hole was commenced 21st November with first samples collected after 5:00am on 24th November. There were issues with the borehole not intersecting the workings; fortunately there was connectivity with the underground atmosphere, otherwise sampling would have been delayed further. Interpretation of data from this sample location in conjunction with results from the vent shaft indicated the likely presence of an ignition source and a possible explosive atmosphere. The same day there was a second explosion.

Had a there been a tube bundle system installed at Pike, although there was a likelihood that tubes may have been damaged there would still have been information available at least from some locations underground.

During post fourth explosion monitoring operations at the Slimline shaft, the low density polyethylene tube used for sampling had to be replaced with copper tube because the plastic tube had melted. The tube used has a melting temperature of the order of 120oC. Ongoing issues were experienced with tar clogging sample lines. Up to 11.8% carbon monoxide and 7.3% hydrogen were measured at this location meaning reliance on gas chromatography for accurate results.

An overland tube bundle system was installed for ongoing gas monitoring and sample collection.

A considerable time after the fourth explosion at Pike River, it was identified that the methane sensor located in the surface évasé at the vent shaft was still in place. This was despite the évasé itself being shifted by the force of the fourth explosion. It is believed that this sensor was installed by securing it with a two metre rope. The sensor was recovered from the évasé and sent to a testing laboratory for further investigation. The external wiring gland was missing and there was evidence of heat and mechanical damage. When power was supplied to the sensor and methane test mixes applied it was observed that the sensor was still capable of measuring methane although the concentrations were not accurate. The fact that the sensor when repowered was capable of making any measurement is remarkable when considering it had been in the path of four explosions including the fourth that had enough energy to move the évasé from its location. This shows that with more attention to installation standards (particularly cable), real time monitoring results may be available even after an explosion.

Page 95 of 154 ACARP C19010 Extension

th A2 Upper Big Branch, USA 5 April 2010

Reference: Page et al (2011); Brune & Goertz 2013; Brune 2013; Phillips 2012; MSHA Upper Big Branch Mine-South Mine ID: 46-08436 April 5, 2010 Accident Public Briefing June 29, 201; MSHA video -http://www.msha.gov/PerformanceCoal/PerformanceCoal.asp; West Virginia Office of Miners’ Health, Safety and Training, Report of Investigation presentation.

An explosion occurred at Upper Big Branch (mine plan Figure 36) at 3:02pm, April 5, 2010 resulting in twenty nine fatalities and two injuries. MSHA determined that the cutting bits on the tail drum of the longwall shearer likely generated hot streaks on the sandstone roof or floor and that this was the most likely ignition source of gas released from a fault zone as a floor feeder near the back of the shields. This ignition of methane did not begin to propagate immediately. The flame from the ignition burned near the longwall tailgate for approximately two minutes and could not be controlled by the miners at the shearer, forcing their evacuation. The methane ignition then triggered a localised methane explosion in the Tailgate 1 North providing the initial explosive energy to suspend float coal dust in the tailgate entries that allowed transition to a coal dust explosion. The flame zone from the coal dust explosion was extensive as seen in Figure 37. Flames did not travel from the tailgate, where the explosion initiated, across the longwall face to the maingate. Pressure and heat (no flames) from the coal dust explosion that propagated did however extend some distance across the longwall face from the maingate end as identified by investigators and shown in Figure 38.

Figure 36: Upper Big Branch mine plan (taken from Page et al, 2011)

Page 96 of 154 ACARP C19010 Extension

Figure 37: Extent of flame from explosion (taken from WVMHS&T Report of Investigation presentation)

Figure 38: Indication of heat at longwall face (taken from Phillips 2012)

This was a massive blast that caused widespread damage underground, yet rescue teams were able to penetrate all the way into the mine and to the longwall face. People standing just inside the mine were blown over by the force of the blast exiting the mine. Dust exited the mine as well. At the time of the explosion, a crew that had been setting up a miner section near the Ellis Portal was traveling inbye in a mantrip on the way back to the North Portal at the end of their shift. Joshua Williams, Roof Bolter, described the experience: “We was coming up the track, and the guy I was bolting with, he said, ‘Man, it's dusty.’ I said, ‘Yeah.’ Then he said, ‘Do you feel a lot of air coming down the track?’ I said, ‘Yeah.’ He said, ‘It wasn't doing that this morning.’ We kept on going, and my ears popped and

Page 97 of 154 ACARP C19010 Extension

I couldn't hear nothing. And then that's when we hit air… started pushing our mantrip back. It was throwing blocks, foam. That's when I laid down on the mantrip and threw my jacket over my head and was starting to get my rescuer out because I didn't know what in the world was going on. …It blew our mantrip. It blew it probably five crosscuts [outby]…we rode the track all the way back out to the Ellis Portal, and then we went outside.”

People at the south portal were unaware that an explosion had occurred. Rescuers were able to use mantrips to renter the mine and reach the TG22 crew inbye the 78 switch. Two miners in the TG22 crew survived the blast the others were overcome by the noxious atmosphere. Rescuers found debris barring the track inbye 78 switch. Ventilation controls inbye were also damaged. They were able to proceed on foot all the way to the longwall. Communications were also destroyed.

It would appear that all workers inbye the crew that were exiting the mine died instantaneously, none of the refuge bays in the mine were activated.

After the explosion persons entered the mine without mine rescue apparatus, some had hand held multigas detectors and later reported that they had been to the tailgate and almost the maingate of the longwall but had to retreat as a result of excessive carbon monoxide levels. From April 5th to April 9th, over 20 mine rescue teams worked around the clock in an attempt to locate and rescue the missing miners. During this time rescue teams were evacuated several times due to flammable gas concentrations and possible ignition sources. Gas monitoring was initially conducted at fans and portals as well as that conducted by rescue teams.

At 12:16 am (April 6), a rescue team reported an air quality reading of 3.2% oxygen, 9,999 parts per million of carbon monoxide and over-range on methane. At 12:37 am a team reported heavy smoke, that they could not determine air flow direction and that all detectors showed over-range on methane and carbon monoxide, with 3.2% oxygen. At 12:45 am all rescue teams were ordered by the command centre to retreat out of the mine due to the explosive mixture of gas and evidence of a fire. All of the rescue teams were out of the mine by 2:30am.

Personnel from MSHA’s Directorate of Technical Support including specialists from the Physical and Toxic Agents Division and the Ventilation Division, arrived at the mine at approximately 1:00am on April 6th with a portable gas chromatograph. The mobile gas laboratory arrived at approximately 3:00 pm, with additional gas chromatographs. Drilling of three boreholes was planned to better assess the atmosphere. Two of these were started on the 6th and the first (Hole 1A) intersected the mine at 4:00 am on 7th April at crosscut 35 on HG 22. Gas levels from this borehole and the Bandytown fan were monitored to determine when safe re-entry of the mine would be possible. Re-entry mine rescue plans were developed while waiting for hazardous gas levels to decrease.

On April 8th at 1:45 am, results for samples collected from the boreholes permitted re-entry into the mine for rescue operations. Four teams re-entered at 4:55 am but at 9:29 am the results from a borehole indicated that there was an explosive mixture of gas in the mine. All rescue teams were ordered to evacuate and were on the surface by 10:55 am. Teams re-entered the mine on April 9th at 12:42 am but again due to results from a borehole indicating that there was a fire and methane air mixture close to the explosive range at 4:44 am the teams were ordered to retreat to the surface immediately and returned to the surface by 6:11 am.

At 9:02 am inert gas was injected into borehole BH 1A. At 2:32 pm nitrogen trucks completed pumping into borehole BH 1A and changed to a nitrogen generator at 2:40 pm and continued pumping. A quantity of nitrogen equal to approximately twice the volume of the HG 22 mine workings had been injected into BH 1A. At 4:15 pm it was determined safe for mine rescue teams to re-enter the mine and two teams re-entered the mine. At 11:40 pm 9th April, the last of the victims was found. Recovery was complete by 5:30am April 12.

Page 98 of 154 ACARP C19010 Extension

Nitrogen was injected into the mine through boreholes to displace combustion gases where ventilation could not reach. No personnel were permitted inside UBB between April 13th, 2010 and June 2, 2010, due to these gases and the potential of active “hot spots” to reignite combustible mixtures. Hot spots of glowing embers were still found in a few locations when the mine was re-entered in June 2010 (and during recovery).

On May 27, all parties involved in the decision making agreed that the mine atmosphere had stabilised sufficiently to allow re-entry, pending the finalisation and sampling of borehole “HG 21-1” near the longwall. Problems arose with HG 21-1, however, when the drill intersected an inactive area of a mine above UBB, causing a delay in the completion of the borehole until June 6. Nonetheless, because of the distance separating the longwall and the portals, exploration of the portal areas began on June 2, with the anticipation that borehole HG 21-1 would be completed prior to any exploration in the longwall area. All parties agreed that the mine atmosphere would continue to be sampled for stability during exploration and recovery work. The decision for re-entry went ahead without results from borehole HG21-1, based on the distance between the longwall and portals.

MSHA (Page et al 2011) determined that the coal dust explosion began approximately at the intersection of crosscut 48 and entry No. 6 in Tailgate 1 North. Explosion forces were generated by the flame in all directions, including back across the longwall face. Evidence indicates that approximately 14 psi travelled back to the longwall tailgate from the coal dust explosion. Explosion damage indicates that the coal dust explosion initially propagated inbye in entries 5 and 6 and outbye in entries 5, 6, and 7. Eventually, all entries and crosscuts in Tailgate 1 North from as far inbye as crosscut 77 exhibited evidence of flame. The flame travelled inbye at about 300m/s while generating a pressure exceeding 18 psi. Flame propagated outbye from crosscut 48 and involved all entries and crosscuts of Tailgate 1 North, outbye to the Old North Mains. The flame also propagated to crosscut 67 in Old North Mains in entries 1, 2, and 3.

The explosion flame travelled outbye at the same time as it was traveling inbye in the tailgate entries. Flame initially travelled outbye in the tailgate entries at about 183m/s, generating a pressure of nearly 6 psi. Before reaching the crossover entries, additional coal dust became involved and flame speeds accelerated to over 300m/s in all tailgate entries. As the flame continued outbye in the tailgate entries, eventually it propagated to the locations where the tailgate intersects the North Glory Mains. The flame speed dropped dramatically at this intersection due to the additional entries and the increased incombustibles in the mine dust. The flame extinguished in this direction about 11 crosscuts outbye the tailgate entries. As the flame propagated outbye in the Tailgate 1 North entries, it turned 90º to the left and entered the crossover between Tailgate 1 North and Headgate 1 North. All entries and crosscuts of the crossover were engulfed in flame, including the entries and crosscuts that turn 90º and head towards and into the North Glory Mains entries.

Flame propagation did not occur along the length of the North Glory Mains but small pockets of flame extinguished as they projected a short distance into the North Glory Mains. From the crossover entries, the explosion flame propagated into Headgate 1 North and turned both inbye and outbye. The outbye portion propagated towards and into the North Glory Mains. The inbye portion propagated inbye as far as crosscut 32. MSHA was unable to take mine dust samples any further inbye in the Headgate 1 North entries and, consequently, could not determine the extent of flame in those inbye entries. As flame entered Headgate 1 North, the destructive pressures propagated inbye with a flame speed of about 366m/s, generating over 20 psi, as indicated by damage to several monorail sections. Although flame did not enter the longwall face, pressures ranging from 7 to 14 psi did travel along the longwall from the headgate.

The flame of the coal dust explosion also travelled toward the face of the TG 22. As the coal dust explosion propagated into TG 22, explosion pressures increased to near 20 psi. Just before entering TG 22, the flame also turned 90º right and entered the crossover entries between TG 22 and HG 22. Initially, the flame resulted in large and extra-large deposits of coke. However, as the flame continued

Page 99 of 154 ACARP C19010 Extension through the crossover entries, coke was not produced. MSHA believes that the flame increased in speed as it continued through these crossover entries. Increased flame speeds decreased the duration of the flame at these locations and coke formation was not possible. This increase in speed could likely be attributed to the increased fineness of the coal dust and the lack of sufficient rock dust through these entries. The flame slowed as it turned into crosscuts. Mine dust samples taken in the crosscuts of the crossover entries included large and extra-large quantities of coke, indicating flame travel.

The flame entered HG 22 at the mouth of the section and turned 90º left and right. The portion of the flame that turned left travelled into HG 22 and propagated to the faces. The flame propagated into HG 22 at speeds approaching 460m/s generating a pressure of approximately 25 psi. Additional coal dust caused increases in the flame speed and pressure. Calculations have shown that explosion pressures were on the order of 52 to 65 psi. Pressure piling occurred as the flame and forces continued to push against the dead end of HG 22. This resulted in a reflected overpressure traveling outbye that could have reached a maximum pressure of 105 psi. The flame consumed available oxygen in HG 22 and, after reaching the faces, was unable to propagate outbye as it extinguished from the lack of oxygen. The flame that turned right travelled outbye to near crosscut 115 in the North Glory Mains, into all entries and crosscuts of the Glory Hole Mains, and turned again and propagated into all entries and crosscuts of the North Jarrells Mains, and all entries and crosscuts in West Jarrells Mains. Pressures throughout these areas averaged about 20 psi with flame speeds of over 305m/s. The flame of the explosion extinguished at the dead ends of West Jarrells mains due to lack of sufficient oxygen for continued propagation.

MSHA estimated that the explosive accumulation of methane that was eventually ignited contained approximately 8.5m3 of methane. When diluted with air to 10 percent, this volume of methane would form an explosive volume of 85m3 cubic metres. The flame of an explosion generally involves a volume that is approximately five times the volume of the initial methane accumulation. The flame from this initial methane explosion affected a volume of about 425m3, or a linear distance of approximately 43m, based on the dimension of the mine openings where the ignition occurred. The methane explosion propagated away from the longwall face. With a flame speed of approximately 91m/s, the methane explosion would have extinguished in about ½-second while generating a maximum pressure of about 4 psi. The flame zone that actually occurred at Upper Big Branch, however, was far greater than 425m3; it contained a volume of about 880,000m3. This flame zone can easily be achieved in a coal dust explosion that generates limited pressure.

Explosion damage to roof pans (Figure 39), aerosol spray containers, light bulbs and displaced structural members were modelled by various means to estimate the explosion pressures involved. The work was performed by State and Federal investigators. Figure 40 shows the extent of force from the explosion. Damage to conveyor belts varied according to the strength of explosion forces and the gauge of the C-channel used in their construction. The greatest damage occurred where they were impacted broadside by explosion forces, which deflected the structural laterally and induced a rotation to hanging belt structure that was clockwise from left hand forces when viewed along belt axis. For the most part, severe damage was confined to crosscut intersections, and damage quickly lessened away from the intersections where the coal pillars provided shelter from wind forces.

According to the MSHA report (Gates et al 2011) coking indicates the area affected by the flame of the explosion. The temperature required for coking to commence varies with the rank of coal, but is of the order of 371º C. Flame temperatures during an explosion can be nearly 982º C, however the flame may only be at each location for approximately 45 milliseconds. The amount of coking that occurs is related directly to the exposure temperature and the duration of that temperature. When objects are exposed to flame for a sufficient duration of time, heat is transferred and produces coking. However, even within the area affected by the flame, coking of the coal does not occur at all locations.

Page 100 of 154 ACARP C19010 Extension

Research on alcohol coke testing indicates that coke is found whenever coal particles are dispersed into a flame, and therefore the presence of coke is a good indication of flame travel. Coke, as measured by the Alcohol Coke Test, is found after explosions at an incombustible content of up to 80 percent. The Alcohol Coke Test indicates the quantity of coke in each sample as either none, trace, small, large, or extra-large. Large and extra-large quantities of coke are indicative of flame. The results of the Alcohol Coke Test are shown on the mine map in Figure 41.

Figure 39: Damage to pans (taken from West Virginia Office of Miners’ Health, Safety and Training, Report of Investigation presentation)

Page 101 of 154 ACARP C19010 Extension

Figure 40: Roof pan bending as mapped by investigation team (taken from Phillips 2012)

Page 102 of 154 ACARP C19010 Extension

Figure 41: Areas of coking (taken from Phillips 2012)

Page 103 of 154 ACARP C19010 Extension

A3 Sago, USA 2nd January 2006

Reference: Zipf, Sapko and Brune 2007; Griffin 2013, Brune & Goertz 2013; Gates et al 2007.

There was an explosion on 2nd January 2006 at approximately 6:26 am at Sago within an area that had recently been sealed. At the time of the explosion, 29 miners were underground (Figure 42). Twelve miners lost their lives and one was seriously injured, all others eventually evacuated. One miner (Terry Helms) died of carbon monoxide poisoning shortly after the explosion. The 2nd Left Parallel crew (12 men) were affected by a pressure wave of less than 2 psi without experiencing any known traumatic injuries. They attempted to evacuate but couldn’t due to smoke and toxic gas, so barricaded themselves awaiting rescue (11 fatalities, one serious injury). Many of those who died survived up to 12 hours before succumbing to the carbon monoxide. Randall McCloy Jr survived 41 hours before being rescued. The 1st Left crew (all survived) were affected by a pressure wave of about 2 psi, and were in a direct line with the explosion forces. They were impacted by flying debris and a rush of air, which reportedly lasted for about 8 seconds.

The ignition source for the explosion was attributed to lightning. The explosion occurred inbye the 2 North Mains seals and destroyed all ten of the seals used to separate the area from the active portion of the mine. It is thought that the methane in the sealed area was ignited by a lightning strike that induced a spark in the sealed area. It was estimated that about 4000 m3 of methane was consumed by the explosion. The explosion was believed to have generated pressures in excess of 93psi but quickly dissipated in all directions to much lower values as seen in Figure 43.

Gas monitoring at Sago relied upon the personal monitors of key underground personnel supplemented by carbon monoxide monitors on the belts and methane detectors on equipment. The first indication on the surface that something unusual was occurring underground was a carbon monoxide sensor alarming with a measured concentration of 51ppm.

The pager phone cable was found to be cut or pulled apart, especially where it traversed crosscuts, exposing it to the apparent forces from the explosion. The signal line of the trolley phone system was severely damaged in the area affected by the explosion.

There were three redundant communications systems in use at the mine. The paging phone system failed for the 2nd Left Parallel crew that perished. Although the 2-way radios did not fail, distance and line-of-sight limitations would have likely prevented the 2nd Left Parallel crew from using them to communicate with others in the mine. The third means of communication the trolleyphone system was not functional the day of the disaster and was not available to the 2nd Left Parallel crew. In 2nd Left Parallel, the barricaded miners did not try to call out because all of the communication devices were damaged. The explosion damaged wiring and several pager phones. Pager phone communication to 2 North Mains and 2nd Left Parallel was not possible. It is difficult to say whether the trapped miners on 2nd Left Parallel would have been able to escape the mine, had they been able to communicate with other sections and the surface. Communication would have allowed rescue teams to know where they were and possibly reached them prior to the 11 fatalities. It is also possible that others might have been able to talk them through a means and avenue of escape. The 2nd Left Parallel section crew was inbye the explosion; all others were outbye the explosion.

After the explosion even though monitoring data from some carbon monoxide sensors was no longer available on the surface, these sensors could be heard alarming underground on the mine phone indicating the sensor was working but communication with surface compromised.

Page 104 of 154 ACARP C19010 Extension

Sealed area

Figure 42: Location of survivors and fatalities at SAGO mine (taken from MSHA report).

Immediate entry to the mine was made by mine management officials (without appropriate equipment including gas detection) to assess the situation. When told by someone underground at the time of the incident that he thought there had been an explosion the reply was that there could not have been an explosion and questioned how it could have happened. They found damaged ventilation controls and in an effort to reach the missing miners (no contact with the 2nd Left Parallel crew had been made since the event), they attempted to restore ventilation, using temporary ventilation controls.

The remaining power to the underground was de-energised, the plan being to de-energise the Atmospheric Monitoring System (AMS) as well. The AMS system was equipped with a battery backup that maintained power to the system when there was a loss of mine power. The system would remain energised until it was manually disconnected. That was not done at this time, and the AMS remained energised until discovered by mine rescue teams during exploration forcing the withdrawal of teams. Supplies including gas detectors were sourced from the surface and efforts to restore ventilation were continued. After some progress they decided that there was a potential of another explosion resulting from their actions, since they were forcing fresh air into areas where explosive gases might be present. One of the detectors they had was at its maximum reading and was in malfunction mode. They withdrew to the surface.

Federal (MSHA) and state (WVMHS&T) agencies responded to the accident. At the time most of MSHA’s incident response equipment and man power was responding to a fire at West Elk mine. Mine rescue teams were organised, a command centre was established, and a rescue effort was initiated.

Page 105 of 154 ACARP C19010 Extension

Figure 43: Explosion overpressures following explosion (taken from MSHA report)

Gas concentration measurements were made at mine openings and at 8:40 am measurements from the No. 1 Drift Opening returned 47 ppm carbon monoxide, 0.0% methane and 20.4% oxygen. At 9:10 am a mine worker and a West Virginia Office of Miners' Health Safety and Training (WVMHS&T) mine inspector each took readings with their own instruments, one returned 23ppm carbon monoxide, 0.0% methane and 20.3% oxygen and the other 50 ppm carbon monoxide, 0.0% methane and 20.6% oxygen. By 10:47am, measurements made by MSHA inspectors returned offscale (>500ppm) ppm carbon monoxide, 0.8% methane and 19.8% oxygen.

At 12:17 pm the No. 1 Drift Opening indicated carbon monoxide in excess of 1,999 ppm (maximum reading on instrument used). At 12:20 pm an imminent danger order was issued because of the extremely high carbon monoxide levels detected in the No. 1 Drift Opening. All non-essential personnel were withdrawn from the pit and the surface buildings. Mine rescue team members were mobilised to conduct the sampling of the No. 1 Drift Opening. At about 1:00pm elevated carbon monoxide concentrations were measured outside (330ppm) and inside (130ppm) surface buildings. MSHA personnel directed that all office and nonessential personnel leave the mine site.

A formal monitoring plan was developed requiring two mine rescue team members wearing full apparatus to monitor the gases exiting the mine with results reported to the command centre. Two backup mine rescue team members wearing full apparatus were required to stand at the edge of the pit. Logs of manual carbon monoxide measurements made during the response were found to be inaccurate and MSHA had to provide training on use of hand held gas detectors. At approximately 1:30pm levels dropped in the offices and non –essential personnel were allowed back.

Page 106 of 154 ACARP C19010 Extension

CONSUL sent a gas chromatograph and personnel to operate it to assist with monitoring; it was calibrated and operational by 3:00 pm.

Personnel from MSHA’s Ventilation and Physical and Toxic Agents Divisions left Pittsburgh, Pennsylvania at around 2:00 pm equipped with infrared and electrochemical gas analysers, several thousand feet of 3/8 inch PVC tubing, vacuum pumps, four handheld permissible radios, a gas chromatograph, and the associated computers needed to operate the gas chromatograph and analyse the gas results.

Entry into the mine was delayed because of the elevated levels of carbon monoxide and methane. Preparations were started to drill a borehole into the 2nd Left Parallel section for sampling and communications purposes. A downward trend in the levels of dangerous gases from the drift openings was observed and after 4:00 pm the mine operator submitted requests to send mine rescue personnel into the mine. MSHA and WVMHS&T denied these requests based on the carbon monoxide exiting the mine still being too high reflecting a substantial risk of fire and the possibility of another explosion.

Readings continued trending downward (confirmed by gas chromatograph) and while they were still at dangerous levels, it was determined that they were low enough to allow rescue efforts to commence. At 4:55 pm a plan for the start of exploration was submitted and approved by MSHA and WVMHS&T. There was a high degree of risk associated with this decision and it was discussed with all parties including the mine rescue teams before they started underground.

At 5:15 pm the carbon monoxide at the No. 1 Drift Opening was still dangerously high at 1,740 ppm, but the downward trend had been continuing for several hours. At 5:25 pm a rescue team entered the mine through the fan house and proceeded inbye exploring the mine.

MSHA’s Ventilation and Physical and Toxic Agents personnel arrived at the mine site at approximately 5:15 pm. A sampling line had to be extended to the No. 1 Drift Opening since a previously installed line was plugged. By 7:20 pm (~13 hours after the explosion) MSHA’s was set- up and monitoring the mine atmosphere exiting No. 1 Drift Opening. Gas trends were generated with measurements recorded about every 15 minutes during the rescue operation.

The AMS which was thought to be de-energised had remained energised which was identified by rescue teams and resulted in an order at 2:40am to retreat from mine. By the time the AMS was de- energised, the borehole into the 2nd Left Parallel section was an hour from holing through. Re-entry was delayed due to risk of holing through initiating an explosion. The borehole intersected workings at 5:53 am and results for collected sample returned 1,052ppm carbon monoxide and 20.4% oxygen. Rescue teams re-entered the mine more than four hours after the order to withdraw based on the energised AMS was given.

At 5:20 pm the rescue teams located the first victim. Not long before midnight the other 11 victims and sole survivor were found. By about 1:00 am the survivor had been transported to the surface and placed in an ambulance. The command centre and the rescue teams discussed the recovery of the deceased miners. Normal procedure would be for the area to be re-ventilated prior to any recovery, to limit the exposure of rescue team personnel to any hazards. However, rescue team members volunteered to re-enter the mine and, under apparatus, recover the deceased miners. By around 9:22 am the victims had been recovered and transported to the fresh air base. Shortly thereafter, the mine rescue teams and the victims were transported to the surface.

During mine recovery, monitoring was conducted at the No. 1 Drift Opening, Borehole No. 1, and eventually, through a series of additional boreholes. Monitoring continued until the situation was deemed stable and determined safe for miners to re-enter the mine and restore ventilation. On January 5th, Borehole No. 2 was completed into the track entry at 31 Crosscut, No. 5 Belt. It was used to

Page 107 of 154 ACARP C19010 Extension monitor air quality in 1st Left. Boreholes No. 4 – 7 were drilled into the 2nd Left Mains. Borehole No. 4 was started on January 6th and completed on January 8th. Borehole No. 7, the last borehole drilled was started on January 17th and completed on January 19th.

The results from analysis of the mine atmosphere remained favourable throughout the mine recovery. On January 21st, 2006, with ventilation established to the boreholes at the inbye end of the 2nd Left Mains, mine rescue/recovery teams entered the mine to examine and re-establish ventilation following the approved plan developed by the operator.

Page 108 of 154 ACARP C19010 Extension

A4 Willow Creek, USA, 31st July 2000

Reference: Griffin 2013; Brune & Goertz 2013; Brune 2013; McKinney et al 2001

Four explosions occurred at the Willow Creek Mine over 31 minutes beginning 11:48pm 31st July 2000. Figure 44 shows the area affected by the explosions. Most likely the first explosion was initiated by a roof fall in the worked-out area of the D-3 longwall panel gob igniting methane and other gaseous hydrocarbons present. This also resulted in an ongoing fire. Personnel remained on the D-3 longwall section to extinguish the fire near the base of the shields, believing that just a roof fall had occurred. Eventually, liquid hydrocarbons became involved in the fire. Conditions worsened in the face area just prior to the second explosion; the fire was burning more intensely in the gob and the fire roaring behind the shields could be heard. Two closely spaced explosions occurred at approximately 11:55 pm. A fourth explosion occurred at 12:17 am on August 1. Two fatalities occurred as a result of the second and third explosions. There were also eight miners injured. A Personal Emergency Device (PED) system was in use at the mine, and was instrumental in alerting miners working in active and remote areas of the mine to evacuate following the initial explosion. These miners all safely exited the mine.

The communication system remained operational during the emergency with miners able to communicate with the surface using the pager phone on the section. The AMS attendant observed communication failures with many of the sensors surrounding the D-3 section following the first explosion, although some remained operational. Elevated carbon monoxide readings occurred in the bleeder entries and in the D-3 No. 1 headgate entry. The monitors near the headgate bleeder connector regulators experienced a communication failure. Data from the carbon monoxide monitors near the tailgate regulators at the bleeder entries indicated off scale concentrations (greater than 50 ppm) shortly after the explosion. Data from the carbon monoxide monitor at MPL B2 showed that the concentrations began to increase approximately 21 minutes after the explosion. Within an additional two minutes the readings were in excess of 50 ppm (off scale). Data from the carbon monoxide monitors in the No. 1 headgate entry outbye the face revealed that elevated concentrations of carbon monoxide occurred at the monitor locations near the longwall face. Data from each outbye sensor also showed elevated concentrations.

The fire provided the ignition source for these subsequent explosions. A mine rescue and recovery operation was conducted and all remaining survivors and the deceased were recovered by 4:00am. The mine surface openings were sealed at approximately 10:30 am on August 1with no plan to re- enter or reopen the mine (although mine was recovered in Oct 2001).

Men were sent to the mine return portals to monitor gases at approximately 1:00 am. After checking the three return portals, twice each, they returned to the command centre to report their findings. At approximately 1:30 am men exited the mine in the D-3 section mantrip. They provided information concerning at least two of the injured miners still underground. A decision was made to send a mine rescue team to the D-3 longwall section. At one point the team encountered light smoke and 4.6 to 4.9 % methane. Elevated methane concentrations and light smoke were present on the face; however, there were no visible flames. The rescue team located and rescued injured miners and located and retrieved the two fatality injured miners.

Page 109 of 154 ACARP C19010 Extension

Figure 44: Area affected by explosions (taken from McKinney et al 2001)

Although difficult to determine as no re-entry was conducted for investigation purposes as little as 1.4m3 of methane diluted to 6.5% was involved in the initial explosion generating pressures estimated at approximately 5psi near the origin, 3psi exiting the headgate into the bleeder entries,2psi reaching the tailgate bleeder regulators and 0.5psi across the longwall face.

Evidence indicates that forces generated during the second explosion were of a lower overall magnitude than those of the first explosion. However, injuries were more significant due to the proximity of miners to the origin of the second explosion. MSHA considered that the third explosion was the most powerful explosion. As with the first explosion, obstructions prevented the full thrust of the explosion from propagating outbye along the maingate fringe of the goaf. No miners underground recalled the fourth explosion; it was identified by a spike on the fan chart.

Page 110 of 154 ACARP C19010 Extension

A5 Endeavour Colliery, NSW 28th June 1995

Reference: Anderson, Urosek, and Stephan (1997), NSW DMR MDG No. 1007 (1996)

On June 28, 1995 at approximately 9:50 am an explosion occurred in the 300 Panel at Endeavour Colliery (see Figure 45). A large goaf fall in the goaf occurred prior to the explosion, propelling methane into the mine workings. The men at the face reported a windblast from the roof fall. The methane mixed with mine air and some of the methane was ignited. The ignition source was not definitively identified, however expert investigation identified that it was most likely located in 21 cut through between 1 and 3 headings and Anderson Urosek and Stephan (1997) consider the shuttle car cable connector that was in a non –flameproof condition as the most likely source of ignition. There were 30 miners underground at the time, all evacuated safely.

The explosion was consistent with approximately 6 m3 of methane mixed with air to about 6 -7% composition. This would indicate that the total volume of the flammable atmosphere that was ignited was only about 100 m3 generating a peak flame speed of 122 m/s and pressures of up to 4psi in the 300 Panel and about 0.5psi outbye. There was significant damage to all brattice stoppings within the panel and also to plasterboard stoppings (in poor condition prior to explosion) in 8 West; some 2 km from the 300 Panel. The overpressure was sufficient to raise dust in the 8 West headings. Overpressures of similar magnitude to the explosion may have been generated by the roof fall. One miner had the radio control unit he was carrying, a reasonably heavy object, torn from his hands by the wind blast and others were knocked to the ground. A shuttlecar driver had his helmet blown off and the bracket holding his cap lamp to the helmet was torn off. Some of the workers were knocked down by the effects of the explosion but none suffered severe physical trauma.

Communications remained operational with the surface advised via telephone that the explosion had occurred. The phone in the crib room was also heard ringing by two disorientated miners that allowed them to establish where they were in low visibility.

The colliery had a six point tube bundle system capable of analysing for methane and carbon monoxide which continued to provide results post explosion. The nearest sampling point to the 300 Panel was located some 5 km outbye. There is no reference to damage to the tube bundle identified on re-entry. The distance between the closest sampling point and the site of the explosion along with low overpressures generated are likely to be reasons for its unaffected operation (although locations less than ideal for interpretation post event hence need for boreholes).

From June 29 through July 10, personnel did not enter the mine with samples analysed and collected from the six points using the tube bundle system during this period. Additional information was required and boreholes were drilled into the 300 Panel with samples collected and analysed using the Department’s mobile laboratory. On 10th July the atmosphere at the underground monitoring locations and at the boreholes appeared stable with minimal risks of combustion occurring underground. Plans were developed to begin the recovery of the underground workings including advancing an existing tube bundle monitoring location.

Page 111 of 154 ACARP C19010 Extension

Figure 45: Endeavour Colliery (taken from MDG 1007)

Page 112 of 154 ACARP C19010 Extension

A6 Moura No.2 Mine, Qld 7th August 1994

Reference: Stephan, Urosek and Giardino (1994); Golledge et al (1995); Windridge (1995).

At approximately 11:35 pm on Sunday 7th August 1994 there was an explosion in the Moura No. 2 Mine that claimed the lives of eleven of the twenty one men underground at the time. The eleven men who died were all deployed in the southern side of the mine, in the region of 5 South and 512 seals (see Figure 46). The under manager was talking on the phone to a section electrician in 5 South at the time of the explosion. The phone was cut off simultaneously to the explosion. Those that escaped were in the northern side. The mostly likely scenario was determined to be that the flammable atmosphere of methane gas in the recently sealed 512 panel was ignited by a concealed spontaneous combustion event within the panel.

A second more massive explosion occurred at 12:20 pm on Tuesday 9th August and the mine was sealed from the surface without any re-entry to the mine. The bodies have never been recovered. Even though it was not possible to re-enter the mine after the two explosions, there was comprehensive gas monitoring data available pre and post explosion from a combination of sources, including tube bundle data that enabled a detailed analysis to be undertaken.

The volume of methane involved in the explosion was estimated to be less than 70 m3, based upon the effects on the surviving miners. The pressure wave that the crew in 5 south experienced would have been approximately 4 psi, with the original explosion estimated to only be about 8 psi. Some ventilation controls within 5 South were damaged by the explosion but not elsewhere.

Page 113 of 154 ACARP C19010 Extension

Figure 46: Moura No. 2 Mine plan

The impacts of the explosion on the men in the Northern part of the mine varied from nothing discernible to being bowled over by the pressure wave. A worker was knocked over in 1 NW. Brattice here shook but did not break. Another worker in 1 NW near the feeder Xcut felt his ears pop and brattice vibrated, airways filled with dust, followed a minute later by smoky green brown haze which had an acrid taste and burnt the eyes. He was able to phone the surface from the crib room phone. A number of other miners in the vicinity had a similar experience. Another worker in 1 NW walking toward the miner got blown into the rib, and then his ears popped, he was able to retreat to the crib room. A miner walking the belt road Dip 1 boot end, was hit by a blast of air and dust knocking him onto the “seat of his pants”. He stopped the belt near No. 3 Xcut. The crew escaping from 1 NW found that initially visibility was 5 -10 metres then dropped to about 1 metre, they used the water pipes hanging from the roof to navigate their way to the main dip. The ten men in the northern part of the mine were able to escape through smoke.

Page 114 of 154 ACARP C19010 Extension

No damage was done to either fan although an explosion relief door was blown about four metres away from the fan housing. The explosive forces apparently caused the detachment of each of the Victaulic pipe range systems connected to the surface boreholes at 520 and 5 South which became subsequent sample points after the explosion. Approximately 50,000m3 of methane which was being drained from the mine each day was thereafter a contaminant of the underground mine atmosphere.

As part of the subsequent investigation (limited to the surface) boreholes were sunk in order to attempt the collection of photographic evidence in the form of video recording. These were located outside seals in 1 Heading of 511 Panel (a brick seal) and 2 Heading of 512 Panel (a Tecrete seal). A borehole attempted in the vicinity of a seal in the 5 South panel proved unsuccessful. Evidence obtained from the boreholes indicated the following:  both the 511 and 512 seals originally in the vicinity of the boreholes were absent;  there appeared to be seal debris outbye the 512 seal at a short distance from the original seal location;  there appeared to be bricks from the 511 seal inbye the original seal location; and  a number of roof bolts that had provided reinforcing for the 512 seal appeared to be bent outbye;  there was no evidence of the presence a crib table and Tecrete batcher (mixer) which had been observed in the vicinity of the 512 seal by Blyton during the afternoon shift of Sunday 7th August; and  a number of props were still standing in the vicinity of both seals.

According to Windridge (1995) “these observations indicate an explosion occurred within the 512 Panel. The proximity of seal debris to the 511 and 512 seals may indicate those seals to have been destroyed by a relatively weak explosion, and, by inference, the first explosion. These indications are, however, by no means certain and do not preclude the possibility of explosion initiation external to the 512 Panel.”

The second explosion that occurred just over 36 hours later was more powerful than the first. The plume from the second explosion exited the portals well in advance of exiting the fanshaft. Material was ejected from the fanshaft – steelplate, steel inspection ladder from fanshaft and wire cable. The plume from the fan was estimated to reach 300 m into the air, it was observed to gush out under extreme pressure, black initially from 30-40 seconds, then white for 20 seconds then a grey for about 50 seconds. The ducting linking the mine fan to the shaft was destroyed, some sections being reportedly launched into the air. Large volumes of dust, smoke and gases, including carbon monoxide, were forcefully emitted from each of the entry tunnels into the mine. The surface facilities including the emergency control room, the gas monitoring room and the bathroom, although being over 250 metres away and to the side of the mine openings were covered with dust. The prevailing wind brought products of combustion from the mine to the surface facilities. Carbon monoxide levels around the surface facilities rose to over 400 ppm and required the use of self rescuers and immediate evacuation of the area. Contamination of the atmosphere around the surface buildings continued after the main blast with smoke continuing to issue from the underground tunnels. This made the Emergency Control Room unfit to use and, as a result, gas analysis equipment was relocated to a safe position several kilometres away in the open cut mine office complex. Following the second explosion it was considered that there could be no survivors and to prevent further explosions the mine was sealed.

Page 115 of 154 ACARP C19010 Extension

Moura No. 2 Mine had a tube bundle system and at the time of the explosion eleven underground points as well as the room the system was located in on the surface were being monitored. The monitored locations are given in Table 6 and shown in Figure 47.

Table 6: Tube bundle sampling locations, Moura No. 2 Mine

Tube Tube Name Location in Mine Number 1 4 SOUTH 4 South Return

3 FAN NTH RTN Main Fan North Return 4 FAN STH RTN Main Fan South Return 5 512 SEALS Behind 512 seals. Moved from 512 Bottom Return. 6 5 STH BOTTOM 5 South Bottom Return 7 5 STH TOP 5 South Top Return 8 1 NW RETURN 1 North West Return 9 17 CT DIPS 17 Cut Through Dips 14 PUMP ROOM Mine monitoring room on surface 16 512 TOP RTN 512 Top Return. Moved to 510 South Return 18 510 NTH RTN 510 North Return 19 DIPS NTH RTN Dips North Return

Page 116 of 154 ACARP C19010 Extension

2 NTH WEST

5 NTH 5 NTH WEST

2 NTH

1 NTH WEST 3

(1) 3 NE 8 (11)

(3) 4

R (12) 18 6 STH 510 D

D 19 5 (2) D (7)

D R 16

R D R

D D 1 2 NE D

D D

511 D D D M

M D

R

R

D (10)

(6) D R D D 9

D 512 D 3 STH (8)

D 6 D

D D R 2 STH

7 D D (5)

D

(4) D D D 4 STH A

D 5 STH D

(9) D

D 1 STH D

D 4 STH D

D

D

D

D

D

4 STH B

Figure 47: Sampling locations Moura No. 2 Mine

Page 117 of 154 ACARP C19010 Extension

The lengths of the tubes varied depending on the sampling location with lag times varying between 10 and 73 minutes. Moura No. 2 performed integrity testing of the tubes by introducing gas of known concentration at the sample point and comparing with the concentration measured at the surface. The time taken for the sample to reach the surface (lag time) was also calculated during this testing. Integrity testing was performed the day of the explosion with results displayed in Table 7.

Table 7: Integrity testing and draw time determination, Moura No. 2 Mine

Tube Location Name Reported Draw Time (minutes) Observed Number (Survey 7/8/94) Carbon Monoxide (ppm) 1 4 SOUTH 10 43

3 FAN NTH RTN 17 44.6 4 FAN STH RTN 13 41.6 5 512 SEALS 44 44.2 6 5 STH BOTTOM 43 43.5 7 5 STH TOP 40 43.6 8 1 NW RETURN 18 29.1 9 17 C/T DIPS 10 43.7 14 PUMP ROOM 0 N/A 16 512 TOP RTN 73 44.3 18 510 NTH RTN >10 hours 23.8 19 DIPS NTH RTN 14 40.7

According to Golledge et al (1995) ‘Most points show 40 to 44 ppm of carbon monoxide indicating little or no leakage. Point 8 (1 North West Return) and Point 18, (510 North Return) show considerably lower values indicating contamination by air leaking into the tubes at some point other than the sampling point’. The tube at Point 18 also returned a lag time greater than ten hours for the leak test and any data from this tube could not be considered reliable.

The tubes were unprotected but installation was reportedly performed to a high standard. Analysis of the pressure drop within the tube bundle gas sampling lines coupled with the results being obtained from the monitoring system indicated that some tubes were intact in their original locations whereas others had clearly been damaged and were reporting from completely different points in the mine. This prompted the drilling of four boreholes;

 512 goaf area  vicinity of the 5 South face  a 5 South/510 roadway junction, and  a Main Dips intake airway.

The boreholes were to ensure gas samples were available from known locations underground. The 512 goaf hole was reported to be blowing outwards and having a bitumen smell while the other holes

Page 118 of 154 ACARP C19010 Extension exhibited negative pressure consistent with the main fan operating. Samples were manually collected from these boreholes and analysed using the onsite gas chromatograph.

Difficulties were experienced during early borehole sampling operations. Problems included air ingress, methane ingress from overlying coal seams into damaged or uncased bores and setting the borehole sampling tube into the pit roadway. The latter caused some significant problems when excess lengths of tube were lowered into new boreholes. This resulted in some instances of inadequate purging of the sample tubes. Borehole G-512B did not intersect the intended roadway. Also a borehole attempted in the vicinity of a seal in the 5 South panel (for video evidence) proved unsuccessful.

Although not all tubes were sampling from intended locations, results from the system and subsequent gas chromatograph analysis of samples collected via the tube bundle system provided valuable information on the status of the underground environment. Subsequent conversations with the Chief Inspector of Coal Mines at the time of the incident revealed that the tube bundle results were the reason rescue teams weren’t sent into the mine, and that the levels of methane measured in the mine following the initial explosion were not expected (it wasn’t thought that seals other than 512 would be compromised and release methane into the workings). This decision potentially may have saved the lives of rescue teams as there was a second explosion. It would appear without the tube bundle results rescue teams would have been deployed.

The tube bundle results for Fan North Return (Tube 3) and Fan South Return (Tube 4) showed very high levels of methane (approximately19%) and greater than 1000 ppm carbon monoxide. It was deemed important by the incident management team to verify these results and a sample from the fan was collected (from the surface) and analysed using the gas chromatograph. Calculations were made of the expected concentrations at the fan based on results from the two tubes sampling from each of the roadways leading to the ventilation shaft and approximate air flow in each. This allowed comparison with the sample collected at the fan from the surface. The results showed good general agreement between the two sets of analysis, especially methane and carbon monoxide. On this basis, and considering the remoteness from the 510/5 South area, it was determined that these points were unaffected physically by the explosion and the results from the tube bundle system from these two points were reliable. The subsequent Simtars’ investigation supported this finding based on changes to lag time determined by when first products of the explosion were seen.

Samples taken from the boreholes and the damaged tube bundle system revealed the presence of explosive mixtures of gases and very high carbon monoxide concentrations in several places in the mine continually up until the second explosion. This was one of the factors which prevented the deployment of rescue teams underground and necessitated keeping personnel away from the mine entries.

As part of its investigative report Simtars continued evaluation of likely monitoring locations following the initial explosion and reported the following.

To determine when the first signs of post explosion gases reported to the surface for each tube the tube bundle data was reviewed. The time was adjusted for this and the previous sample by the lag time determined for the tube on the day of the explosion. It makes sense that the explosion had to occur between these samples (based on the fact that combustion gases were distributed throughout the mine almost immediately). The clock on the computer logging the data was found to be seven minutes slow so times were further adjusted for this. This gives an estimation of the time of the explosion based on the monitoring results. These results are displayed in Table 8 which shows there were several tubes that returned an explosion time corresponding with the observed time of 11:35 pm, and others that did not.

Page 119 of 154 ACARP C19010 Extension

Those that did agree were labelled as Group 1. This indicates that there was no change in the draw time of these tubes as a result of the explosion. Group 1 tubes had a common window of 23:28 to 23:37 corrected for lag times and slow computer clock. Group 1 tubes were all located well away from where the explosion most likely occurred.

Group 2 tubes were those that had the explosion occurring before or much later than it actually did (based on the original lag times). The tubes that returned an explosion time before it actually happened (Tubes 5, 7 and 16) were likely to have had a change that resulted in a reduced lag time, i.e. likely to have been severed. The lag time for these tubes was adjusted based on the known explosion time. Similarly for Tube 6 where results indicate the explosion occurred much later than it really did, the lag time was increased indicating that the tube had been restricted. These results alone did not indicate from where the tube was sampling. Table 9 shows the lag times calculated for tubes post explosion that align the results first seen at each tube with the reported time of the explosion.

Page 120 of 154 ACARP C19010 Extension

Table 8: Time of Explosion Based on Tube Bundle Results

POINT LOCATION ADJUSTMENT FOR ERROR IN GROUPING TIME OF COMPUTER CLOCK TIME OF FIRST REPORTED EARLIEST LATEST (ADD 7 MINUTES) PREVIOUS RELIABLE LAG TIME OF TIME OF SAMPLE CHANGE TIME (MIN) EXPLOSION EXPLOSION

COMPUTER CLOCK EARLIEST LATEST

1 4 South 23:31 23:45 10 23:21 23:35 23:28 23:42 1

3 Fan North Return 23:32 23.47 17 23:15 23:30 23:22 23:37 1

4 Fan South Return 23:33 23.48 13 23:20 23:35 23:27 23:42 1

5 512 Seals 23:49 00:05 44 23:05. 23:21 23:12 23:26 2

6 5 South Bottom 02:04 02:22 43 01:21 01:39 01:28 01:46 2

7 5 South Top 23:37 23:52 40 22:57 23:12 23:04 23:19 2

8 1 North West Return 23:38 23:53 18 23:20 23:35 23:27 23:42 1

9 17 CIT Dips 23:26 23:40 10 23:16 23:30 23:23 23:37 1

16 512 Top Return 23:57 00:14 73 22:44 23:01 22:51 23:08 2

19 Dips North Return 23:30 23:44 14 23:16 23:30 23:23 23:37 1

Probable Explosion Windows After Correction from Computer Time to Real Time for Group 1 Points (See Text)

23:28 23:37

Page 121 of 154 ACARP C19010 Extension

Table 9: Post Explosion Lag Times

POINT LOCATION EXPLOSION WINDOW FOR ADJUSTED WINDOW TO MATCH ADJUSTMENT PRE-EXPLOSION POST-EXPLOSION PRE EXPLOSION LAG TIME GROUP 1 RESULT REQUIRED TO PRE- LAG TIME (MIN) LAG TIME (MIN) EXPLOSION LAG EARLIEST LATEST EARLIEST LATEST TIME (MIN)

5 23:12 23:28 23:25 23:40 -13 44 31 512 Seals

6 5 South Bottom 01:28 01:46 23:23 23:41 +115 43 158

7 5 South Top 23:04 23:19 23:24 23:39 -20 40 20

16 512 Top Return 22:51 23:08 23:23 23:40 -28. 73 45

Page 122 of 154 ACARP C19010 Extension

17 Cut-through Dips (Tube 9) and 4 South (Tube 1) tubes were located in close proximity but were in separate air splits. 17 Cut-through Dips monitored a minor air split used to ventilate the southern flank of the main intake roads, whereas 4 South monitored the return from the 4 South district. Because of the distance from the site of the explosion and the results of the analysis on the two fan return monitoring points, it was determined that it was unlikely these points had been severed. It was, however, concluded that the 4 South tube had been relocated into the same air split as the 17 Cut- through Dips tube as the tube bundle analysis for these two points were almost identical after the explosion (Figure 48). Had the sample points remained in different air splits, this would be almost impossible. For Tube 1 to be in the same air split would only have required movement of no more than 10 metres. Evidence from survivors identified a high air velocity sufficient to blow a man over in the main belt road for a significant period of time after the explosion. This could have been sufficient to blow the tube out of the 4 South Return air split, around the corner into the Dips North Return air split. Post explosion measurements showed very high concentrations of methane (up to 43%) between 11:40 pm on 7 August 1994 and 4:00 am on 8 August 1994. There were also high levels of carbon monoxide (exceeding 1 000 ppm at times) and carbon dioxide and reduced oxygen levels.

Figure 48: Post Explosion Results Tubes 1 and 9

The Dips North Return tube (Tube 19) monitored an air split used to ventilate the area of gob stoppings to the north of the main dips and east of the fan. It was the most remote point of the tube bundle system from the 512 area and as such deemed intact and unmoved after the explosion. These post explosion results show similar trends to Tube 9 at 17 Cut-through Dips, but with reduced methane and carbon dioxide peaks. Methane reached a maximum of about 10% at the same time as oxygen reached a minimum of about 17%. Carbon monoxide exceeded 1000 ppm for nearly 2 hours.

The 1 North West Return Tube (Tube 8) was located in the main northern return air split about 150 metres west of the upcast shaft. For reasons similar to those applying to Tube 19, it was deemed reasonable to expect that Tube 8 was also undamaged after the explosion. Further support for this came from a comparison with results from Fan North Return (Tube 3) as the majority of the air flow into the Fan North Return came from that passing North West Return. The complicating issue associated with this point was the leakage observed during the integrity tests performed on the day of the explosion.

The remaining tubes were all located closer to the site of the explosion where damage could be expected. The assumption that these were damaged was supported by the work done in calculating the explosion time based on tube bundle results. No evaluation of Tube 18 was made due to the issues with its performance identified during the pre-explosion integrity testing.

The investigation team used the work of Litton (1983) which reported the lag time as proportional to the square of the length tube. Knowing the pre and post explosion lag times, the severed length of tube was estimated. For the 5 South Top Return tube it was calculated that the tube was severed 900 metres from the original sampling location. This would mean the tube was sampling back at the 4 South Panel at about 13 or 14 Cut-through. There was a tube bundle junction box in this location and it was considered likely that it would be more susceptible and therefore more likely of damage than individual tubes.

It was also considered likely that the other tubes assumed to be severed were damaged in this location, but some pinching or restriction existed due to the longer draw times than Tube 7. The post explosion Page 123 of 154 ACARP C19010 Extension data was used to support this proposition with these points showing very similar trends but a slight difference in timing attributed to different lag times, particularly for Tube 6. It was considered that the similar results (Figure 49) indicated the tubes were sampling the same atmosphere. For this to be the case the tubes had to be severed somewhere outbye of 17 Cut-through 5 South as this is the point that all the tubes ran to, and from here were run together to the surface.

A summary of the likely sampling locations post explosion is given in Table 10.

Figure 49: Post Explosion Results Tubes 5, 6, 7 and 16

Table 10: Likely sampling locations following initial explosion

SAMPLING ORIGINAL STATUS POINT LOCATION

1 4 South Relocated to same air split as Point 9: 17C/T Dips

3 Fan North Return Unaffected

4 Fan South Return Unaffected

5 512 Seals Severed in 4 South Level

6 5 South Bottom Returns Severed in 4 South Level and severely pinched

7 5 South Top Return Severed in 4 South Level

8 1NW Return Unaffected (integrity doubtful)

9 17 C/T Dips Unaffected

16 512 Top Returns Severed in 4 South Level

19 Dips North Return Unaffected

The post incident investigation also revealed an issue with possible cross contamination between samples. The 5 South Bottom Return tube (Tube 6) returned persistently high levels of carbon monoxide and high oxygen post explosion. Tube 6 was identified as being pinched and therefore had a very low flow rate through the tube. It was proposed, and confirmed by testing, that a sample location could be contaminated by gas from other sample locations when the tube in question had a restricted flow.

No records are available on the vacuum pressures pre and post explosion for any of the tubes. Page 124 of 154 ACARP C19010 Extension

During the post incident investigations conducted onsite it was identified that the oxygen paramagnetic analyser may have been damaged/influenced by the second explosion with significant errors associated with measurement that were not apparent prior to or following the initial explosion.

Even though the system was compromised by the explosion, valuable information can be gained from the results. This information is critical to incident management teams responsible for emergency response. Mine sites should consider the resources required (people with the necessary skills and knowledge) to make these evaluations and determinations to support the decision making process. It must be pointed out that the work conducted by Simtars in reviewing this data was conducted post the event and without the pressure that goes with responding such events. It took an appreciable amount of time and knowledge and it is not envisaged that mine sites with the current level of preparedness would be in a position to make such evaluations and determinations within a timeframe that could be applied to results generated during response to the incident. This is an area needing improvement within the industry.

Page 125 of 154 ACARP C19010 Extension A7 Moura No.4 Mine, Qld 16th July 1986

Reference: Lynn (1997); Green and Upfold (1987).

At about 11:05 am on 16th July 1986 an explosion in Moura No. 4 Underground Mine killed 12 men who had been extracting pillars in the Main dips section. Eight other men who had been underground at the time, survived. Five men in the 3 South area were advised by telephone to make their way to the surface. The other three were out of contact but made their way to the surface. It wasn’t until carbon monoxide was detected that it was identified an explosion had occurred. A roof fall had been anticipated and initial observations were attributed to wind blast associated with a fall.

The inquiry found that a roof fall had occurred in the goaf and that the wind blast from the fall blew a mixture of methane, air and coal dust into the working area. An explosive atmosphere developed in the working area and in particular around the deputy’s flame safety lamp. An ignition occurred which created a violent explosion throughout the section. The explosion was quenched by the presence of a water barrier in the belt roadway and substantial quantities of water in swilleys in other roadways.

Observers on the surface saw a large dust cloud emerge from the mine portals. There was an interruption to the power underground and almost immediately damage was reported to the mine fan ducting, the internal baffles being blown some 25m.

Rescue efforts commenced immediately, the conditions were found to be very dusty and a strange smell was present. A high concentration of carbon monoxide was detected at the mine portals before the fan was restarted. The rescue team that penetrated furthest was halted by blast debris and poor visibility. The team found that carbon monoxide was high and methane levels were rising.

At 12:25pm near 8 ct, visibility was about 20m with greater than 700 ppm carbon monoxide present and no methane. At 1 pm it was reported that the “Taj Mahal” was completely destroyed (there had been a roof fall previously to a height of about 8m in the supply road and a major steel structure known as the “Taj Mahal” was erected to protect persons travelling underneath), carbon monoxide was detected at greater than 700 ppm along with 2.2 % methane. At this time more than 3000ppm carbon monoxide was detected in the fan portal. There had been extensive damage to the ventilation system.

There was significant damage within the panel but the energy of the explosion had largely dissipated by the time it had reached the exit to the panel. The shuttle car was found in No.3 heading hard against the rib, in an unnatural position. A diesel rover was found inbye 26ct in No. 4 heading against the rib. The bonnet was blown off, the vehicle was tilted at 10o and the roof of the vehicle had been hooked onto a roof bolt. The seats were extensively burned. The blast caused massive damage in the belt roadway, moving the conveyor frame sideways at 24 ct and breaking the frame at about 23 ct moving the entire frame outbye 22 ct. The water barriers situated between 23 and 24 ct were blown out with the frames found at about 20 ct. Water in the headings 1, 2 and 3 at the swilleys had been moved up the dip as far as 19 ct. All stoppings inbye 22ct were destroyed. All brattice stoppings inbye 22 ct were burnt. It is estimated that in excess of 10,000 m3 of roof material fell. The explosion displaced the explosion panels on the ventilation ducting at the fan. Green and Upfold reported the extent of the flame, heat damage and over pressure as shown in Figure 50- Figure 53.

Page 126 of 154 ACARP C19010 Extension

Figure 50: Moura No. 4 Extent of Flame (taken from Green and Upfold 1987)

Figure 51: Moura No. 4 Heat Damage on Plastic Materials (taken from Green and Upfold 1987)

Page 127 of 154 ACARP C19010 Extension

Figure 52: Moura No.4 Estimated Explosion Pressures (taken from Green and Upfold 1987)

Figure 53: Moura No.4 Major Explosion Airflows (taken from Green and Upfold 1987)

Moura No. 4 Mine had a six point tube bundle system monitoring carbon monoxide and methane that was working effectively up to the time of the explosion. The system was however powered by the main electricity supply to the mine which was isolated following damage to the ventilation fan, so stopped working when power was lost. This led to the Warden’s Inquiry report commenting that it was desirable for the power supply to such equipment to be independent of the normal supply service. This form of protection can be considered as probably the easiest to achieve to ensure the ongoing operation of tube bundle monitoring following an incident.

There is little information related to the extent of damage to the tubes underground although the Inquiry recognised that the continued use of this equipment even though some of the tube installation had suffered damage could have provided invaluable information concerning the conditions at the time of and immediately after the explosion. Reference is also made in the Inquiry report to a new sample tube point being established during the recovery operations to monitor an area of smouldering coal that was sealed and inerted.

There were grave doubts about the safety of the twelve men unaccounted for but the possibility of a second explosion caused the suspension of further rescue attempts until an accurate assessment of the mine atmosphere could be completed. Arrangements had been made to bring a gas chromatograph Page 128 of 154 ACARP C19010 Extension and chemists from Simtars located just outside Brisbane, however, they were not expected on site for some time. While waiting for their arrival analysers from Mines Rescue were installed to analyse samples. The gas chromatograph arrived on site at 7:30 pm with the first results being available to the control centre shortly after 10:00 pm.

Initial measurements using the Mine Rescue analysers from the main fan drift returned results over range for carbon monoxide (greater than 5000ppm), 1.7% methane and 18% oxygen. To get additional information from within the mine drilling of a bore hole was commenced to intersect the Main Dips Section at 27 ct 4 Heading. The borehole was completed at 4:00 am on 17th July.

Upon the control team being satisfied that the mine atmosphere was not explosive (based on gas trends), rescue team 3 entered the mine at 9:33 a.m. on the 17th July, to explore the workings. Teams 4 and 5 continued exploration and team 6 later confirmed that an explosion had taken place, all lives had been lost and that extensive damage had occurred to the ventilation system. This team located all but two of the bodies of the missing miners in the area of the Main Dips Section. Many difficulties were experienced during these operations, including poor visibility and blast debris. The rescue attempts were hampered by emission of methane from the sealed 4 South panel and increased the level of methane in the Main Dips when the atmospheric pressure fell. Because of these conditions rescue operations had to be halted and it was not possible to recover the bodies from the mine at that time.

On Friday 18th July, rescue teams 7 and 8, continued exploration and recovery work, while other commenced temporary repairs ventilation control devices. Operations were ceased when it became evident that an active fire was present inbye 19 ct. A decision was made to suspend further recovery until a more reliable sample point could be established at 25 ct 1 Heading.

On Saturday 19th July, action was taken to expedite the arrival of inertisation equipment (“Mineshield” from NSW Mines Rescue Service). Additional borehole drilling was commenced to facilitate further sampling, water injection into the goaf and nitrogen injection into the workings. Drilling and monitoring was continued throughout the remainder of that day.

At approximately 8:00 am, Sunday 20th July, the Mineshield equipment and technical personnel arrived on site. However, the fuel for the vaporising unit was delayed. Attempts were made to inject liquid nitrogen directly down two bore holes but proved unsuccessful as back pressure in the boreholes caused the liquid nitrogen to force its way back to the surface via cracks in the subsoil. This eventually froze the ground and blocked the borehole. Water injection was also proving difficult through blockages in the uncased bore holes. Both operations were abandoned and recovery of the blocked holes by reaming and casing was commenced in anticipation of the nitrogen vapourising unit becoming operational and the arrival of further quantities of liquid nitrogen.

Borehole recovery and mine atmosphere sampling continued into Monday 21st July, when borehole sample results at 10:00 indicated the mine atmosphere about the Main Dips Section was not explosive. Further samples one hour later provided similar information. As a result, rescue team 9 accompanied by a District Union Inspector, The Mines Rescue Superintendent and the Government Mines Inspector entered the mine. During this inspection concern arose about the accuracy of sample results received up to that time because a thick bluish smoke and a “fire stink” were detected. These signs indicated the existence of an active fire inbye of 22 ct.

Further exploration was suspended and attempts were made to inject nitrogen gas. The first significant injection rate of five tonnes per hour was achieved at approximately 6:00 pm. This rate was increased gradually to 14 tonnes per hour at 8:00 pm causing the oxygen levels to be slightly reduced. However, this rate could not be maintained due to the difficulties of getting sufficient nitrogen to the site. It was evident that the natural ventilation flow in the unsealed panel was diluting the nitrogen and it was calculated that to reduce the atmosphere to 12% oxygen would require an injection rate of 18 tonnes per hour which could not be guaranteed.

Page 129 of 154 ACARP C19010 Extension

On Tuesday, 22nd July, water injection to the goaf area was recommenced to reduce the area to be inertised by nitrogen. Rescue teams 10 and 11 entered the mine to locate the source of smoke and to erect brattice seals to reduce the quantity of air flow in the panel. Whilst these teams were underground, the nitrogen injection rate was set at 10 tonnes per hour.

In 24 ct between 2 and 3 Headings a large area of smouldering floor coal as well as evidence of burnt out props was discovered. A new sample tube point was established inbye and all roads were sealed by brattice. Drilling of a borehole was commenced directly over the heating to allow the injection of nitrogen vapour into the area. On Wednesday, 23rd July, at 8:20 am, the hole was completed and nitrogen, at the rate of three tonnes per hour, was pumped through the drill stem. With sufficient quantities of nitrogen on site and additional supplies in transit, it was decided to attempt to recover the bodies.

The nitrogen injection rate was increased and five rescue teams were prepared for the recovery operation. By 1:00 pm, oxygen levels had been reduced sufficiently to allow the operation to commence. Physical conditions were extremely arduous with high temperature and humidity, very poor visibility and extensive blast debris. In spite of these conditions all of the bodies including the two bodies which had not previously been located were recovered. The last of the bodies was transported to the surface by 5:15 pm. The “mineshield” equipment was shut down at approximately 5:30 pm.

Page 130 of 154 ACARP C19010 Extension A8 Appin Colliery, NSW 24th July 1979

Reference: Goran,1979; Kininmonth, 2010; Kininmonth, 2010; Disasters – Appin Colliery Gas Explosion – 1979, http://www.illawarracoal.com/appindisaster.htm; MacKenzie-Wood and Ellis 1981.

At 11 pm on Tuesday 24th July 1979 an explosion of methane gas in K Panel of Appin Colliery resulted in the deaths of 14 workers. This disaster was associated with a changeover of ventilation (Figure 54) that was intended to create a situation whereby a central intake, in a three heading development panel, would be shielded from intake gas by returns on either side. Prior to the changeover (stage one) the panel had two intakes and one return. Stage two was the planned post changeover ventilation flow. Stage three was the completion of the changeover with a fan, which had been placed in B Heading being used to ventilate the stub. The changeover was only partially completed before the explosion occurred. The investigation into the disaster found that the whole of B Heading inbye cut through no. 3 was virtually unventilated as the brattice across 3 cut through between LW 8 Maingate and A Heading was still in place and the new temporary brattice across B Heading had been erected. There was consequently a build-up of methane in B Heading that was ignited. The ignition source was thought to be either the deputy’s flame safety lamp or the fan starter box that was found to not be fully secured. It was postulated that a spark from the starter box could have flashed back to the face via the flume line.

At the time of the explosion the control officer was in his office at pit bottom about 5 kilometres from the site of the explosion. He heard a noise like a vibration and the lights went off. He could see dust. An electrical engineer was working on a junction box in D Heading of White Panel approximately 26m outbye of its intersection with B Heading of K Panel; he was about 800m from the explosion. At the time of the explosion he heard a noise which at first he thought was somebody shotfiring along the K Panel track road. The noise was followed by a cloud of dust and a shock wave. This knocked him over to a crouching position behind the man car. Almost immediately there was a heat wave. He did not see any flame. The shock wave caused him to lose his hat and he put up his hands either to retrieve it or hold it on. He received second and third degree burns to the back of both hands and burns of less severity over the whole area of his head and face. He stumbled to the White Panel crib room on the next cut through and found that the power had gone off. There was a lot of dust in the air.

Page 131 of 154 ACARP C19010 Extension

Figure 54: Appin Ventilation Change Plan (taken from Kininmonth 1981)

The White Panel deputy was in the crib room, about 90m outbye of the intersection of B Heading of K panel and D Heading of White Panel (Figure 55) when the explosion occurred. He was knocked to the floor from his seat at the table by a sudden blast of air. He felt debris flying over him. He had first heard a hissing noise and with it felt a heat wave; he received minor singeing of the hair and eyebrows. After a few seconds he heard a big thud, resembling a roof fall. The hissing meanwhile continued. When the hissing stopped the power went off in the crib room. His oil flame safety lamp remained alight. There was almost no visibility because of grey dust. After treating the electrical engineer he went looking for his crew. The White Panel crew were undoubtedly protected from the direct effects of the explosion by a substantial fall which blocked Longwall 8 maingate 80 m from D Heading of White Panel. Near K Panel heading the deputy was blocked by smoke gushing out of the heading. The six men in the White Panel crew made their way to the crib room. The deputy returned to the crib room and contacted the surface for assistance. He went looking for his men again and sometime after heard them in the crib room, which they had reached by another road. He walked his men out meet transport. The deputy then returned to the crib room where visibility had greatly improved. At the intersection with K Panel there was some marlin (bolt rope) burning. He checked stoppings in the cut through and found a number were down.

The crew in LW 6 (950m from White Panel) heard a muffled thud and what sounded like a gush of air followed by some stone dust. All power was lost, telephone lines went dead and the belts stopped. At the face the air was stagnant and very dusty. They attempted to exit through the maingate towards Red Panel. By the time they reached the crib room they could smell smoke. By this time the smoke was so thick that there was almost no visibility. They were guided by belt structure and the track. At Page 132 of 154 ACARP C19010 Extension the first cut through of H Panel the overcast was down on the outbye side. The air was then clear without dust or smoke.

Figure 55: Affected area of Appin Colliery (taken from Kininmonth 1981)

Rescue personnel arrived at the mine at 12:29am, with two teams going underground at 12:35am. Two rescue teams entered the mine and travelled to White Panel. The White Panel crib room (just outbye the entrance to K panel) seemed to have suffered little damage though there was much dust. The crib room was established as the fresh air base for reasons including it had a telephone connection to the surface. They inspected the electrical engineer’s diesel man car near the entry to K panel, there was not much damage in the area. There were signs of burning on the high tension cable and on the tube bundle system. There was burning in B heading of K Panel around Blue Panel. Later inspections showed much more damage, including a destroyed overcast at the intersection of A Heading of Red Page 133 of 154 ACARP C19010 Extension

Panel and B Heading of K Panel. Drager tube readings of carbon monoxide at the entry to A Heading returned off scale readings (greater than 3000 ppm). Visibility was poor so the rescue team followed the track line to the A Heading crib room. Ten bodies were discovered in the crib room. The crib room had been hit with some force. Inanimate objects had been thrown around, but not for any distance. There were no particular signs of charring or burning. Some of the bodies were covered in soot. Further progress by the team up A Heading towards 2 cut through was impossible beyond the sweeps around the cut through (track deviation); the party could see nothing for dust and smoke. Measurements for methane and carbon monoxide were both very high off scale. Ventilation was needed to clear the smoke and methane.

A rescue team entered B heading after the panel had been re-ventilated and found little damage between B1 and B2 but the stone dust barriers had been completely demolished. Inbye B2 visibility was still bad and there was much debris – cables and old belt. Beyond B3 there were steel vent tubes in ribbons, blown down the heading. Further on they came across the fan lying blasted a substantial distance from its original site and overturned. Inbye 4 ct they found the shuttle car containing a body killed by the blast rather than carbon monoxide poisoning. A Heading was found to be relatively undisturbed at 4 ct, its vent tubes were still intact though a little askew. The rescue operation involved 170 volunteers and lasted 41 hours from 11:45 pm, Tuesday to 5:00 pm, Thursday although the last 12 hours were mainly occupied in routine repair work.

MacKenzie-Wood and Ellis (1981) reported that following the explosion at Appin Colliery the NSW Government mobile gas laboratory was used during the rescue operation and temporary ventilation restoration. Its role was to monitor the return gases issuing from the main fan évasé, to supplement the results being obtained from three underground points by the Colliery's Corex Tube Bundle System. According to MacKenzie-Wood and Ellis (1981) although the gases analysed at the return shaft were more dilute than those obtained by the tube bundle system, changes in concentration were detected by the mobile laboratory some forty minutes before they were reported by the Appin tube bundle system. This was because of the long sample transit time in the tubes and is worth considering during emergency situations

The volume of methane that exploded would have been much less than 400 m3 and probably less than 150 m3 based upon the maximum volume of the mine atmosphere that could have exploded. There is evidence of coal dust being involved in the explosion, and indeed the major component. The explosion was most probably curtailed by the presence of stone dust. Damage to the mine extended well beyond K Panel, the extent of structural damage can be seen in Figure 56 and the extent of flames in Figure 57. The first rescue team into the mine came across two men who were hosing down smouldering items and generally attempting to cool the area. The initial explosion created secondary fires.

Page 134 of 154 ACARP C19010 Extension

Figure 56: Extent of structural damage (taken from Kininmonth 1981) Page 135 of 154 ACARP C19010 Extension

Figure 57: Extent of flame travel (taken from Kininmonth 1981)

Page 136 of 154 ACARP C19010 Extension A9 West Wallsend Colliery, NSW, 8th January 1979

Reference: Ellis C, 1980; Goran, 1979; Mackenzie-Wood and Ellis, 1981.

The time of the explosion was between 4.30 am and 4.55 am on Monday, 8th January, 1979. The last work done prior to this was the morning shift of Saturday, 6th January, ceasing at 1 pm. On leaving the colliery all underground power was turned off. Power was to be restored Monday morning following inspection by two deputies. The main ventilation fan had stopped at about 4:30 pm on Saturday 6th January and had been out of operation for 36 hours. It would appear adequate inspections were not undertaken prior to energising the high tension cables throughout the mine and when energised a violent explosion followed. It was thought to be a methane explosion. Calculations based on the mine gas make identify that during the 36 hour fan stoppage about 90,000 m3 of methane would have been liberated. Some of this would have been lost because there would have been some natural ventilation. If no methane escaped, and formed a 10% methane/air mixture this would fill some 80 kilometres of mine roadway - probable well in excess of the total length of roadways at West Wallsend No. 2 The concept, however, helps to understand the situation – even with some methane having been ventilated out, an explosive fuel mixture substantially throughout the mine existed at the time it was ignited.

The explosion was extensive and violent. Dramatic examples of its violence are obvious; the hurling of the fan elbow over a considerable distance (~100m). Objects were thrown vertically from the upcast shaft, having travelled underground, negotiated a right-hand bend and then having been thrown several hundred feet vertically upwards. A wheelbarrow, a stretcher, an oil drum and a telephone were all found on the surface having been propelled from underground. Buildings on the surface in line and near to the drift suffered damage and even nearby steel structures on the surface were damaged by the blast out of the drift. Underground heavy equipment was moved over substantial distances and some broken. Transformers were moved up to 6m and personnel cars blown off the rails and into the ribs.

Timber roof supports were scattered, steel "W" straps were buckled, and conveyor belting was disrupted. Naturally mixed coal and stone dust were blown away from surfaces and deposited elsewhere in the mine or carried above ground.

Damage extended beyond pit bottom at least as far inbye as the intersection of 3 west and Main North headings a distance of 1800m. The damage in the pit was not uniform but was similar over large areas. From the main north heading over several hundreds of metres the damage did not increase or decrease greatly in severity, indicating not a local but an extensive explosion. The direction of the blast appears to be from north to south, although there are the usual exceptions to the pattern. Some areas suffered much less damage than others e.g. 1 West Heading beyond the first few cut throughs was progressively less affected. The south headings were reported to not have been affected to the same degree as the North headings. The returns in the Main North Headings were much less damaged than the intakes. Timber was largely removed from the intakes but still standing in the returns. The location of the ignition of the explosion is not definitively known but appears to be further inbye than the entries into 3 West Panel beyond the intersection with the Main North Headings.

The mine did not have its own tube bundle system and the NSW Government mobile laboratory was being used. The laboratory was located at the upcast shaft and prepared for operation while temporary repairs were made to allow the fan to operate. Two 3/8 "outer diameter Nylon 11 sampling lines were sealed into the upcast shaft, and monitoring continued as the fan was put into operation. The fan was restarted at 18:59 on 8th February. The next day a light haze became visible at the fan évasé, and the carbon monoxide readings increased suddenly. Within minutes dense black smoke was issuing from the shaft, quickly blocking the nylon sampling lines and delaying the arrival of the gas to the analysers. The increasing heat in the upcast shaft melted the sample lines and sealed them, preventing further gas analysis, and blocked all sample lines.

Page 137 of 154 ACARP C19010 Extension A10 Roof Fall in Drift Mine A

Reference: (Personal correspondence with mine wishing to remain unnamed)

A recent significant roof fall in a drift at a mine disrupted communications and power, however six of the eight tube bundle lines in service that were run in the drift survived. The tube bundle installed was in two bundles of five core tube. The five core tube has five tubes encapsulated in an outer protective sheath. Review of vacuum pressures (Figure 58) and sample flow rates (Figure 59) post incident shows that the tubes were not broken but instead squashed. The two tubes that were not operational after the event were squashed to the extent that sample could not be drawn through them, so no measurements were possible.

When tubes are squashed it is important to remember that with the extra resistance leakage paths not previously favoured may eventuate. Any such leakage would influence the accuracy of the results. The onset of leakage in a sample line monitoring an area normally returning low oxygen concentrations (such as a sealed goaf) is easily picked up by an increase in oxygen. It is however more difficult to determine whether a general body type sample is still monitoring from the intended location as increases in oxygen when original samples contain those close to fresh air are difficult to detect. Figure 60 shows there was no change to the oxygen concentration for the location with approximately 10% oxygen and a slight increase in one of the tubes returning over 20% oxygen, indicating that at least some leakage was occurring in this tube. To be assured that general body type samples were still monitoring from the intended locations, review of other gas components would need to be undertaken in conjunction with changes to vacuum pressure.

This example highlights that additional protection of the tubing, in this case the outer sheathing of the five core bundle, can afford the protection required to allow ongoing tube bundle analysis even after a significant fall. Mine sites should evaluate areas of elevated risk for tubes where additional protection is required.

Figure 58: Tube vacuum pressures

Page 138 of 154 ACARP C19010 Extension

Figure 59: Sample flows

Figure 60: Oxygen concentrations

Page 139 of 154 ACARP C19010 Extension A11 Roof Fall in Drift Mine B

Reference: (Personal correspondence with mine wishing to remain unnamed)

A major roof fall occurred at the bottom of a man/materials drift. The fall severed the direct access communication (DAC) system line and the 100 paired telephone communications line. The air and water services were also cut and broken at this point. No communications or services were usable on the immediate inbye side of the fall. No telephone communications were possible.

This mine had another DAC line running down the belt drift connecting with the other line at the bottom of the drifts. Air and water services were also duplicated in the belt drift, and by isolating inbye and outbye the fall were maintained. Another DAC line is installed in an auxiliary intake shaft.

Initial contact to the control room was by DAC from the outbye side of the fall in the man/materials drift. The DAC was still operational outbye the fall in this drift. The next communications were from the bottom of the belt drift. Throughout the incident the DAC was relied on for communications. Without having the communications and services run in both drifts there would have been very little way of communicating to the surface and air and water to service vehicles for withdrawal or compressed air for entrapment would not have been available.

It must be noted that real time and tube bundle gas monitoring systems remained operational throughout with information available at the surface as both are run down a purpose borehole rather than either drift.

Page 140 of 154 ACARP C19010 Extension A12 Crandall Canyon, USA, August 6th 2007

Reference: Gates et al 2008.

On August 6, 2007, at 2:48 am, a catastrophic coal outburst accident occurred during pillar recovery in the South Barrier section of the mine. There were nine fatalities, six in initial outburst and three fatalities and six injuries from another outburst during rescue. Two independent phone systems located in separate entries were run, however phone communication with the South Barrier section was lost. Immediately after the outburst, attempts were made to reach the South Barrier section by pager phone with no response. PEDs were sent to the miner with a receiver on the section to call the surface without response. It cannot be established if systems had either failed and/or the miners had perished or were otherwise unable to respond.

The coal burst damaged ventilation structures but did not appear to damage the PED loop antenna as the PED was used successfully to communicate with surviving miners and rescuers. The real time gas monitoring system reported communication failure from sensors throughout the South Barrier section.

In total, seven boreholes were drilled from the surface to the mine workings. Each successive borehole provided information as to conditions in the affected area and helped to determine the location of the next hole. The first three boreholes were drilled as the underground rescue efforts were ongoing. The next four boreholes were completed after the accident on August 16. None of the boreholes identified the location of the entrapped miners. There were some issues reported associated with collecting samples from boreholes including holes initially being blocked.

Page 141 of 154 ACARP C19010 Extension A13 Alma Mine No. 1, Aracoma Coal Co., USA, 19th January 2006

Reference: Murray et al 2007.

At approximately 5:14 pm on January 19th 2006 a fire occurred at the 9 Headgate longwall belt take- up storage unit, resulting in the deaths of two miners. The fire occurred as a result of frictional heating when the longwall belt became misaligned. Subsequently the belt was fully involved then coal in the belt entry started to burn. Initial attempts to extinguish the fire failed, and observations at the scene indicated that smoke from the fire was traveling further into the mine via the 2 Section intake air course.

Mine management did not immediately withdraw miners from the affected areas (2 Section and the longwall section) when the atmospheric monitoring system (AMS) generated an alarm signal. Approximately 28 minutes elapsed between the time of the first carbon monoxide alarm and the time evacuation of the miners on 2 Section was initiated. Two miners from 2 Section became separated from the other ten miners in thick smoke when travelling to a personnel door on foot during the evacuation and perished. Three miners returned to the smoke-filled intake air course to search for the missing men, but were unable to find them and re-entered the belt entry.

The longwall crew escaped successfully but as the inbye No. 2 Section development crew was escaping, two crew members became separated from the others and perished.

At least initially, the phones worked after the fire began. Miners at the site of the fire travelled 33m to a mine phone to call outside and direct the Dispatcher/AMS operator to order No. 2 Section (inbye the fire) to evacuate. The Dispatcher called the section on the mine's pager phone. There was no response. He then activated the signal light on the section pager phone. Again, no response. He then used the AMS computer to remotely stop the North East Mains No. 1 belt. Sequence switches progressively shut down inbye belts. At this point, 5:39 pm, almost 28 minutes after the first sensor alarmed, miners on No. 2 Section used the mine's pager phone to call outside to find out why their belt had stopped. At that point they were told about the fire and were ordered to evacuate. The AMS system operated as designed after the fire began, 5:14 pm was when the first AMS CO sensor in the belt entry alarmed and at 5:16 a second sensor alarmed.

Eventually mine pager phone communication with the longwall face (5:50 pm) and AMS communication with carbon monoxide sensors in the No. 9 Headgate longwall belt entry (5:59 pm) failed due to the fire. The cable bundle containing AMS and mine phone circuits were destroyed by the fire.

The mine operator failed to provide a two-way voice communication system in an entry separate from an entry containing the AMS as required by § 75.351(r) (US legislative requirement). The pager mine phone system utilised two wires within the AMS cable bundle. The AMS cable bundle for the carbon monoxide sensors along the longwall belt was installed in the longwall belt entry.

A responding MSHA officer learned of a borehole near the location where the 2 Section miners had transferred from the roadway into the alternate escapeway. She believed the borehole might provide valuable information about the mine atmosphere because it was located inbye the fire area. Another MSHA officer was sent to monitor mine gases at the borehole. The carbon monoxide levels in air exhausting from the borehole were 1,300 ppm at about 11:28 pm on January 19. The carbon monoxide concentration at this borehole reached 1,700 ppm during the initial exploration stages. Carbon monoxide concentrations encountered underground were off-scale on mines rescue gas detectors.

Page 142 of 154 ACARP C19010 Extension A14 #3 Mine Fairfax Mining Co., Inc., USA, September 16th 2002

Reference: Stout, Rinehart and Snyder, 2003.

At approximately 9:20 am September 16th, 2002, a fire occurred at the 3 North West No. 2 (#2-3NW) belt drive. The fire was discovered when Donald Nickels, Chief Maintenance Foreman, encountered smoke while traveling to his mantrip near the 3 North West No. 1 (#1-3NW) belt drive. Donald Nickels called the surface and reported the smoke to Johnny Nickels, Mine Superintendent. Since the fire detection system had not sounded an alarm, Johnny Nickels attempted to contact the section crews on the mine phone. Although the section crews could hear his call, their replies were not received on the surface because fire had already damaged the underground mine phone transmission lines.

Donald Nickels intended to check for problems at the #1-3NW belt tailpiece. However, he smelled smoke as soon as he entered the track entry, which he reported by mine phone to Johnny Nickels. Donald Nickels and Sean Fletcher then entered the adjacent primary escapeway and walked to the surface. A miner recognised Johnny Nickels’ voice paging for both sections over the nearby mine phone, located at the section power centre, but was unable to understand Nickels’ message. He tried answering Johnny Nickels, but Nickels couldn’t hear him. The miner was able to communicate with the other working section and, approximately five minutes later, he called a shuttle car operator on the 1-Northeast Section, who informed him that he still had not heard from Donald Nickels. Johnny Nickels tried again to order an evacuation over the mine phone, but he received no response from the sections.

The flames extended inbye from the #2-3NW Belt head roller for a distance of 9m, to the takeup unit, and across the full width and height of the entry (4.6m x 2.2m) The heat of the fire melted the plastic light covers on the belt starter box (1.2m from the drive area), charred the cap pieces on the roof bolts (a distance of 9m), and charred wood posts alongside the belt.

In addition, heat from the fire melted fire suppression and water supply hoses, communication lines, and electrical cables. Testing indicated that the melting point of the fire suppression hose was 165- 172oC. The insulation on the fire sensor wire was melted off; which, when tested, showed a melting point of 185-196oC.

The communication wiring going to the two active sections passed by the #2-3NW belt drive. The wiring included two-way radio antenna wire and the mine phone wires. When the fire occurred, the insulation on the wires was destroyed and communications were disrupted to the working sections. This disruption in communication caused persons to enter the mine to locate miners who remained unaccounted for during the evacuation.

Page 143 of 154 ACARP C19010 Extension A15 No. 5 Mine, Jim Walter Resources, Inc., USA, 23rd September 2001

Reference: McKinney et al, 2002.

On the 23rd September, 2001, two separate explosions occurred at approximately 5:20 pm and 6:15 pm in the No. 4 Section of the mine, resulting in fatal injuries to thirteen miners. At the time of the explosions, thirty-two miners were underground during a non-producing Sunday afternoon shift. The first of the two explosions occurred near the scoop battery charging station. A large rock fell from the roof and then the entire intersection failed. The fall released methane and damaged a scoop battery causing arcing that ignited the methane. Other miners travelled to and into the No. 4 Section to rescue an injured miner and a second, much larger explosion involving both gas and dust triggered by the section block light system occurred, killing all 13 miners on the section or near the mouth of the section. Investigators determined over pressures up to 12 psi were experienced.

The mine communication system was a two-wire pager phone (cable was PVC coated, 18.5 gauge, copper covered steel conductor wires). Phone cable was supported from the roof in the centre of No. 2 entry from the mouth of No. 4 Section to the phone inbye the power centre. The phone near the No. 4 Section was inoperable after the first explosion. A phone farther outbye the section at the 3 East Turn was operable and was used to call the surface but the call was interrupted by problems with the phone system.

The mine had a monitoring system that monitored carbon monoxide. Five minutes after the first explosion, the system showed communication failures for three monitors located in No. 4 Section and 4 East, caused by damage from the forces of the first explosion.

The phone wire was found under the roof fall. It was damaged by the fall, explosive forces, and heat from the explosion. It is not known if the damage was from the 1st or 2nd explosion but the phone was inoperable after the first explosion. The phone itself was recovered by the rescue team and functional when tested later by investigators.

Page 144 of 154 ACARP C19010 Extension A16 Darby Mine No. 1, Kentucky Darby LLC, USA, May 20th 2007

Reference: Light et al, 2007.

At approximately 1:00 am May 20th, 2006, an explosion within the sealed A Left Section at Darby Mine No. 1 resulted in the immediate deaths of two miners who were located at A Left No. 3 Seal. Three of four miners evacuating from the B Left Section succumbed to carbon monoxide poisoning with smoke and soot inhalation. The explosion was attributed to using an oxy acetylene cutting torch to remove metal roof straps from the roof that intersected two seals. This ignited an explosive gas mixture inbye the seals.

The force of the explosion completely destroyed the seals. The flame was limited to the area behind the seals and did not extend into the mine as seen in Figure 61. The personnel carrier that transported two of the victims to the seals where the explosion was initiated was blown outbye nearly 80 metres (see Figure 62). It was estimated that the minimum pressure exerted on the vehicle was 22 psi. Investigators found the motor and a contactor separated from the vehicle and lying almost 10m away. The three other miners who died attempting to escape were found inbye the No. 3 belt drives and communications were damaged at that location. Whether they may have attempted to use the communications is not discussed in investigation reports but it is possible that if they had been able to do so, they may have been guided to safety.

Gas measurements were made at the fan and mine entrances. MSHA inspector measured gas concentrations at the fan (no carbon monoxide detector at site previously) 2.6% methane and over 500ppm carbon monoxide. An air sample sent for analysis returned 0.23% methane, 19.26% oxygen and 6,162ppm carbon monoxide. During rescue operations carbon monoxide concentrations were measured between 80ppm and off scale.

Page 145 of 154 ACARP C19010 Extension

Figure 61: Plan of Darby Mine No. 1 showing extent of flames and victim locations (taken from Light et al 2007)

Page 146 of 154 ACARP C19010 Extension

Figure 62: Damaged personnel carrier on top of conveyor belt debris (taken from Light et al 2007)

Page 147 of 154 ACARP C19010 Extension A17 Truck Fire in Gold Mine

Reference: Personal communications

A truck fire not far from the portal in a gold mine resulted in the loss of communications. As is standard practice in gold mines the underground workers waited out the fire in a refuge chamber. In a review of the event the mine determined the best solution to the loss of communications to be to running an additional phone line via a completely separate route into the underground refuge chambers. The cable chosen for this additional phone line was a fire resistant low smoke cable.

Page 148 of 154 ACARP C19010 Extension Appendix B: Conditions likely to be experienced during and after an event

Page 149 of 154 ACARP C19010 Extension B1 Explosions

Foster and Miller (2009) conducted an extensive review of coal mine explosions that had occurred in the USA as well as data from forty years of explosion testing conducted by NIOSH and the former Bureau of Mines at the lake Lynn Experimental Mine. They reported a worst case estimate of 45psi total pressure at the blast origin (not including within sealed areas) with explosions reviewed ranging from 4-22psi for methane only explosions and 12-20psi for explosions involving coal dust. Note that the Foster and Millar review was completed prior to the Upper Big Branch explosion in which reflected pressures in areas where pressure piling occurred were estimated to be up to 105 psi.

There are many variables that affect the magnitude of the over pressure experienced during an explosion. The volume of gas or dust involved, the concentration of the gas (or dust), the layout of the workings and the distance from the ignition point are all influencing parameters.

When gas burns it releases a certain amount of energy per unit volume burned. The bigger the volume that burns the more energy released. Lake Lynn testing shows that peak over pressures are generated from methane concentrations of 9.5% in air, and that higher or lower concentrations generate lower pressures. Because of the pressure wave associated with the propagation of the flame through an explosive mixture the unburnt gases are compressed and therefore more fuel is available and more energy is released per volume burnt resulting in higher pressures. Therefore the confinement which relates to the layout and dimensions of the workings influences the pressures developed.

Zipf, Sapko and Brune (2007) investigated possible explosion pressures in sealed areas following the explosion at Sago. They reported that combustion of stoichiometric (10%) methane/air mix in a closed volume raises the absolute pressure from 14.7 to 132psia. Maximum pressures generated by coal dust explosions are 129psia similar but slightly less than for methane. These pressures can be amplified by way of pressure piling and reflected waves. Pressures of up to 640psig were reported as possible for detonation with reflected waves. Factors given as influencing gas explosion strength included; methane concentration, turbulence, reflection waves, homogeneity of the mixture, overall volume and degree of confinement.

The greatest overpressures are closest to where the explosion initiates and decrease with distance from the ignition point. The rate of decrease is dependent on the layout of the workings with open cut- throughs rapidly decreasing pressure.

For these reasons it is difficult to estimate exactly what pressure communication and monitoring systems will be exposed to in the event of an explosion. It also highlights the variability and although the site of the ignition may experience pressures that hardware is unlikely to be engineered to survive, over pressures elsewhere within the mine will be less but the need for ongoing monitoring and communications still exists. It must be remembered that pressures as low as 2 psi will propel unanchored devices away from pressure wave.

Although the temperatures of moving flame fronts in a methane explosion can exceed 1000oC, the flame is moving so quickly that the objects in its path are only exposed to the elevated temperatures for such a short period (milliseconds) that the temperature rise of the object can be minimal. Calculations made by Foster and Miller (2009) indicate that cables with insulation that is more than 1mm thick and can tolerate transient temperatures of up to 100oC should not be compromised.

An area that is harder to quantify relating to explosions is damage caused by debris that is picked up by the pressure wave. Observations following explosions have shown that the type of debris picked up or moved in an explosion can vary from small objects like coal and rock particles to large structures such as vehicles and conveyor structure. Debris is capable of destroying the sensor or communication hardware or severing the transmission cable/sample tube.

Page 150 of 154 ACARP C19010 Extension

B2 Fires

As reported by Foster-Miller (2009) the surface temperature of burning coal can reach 900°C, which would be sufficient to cause significant damage to any communications or monitoring system. Unlike an explosion this temperature can be sustained, and influence the temperature of the surrounding atmosphere. For large but localised fires roof temperatures can exceed 500ºC. For smaller fires roof temperatures can exceed 260ºC.

The significance of this is that temperature limits for typical plastic materials used for wire and cable insulation, or plastic conduit, are in the 120-200°C range over several hours. So even though not directly consumed by the fire, transmission can be compromised by elevated temperatures in the area of the fire. Standard tube bundle tube used for sampling at Pike River following the fourth explosion in the area of an identified fire melted and had to be replaced with copper.

Investigation of the Wilberg mine fire revealed that a cable secured to the mine roof was severely burned while an adjacent cable that accidentally fell to the mine floor when its attachment strap melted was not. This shows how location of cables and other communications and monitoring components with respect to elevation in the mine workings is significant when considering survivability.

Page 151 of 154 ACARP C19010 Extension B3 Roof Falls, Outbursts and Pillar Failures

Like explosions and fires the extent and intensity of these types of events is wide ranging, from non- damaging to the filling of a whole roadway. In the extreme cases the hardware and any transmission means can be completely buried and destroyed, by physical forces acting on them, made worse by irregular or sharp surfaces, compromising communication and data and sample transmission inbye of that point. There is also the possibility of compromising the systems in less dramatic events if hit by projected or falling debris. If transmission media is suspended from the roof, any significant movement if not allowed for stretches the media, possibly to a point it is no longer functional. For example, a run of 100 feet, if mounted on an eight feet high mine roof and carried to the floor, could require as much as five feet of lengthwise compliance. This would affect connections and the wire or cable itself. Cabling installations in conduit would need to provide similar compliance. Ten percent extra slack in long runs of cable is often recommended for this purpose. The same goes for transmission media fixed to the rib if rib failure was to occur.

An extreme case associated with roof falls is windblast. ACARP Project C6030 (Fowler and Sharma) reported the following physical damage associated with the initial goaf fall in Longwall Panel No. 1 at Moonee Colliery (Table 11).

Table 11: Physical Damage Associated with Windblast at Moonee Colliery

Element Effect Plaster Board stoppings Seven destroyed Brattice ventilation “wing” (maingate) Displaced nearly 300m outbye Overcasts Superficial damage Tailgate regulator door Badly damaged Water barriers (first ten rows outbye the face) All tubs blown down Water barriers (greater than ten rows outbye) Occasional damage to tubs Computer keypads (maingate) Ripped from connections; One found 30m outbye, other not located Pick block inserts and picks Left on maingate drive –Found scattered to boot end Hydraulic chock return pressure line Perforated through steel armouring and leaking (Butterfly plate found in vicinity) Plastic print canister from underside DCB Ripped off Maingate lights Two found 10m on the return side of the maingate chock

Reported by Fowler and Hebblewhite (2003) as part of ACARP Project No. C8017 the maximum values of windblast parameters shown in Table 12 were measured during the mining of longwall panels no’s 1 to 7 at Moonee Colliery.

Table 12: Maximum Recorded Windblast Parameters

Parameter Maximum Value Peak wind blast velocity 123 m/s Maximum excursion (air flow distance) 197 metres Peak overpressure 35kPa (5.1psi/s) Impulse 89kPa.s (12.9psi.s)

The maximum peak overpressure recorded during the project was 35kPa (5.1psi). As detailed in the report, when a pressure pulse of this intensity strikes a flat surface, such as a stopping, at normal incidence (head on), the instantaneous (peak) value of the reflected overpressure may be equal to twice the peak overpressure plus approximately 2.4 times the dynamic pressure, a total of about 88 kPa (12.8psi). Page 152 of 154 ACARP C19010 Extension

B4 Inundation

Inundation can trap miner workers by impeding escape routes from the mine. Knowledge of where miners are trapped is significant to rescue operations and therefore maintained communications essential. A substantial head of water over the lowest point could occur with miners trapped in a high spot. In the Quecreek incident in the USA in 2002 the maximum depth of inundation in the mine was approximately 100 feet which equates to a water pressure of around 50psi. When roadways are completed flooded it is unlikely that wireless communications would be operational and flooded roadways block RF signals. The duration during which services would need to be maintained could be a few to several days. At Quecreek, the first air/communications borehole reached the trapped miners about 7 ½ hours after the accident, and the physical rescue of the miners began about 100 hours after that.

The water that issued from the face in the Gretley disaster (Staunton, 1998) did so with enough force to move a continuous miner (reported to weigh between 35 and 50 tonnes) 17.5m back down the heading where it jammed. Inundations can impose pressures of over 100 psi for severe inundations with deep water.

Even in what could be considered normal operations small scale flooding or water pooling could have an impact on signal transmission. Any long term services run through potentially wet ground should be suitably designed to withstand water ingress.

Page 153 of 154 ACARP C19010 Extension Appendix C: Copy of mine survey

Page 154 of 154 0

Queensland Mines Rescue Service ACARP Project 51041 Emergency Response: Mine Entry Data Management [Extended]

Industry Risk Assessment 6 & 7 August 2012

Greg Rowan Principle Consultant Rowan & Associates Pty Ltd Ph: 07 3282 6886 Fx: 07 3283 6885 Mb: 0419 641 476 [email protected]

24 August 2012 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 1

Contents

Objective ...... 2

Scope ...... 2

Process ...... 2

Disclaimer...... 3

Acknowledgments ...... 3

Data Acquisition and Reporting System Process Maps ...... 4

Hazard Analysis ...... 6

Unwanted Events Register ...... 7

Bow-Tie Risk Assessment – Loss of Critical Data ...... 13

Action Plan ...... 16

Figure 1 Data Acquisition and Reporting Systems ...... 4

Figure 2 Data Requirements ...... 4

Figure 3 Critical Data Types ...... 5

Figure 4 Critical Data Sensors...... 5

Figure 5 Important Data Types ...... 6

Figure 6 Important Data Sensors ...... 6

Figure 7 Agreed Action Plan ...... 16

Rowan & Associates Pty Ltd Private & Confidential

QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 2

Objective

The objective of the Queensland Mines Rescue Services ACARP Project 51041 [extension] is to assist the Australian Underground Coal Mining Industry assess and identify -

• The general status of Australian underground coal mines environmental monitoring and communications systems capabilities and capacities to provide adequate information after an incident

• What structural design specifications and strategic positioning considerations (including contingences) for environmental monitoring and communications systems would be considered best practice for emergency response

• The current status of existing systems available and suitable for Australian underground coal mines relevant to the scope

• The current status of any research and development being carried out or pending applicable to the scope

• What further specific research is needed to assist industry in implementing the functional specification developed in the original C19010 project and recommendations of how such research could best be achieved

Scope

To achieve these objectives, the project management team developed a work program that included as one of its steps to -

Conduct an Industry Risk Assessment to assess and analyse the potential incidents which could impact effective communications and mine monitoring systems.

This report documents the findings, outcomes and suggested action plan from this risk assessment.

Process

The industry risk assessment detailed above was conducted at the Novotel Hotel, 200 Creek Street, Brisbane, on 6 and 7 August 2012. The Attendance and Action Owners Register included in this report details those present at each session and their assigned actions as outcomes of the assessment.

The risk assessment process consisting of the following steps –

1. Agreeing a defined Scope for the risk assessment [as per Project Proposal]

2. Defining the components of an effective Data Acquisition and Transmission System [captured through a series of Process Maps included]

Rowan & Associates Pty Ltd Private & Confidential

QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 3

3. Detailing and classifying the requirements for data capture and transmission within this system [captured through a series of Process Maps included]

4. Defining the Critical data sources, the types of sensors that collect that data and means for the transmission of that data [captured through a series of Process Maps included]

5. Identifying the types of incidents and their consequences that could threaten the capture, supply and transmission of that Critical data [Hazard Analysis captured in Unwanted Event Register included]

6. Identifying the current preventative and mitigation controls to manage these threats [captured through a Bow-Tie Analysis included]

7. Identifying potential new or improved controls to better protect the capture and supply of critical data for emergency management [included in the Bow-Tie Analysis]

Where specific Action Items were identified, these were assigned to accountable persons and recorded for future consideration of the project management team.

Disclaimer This report was prepared solely for the purpose set out herein and it is not intended for any other person or use. The report is accurate to the best of our knowledge and belief but Rowan & Associates Pty Ltd cannot guarantee the completeness or accuracy of any descriptions or conclusions based on information supplied to it during the conduct of its work. Whilst all reasonable care has been taken in the preparation of the report, all responsibility is disclaimed for any loss or damage, including but not limited to that suffered by any party represented during the conduct of this review arising from the use of this report or suffered by any person for any reason whatsoever.

Acknowledgments The facilitator would like to thank all the personnel who participated in the review who gave freely and openly of their time and expertise. Without such open cooperation and professionalism, the conduct and execution of such multi-party reviews can be difficult and opportunities for shared learning and positive action planning can be compromised. The attendees are to be commended for the outcomes of this assessment.

Rowan & Associates Pty Ltd Private & Confidential

QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 4

Data Acquisition and Reporting System Process Maps The various components of an effective Data Acquisition and Reporting System are detailed in Figure 1. The scope of this risk assessment was to focus on the Data Requirements component of the overall system, including and the threats to the capture, transmission and reporting of this data.

Figure 1 Data Acquisition and Reporting Systems

The Data Requirement component is further broken down into its various aspects as detailed in Figure 3.

Figure 2 Data Requirements

Rowan & Associates Pty Ltd Private & Confidential

QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 5

Critical and Important data types were then separated and the various types of sensors and transmission media used to collect and report then identified. These elements are detailed in the following series of figures.

Figure 3 Critical Data Types

Figure 4 Critical Data Sensors

Rowan & Associates Pty Ltd Private & Confidential

QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 6

Figure 5 Important Data Types

Figure 6 Important Data Sensors

Hazard Analysis A detailed hazard analysis was then conducted to identify the threats to the data sensors and transmission methods that may potentially interfere with the collection and provision of critical data to the post-incident management team.

These threats were analysed for their consequences and potential magnitude. The results are collated in Table 1 Unwanted Events Register.

Rowan & Associates Pty Ltd Private & Confidential

7 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Unwanted Events Register

Unwanted Event / Magnitude / Incident Primary Secondary Consequences Description Consequence Description Consequence Type Consequence Type Risk Description Rating

Loss of Sensors / Data Explosion Over-Pressure Safety Financial 5 Sources Blast debris Safety Financial 4 Heat Safety Legal & Regulatory 5 Fire Safety Legal & Regulatory 5 Dust Safety Legal & Regulatory 4 Contaminants Safety Legal & Regulatory 4 Oxygen Deficiency Safety Legal & Regulatory 4 Strata failure Safety Financial 4 Ventilation changes - damage to VCDs, short-circuits Safety Legal & Regulatory 5

Blast damage to underground infrastructure Safety Legal & Regulatory 5

Blast damage to surface infrastructure Safety Financial 4

Tube damage Safety Legal & Regulatory 5 Imposition of Exclusion Zones Safety Legal & Regulatory 4 Fire Heat Safety Legal & Regulatory 5 Contaminants Oxygen Safety Legal & Regulatory 4 deficiencies Ventilation changes - short circuits, buoyancy impacts, Safety Legal & Regulatory 4

Strata failure Safety Financial 4

Rowan & Associates Pty Ltd Private & Confidential

8 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Unwanted Event / Magnitude / Incident Primary Secondary Consequences Description Consequence Description Consequence Type Consequence Type Risk Description Rating

Destruction of infrastructure [including VCDs] Safety Legal & Regulatory 4

Inundation - Drown [the sensor] Safety Financial 5 inrush Ventilation changes - blockages of airways Safety Legal & Regulatory 4

Destruction of VCDs Safety Legal & Regulatory 4 Destruction of infrastructure Safety Legal & Regulatory 4 Strata failures Safety Legal & Regulatory 4 Inundation - Drown [the sensor] Safety Financial 5 flooding Ventilation changes - blockages of airways Safety Legal & Regulatory 4

Wind Blast Over-Pressure Safety Financial 4 Flying debris Safety Financial 4 Dust Safety Legal & Regulatory 3 Contaminants [flushed from goaf] Safety Legal & Regulatory 3 Ventilation changes - destruction of VCDs; Safety Legal & Regulatory 3

Damaged infrastructure Safety Financial 3 Oxygen deficiencies - goaf gases flushed into airways Safety Legal & Regulatory 4

Water inrush - flushed out of goaf Safety Financial 3 Outburst Over-Pressure Safety Financial 4 Flying debris / coal Safety Financial 4 Dust Safety Legal & Regulatory 4 Contaminants expelled from seam Safety Legal & Regulatory 4

Rowan & Associates Pty Ltd Private & Confidential

9 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Unwanted Event / Magnitude / Incident Primary Secondary Consequences Description Consequence Description Consequence Type Consequence Type Risk Description Rating

Ventilation changes - destruction of VCDs; Safety Legal & Regulatory 4

Damaged infrastructure Safety Legal & Regulatory 5 Oxygen deficiencies Safety Legal & Regulatory 4 Water inrush - flushed out of goaf Safety Legal & Regulatory 3 Strata Failures Crushed / Destroyed sensors Safety Financial 5 Crushed / destroyed tube bundles Safety Financial 5 Dust Safety Financial 3 VCD damage Safety Legal & Regulatory 3 Ventilation changes - pressures, Safety Legal & Regulatory 4 Limits access / trafficability Safety Financial 3 Compromised Gas Inundation of methane into mine Drainage workings Safety Legal & Regulatory 4 infrastructure Collision / Destruction / Disabled Safety Financial 4 Impacts Spontaneous Evacuation triggers Safety Financial 4 Combustion Contaminates Safety Legal & Regulatory 3 Sensor heads Out -of-Range Safety Financial 5 Heat Safety Legal & Regulatory 4 Oxygen deficiencies Safety Legal & Regulatory 4 Exclusion Zones Safety Financial 3 Inadequate NO data Design & Installation of Safety Legal & Regulatory 5 Monitoring Systems installed Incomplete / Ambiguous / Contradictory data Safety Legal & Regulatory 4

Rowan & Associates Pty Ltd Private & Confidential

10 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Unwanted Event / Magnitude / Incident Primary Secondary Consequences Description Consequence Description Consequence Type Consequence Type Risk Description Rating

Loss of Transmission Media Explosion Blast debris Safety Legal & Regulatory 5 Heat Safety Legal & Regulatory 5 Fire Safety Legal & Regulatory 5 Dust impacts on wire-less nodes Safety Financial 3 Strata failure - falls on transmission infrastructure Safety Financial 4

Blast damage to underground infrastructure - including communication networks voice Safety Financial 5 data, visual]

Blast damage to surface infrastructure Safety Legal & Regulatory 4

Fire Heat Safety Legal & Regulatory 4 Contaminants - tars in tube Safety Legal & Regulatory 4 bundles Fire damage to underground infrastructure - including communication networks [voice, Safety Legal & Regulatory 5 data, visual]

Strata failure Destruction / compromised transmission networks Safety Legal & Regulatory 4

Inundation - Strata failures Safety Legal & Regulatory 4 inrush Destruction of infrastructure Safety Legal & Regulatory 4 Short-circuits of electronics Safety Legal & Regulatory 5 Flood tube bundles Safety Legal & Regulatory 5 Blockage of wireless signals Safety Legal & Regulatory 4

Rowan & Associates Pty Ltd Private & Confidential

11 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Unwanted Event / Magnitude / Incident Primary Secondary Consequences Description Consequence Description Consequence Type Consequence Type Risk Description Rating

Overload water traps Safety Legal & Regulatory 4 Wind Blast Pressure wave Safety Legal & Regulatory 4 Flying debris Safety Legal & Regulatory 4 Dust on wireless nodes Safety Legal & Regulatory 3 Damaged infrastructure Safety Legal & Regulatory 5 Loss of Power to PLCs Safety Legal & Regulatory 5 Outburst Over-Pressure Safety Legal & Regulatory 4 Flying debris / coal Safety Legal & Regulatory 4 Dust Safety Legal & Regulatory 3 Damaged infrastructure Safety Legal & Regulatory 4 Strata Failures Crushed / Destroyed transmission media Safety Legal & Regulatory 5

Crushed / destroyed tube bundles - see above Safety Legal & Regulatory 5

Dust Safety Legal & Regulatory 3 Limits access / trafficability Safety Legal & Regulatory 3 Collision / Destruction / Disabled Safety Legal & Regulatory 4 Impacts Spontaneous Heat - damage to transmission Combustion infrastructure Safety Legal & Regulatory 3

Exclusion Zones Safety Legal & Regulatory 4 Inadequate NO data Design & Installation of Safety Legal & Regulatory 5

Monitoring Systems Lightning strike Destruction / compromised transmission networks Safety Legal & Regulatory 4

Rowan & Associates Pty Ltd Private & Confidential

12 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Unwanted Event / Magnitude / Incident Primary Secondary Consequences Description Consequence Description Consequence Type Consequence Type Risk Description Rating

Loss of Mains Loss of data sources Loss of Power Safety Financial 5 Supply Loss of Local Loss of transmission capabilities Supply - Internal Safety Financial 5

Power Trip Failure of UPS - Safety Financial 5 surface Failure of UPS - Safety Financial 5 sensors Failure of Loss of data sources, Failure of Data Management Hardware ; transmission media and data Safety Financial 5 Systems Firmware ; management systems Software Loss of - Loss of Control Room; IT Infrastructure Infrastructure; Communications hub Safety Financial 3 - Destruction / compromised transmission networks

Loss of Access by - Enacting of Exclusion Zone Exclusion requirements - Loss of access to Safety Financial 4

communications and reporting infrastructure

Table 1 Unwanted Events Register

Rowan & Associates Pty Ltd Private & Confidential

13 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Bow-Tie Risk Assessment – Loss of Critical Data

Operation Qld Mines Rescue Service : ACARP Project 51041 Unwanted Top Loss of Critical Data Source Event

Process Step Research Project

CONTROL INFORMATION

Current Current Cont. Cont. Possible New Req’ New Causes of the Risk Control Cont. Mitigation / Control Cont. Detail Preventative Qualit Qualit Improvements or Control d Cont. Top Event Rating Type Effect Recovery Type Effect Controls y y New Controls Type Qual Effect Controls Goaf Fall - - Ingress Protection Replace sensor - "Hardened" sensors windblast Rating [IP55] Subs >90% High with higher standards of IP protection such as Roof Fall - - Ad-hoc Location & Alternative defined in German windblast Placement of sensors monitoring regimes Standard DIN 40050-9 where Impact [boreholes, bag 60- - Ad-hoc Methods of Engin Medium Protection ratings can samples, manual 90% Securing sensors and sampling] be added and Dust and tube bundles in place Water Protection 60- Mediu Engin Ratings can be Use of 90% m SOP for failures of Admin / 60- increased Explosives monitoring systems Medium - Fact-based approach Proced 90% Substit >90 to sensor location and High Outburst Ad-hoc design and secured in ute % installation of consideration of blast pressures / directions / redundant systems Admin / 60- Medium diffusion [tube bundles & Proced 90% Over- telemetric & -Planning & Installation 5 Pressure mechanical] of redundant, Bore-Hole based systems Pressure - OEM Manufacturing - Replace / Vessel Failure Stnds Replace pressure 60- Engin >90% High vessel Subs Medium - Installation SOP for - Re-install pipe 90%

Destructive Force pipe ranges range

Gas Ignition - Ad-hoc Location / - SOP for failures of - Design of mechanical Placement of sensors monitoring systems protection and/or - Ad-hoc design pressure relief and/or - Ad-hoc methods of and installation of dampening from impact Coal Dust securing sensors and Admin / 60- Mediu redundant systems Admin / 60- by debris >90 Explosion Medium Engin High tube bundles in place Proced 90% m [tube bundles & Proced 90% % telemetric & - Design of location and mechanical] securing of sensors out Hybrid of blast direction Explosion

Mobile Plant - Ad-hoc Location / Alternative - Design of mechanical Impact / Placement of sensors Admin / 60- monitoring regimes 60- protection and/or >90 Flying 3 Low Engin Medium Engin High Proced 90% [boreholes, bag 90% pressure relief and/or % Debris - Ad-hoc methods of samples, manual dampening from impact

Rowan & Associates Pty Ltd Private & Confidential

14 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Vehicles securing sensors and sampling] by equipment or tube bundles in place vehicles

- Design of location and securing of sensors out of traffic flows Roof - Ad-hoc Location / Alternative - Design of location / Placement of sensors monitoring regimes securing of sensors [boreholes, bag away from potential fall - Ad-hoc methods of samples, manual areas Rib Strata securing sensors and Admin / 60- sampling] 60- >90 4 Low Engin Medium Engin High Failure tube bundles in place Proced 90% 90% - Design of "hardened" % sensors and transmission media to Goaf Falls withstand / survive fall of ground Fires - Australian Standard - Alternative - Develop and publish AS2290.3 for monitoring regimes an Industry Guideline Explosions Calibration and [boreholes, manual for the design, Maintenance of sampling] installation, Oxidation - Underground Gas commissioning, Spon Monitoring Systems - External Gas operation and Combustion Monitoring support maintenance of mine - Existing systems 60- Mediu Admin / 60- environment monitoring >90 Heat 5 Hot Works Engin Low Engin High designed through 90% m - Change Proced 90% systems that % risk-based methods, Management incorporates the likely GAG Exhausts however, monitoring Procedures [potential] post-incident systems are designed environments that the Use of for normal operating - SOP for failure of system may be

Explosives parameters not Monitoring Systems required to operate in. Friction emergency circumstances - Emergency - Review existing Products of Preparedness and sensor technologies for Combustions - Current sensor Response Plan capability to withstand Generation of design includes changed environmental Dust (limited) condition as a result of Increases in considerations of an incident Contamina 60- Mediu Admin / 60- >90 4 Humidity working Engin Low Engin High nts environments 90% m Proced 90% - Approach OEMs to % Reductions in including Ingress improve the detection Oxygen Protection "bandwidth" Not Fit for Purpose Post - Approach OEMs to Incident increase standards of IP Inertisation protection such as Change of Operating Environmentof Change defined in German Post Fly Ash Standard DIN 40050-9 Incident where impact Decisions Foams 60- Mediu Admin / 60- protection ratings can >90 4 Engin Low Engin High / 90% m Proced 90% be added to increased % Interventi Gels Dust and Water ons protection ratings

Ventilation - Mandate risk-based Modifications approach to use of Pre- VCD Destroyed drilled boreholes Change to / Altered 60- Mediu Admin / 60- >90 3 Engin Low Engin High Ventilation Fans impacted 90% m Proced 90% % / damaged

Rowan & Associates Pty Ltd Private & Confidential

15 QMRS: ACARP Project 51041 Mine Entry Data Management [Extended]

Fire [impact on Vent Pressures] Airway Blockages [falls / debris / flooding] Inability to Conduct Maintenance [clear filters, empty water- Loss of Admin / 60- Admin / 60- >90 4 traps] Low Low Engin High Access Proced 90% Proced 90% % Inability to Calibrate Inability to Leak Test Dust

Blockages Moisture of Sensor Admin / 60- >90 5 Low PPE >90% Low Engin High Sample Bagged-off / Proced 90% % Paths Isolated Tar

Exposure to - Australian Standard - Use of - Review existing High AS2290.3 for redundancies- Use sensor technologies for Concentrations Calibration and of bore-holes capability to withstand Maintenance of changed environmental De- Exposure to Underground Gas condition as a result of Sensitised Admin / Mediu >90 3 Low Levels of Monitoring Systems- >90% Separate >90% Medium an incident- Approach Engin High [recovera Oxygen Risk-based Proced m OEMs to improve the % ble] Emergency detection "bandwidth" Exposure to Preparedness & excess Response Systems moisture

Exposure to Poisoned contaminates

Contamination [non- Admin / 60- 60- >90 5 Low Separate Low Engin High recoverabl Chemical Proced 90% 90% % e damage to sensor

Cross- False Sensitivity responses Admin / 60- 60- >90 4 Low Separate Low Engin High [false Proced 90% 90% % readings]

Table 2 Risk Assessment – Loss of Critical Data

Rowan & Associates Pty Ltd Private & Confidential

QMRS: ACARP Project 51041 Mine Entry Data Management [Extended] 16

Action Plan Item Suggested Improvement Action/s Plan Accountable Completion Person Date 1 Compile Report on outcomes from Industry Risk Assessment to assess and Greg Rowan 24 Aug12 analyse the potential incidents which could impact effective communications and mine monitoring systems complete

2 Identify and assess studies and research already conducted or in progress Darren Brady tba relevant to the research questions and evaluate their potential effectiveness for industry. This should include the international mining industry, and where applicable, industries external to the coal industry

3 Identify and collate current and available information from recent incidents David Cliff tba relevant to the research questions

Identify and collate existing recommendations from relevant incident investigations and Level 1 exercises (post Moura No. 2)

Identify and assess studies and research already conducted or in progress relevant to the research questions and evaluate their potential effectiveness for industry. This should include the international mining industry, and where applicable, industries external to the coal industry

4 Review the current status at existing mines in NSW and Queensland based Geoff Nugent tba on the research questions and highlight best practice where possible. (Industry workshop for operations to present existing standards Based on research questions NSW and QLD)

Identify and assess studies and research already conducted or in progress relevant to the research questions and evaluate their potential effectiveness for industry. This should include the international mining industry, and where applicable, industries external to the coal industry

Industry knowledge transfer - information transfer via industry workshops, seminars, conference presentations and the development of a report with recommended specifications that can also be provided to the operations, service providers and manufacturers

Provide recommendations on specific research required to assist industry achieve the recommended specifications

5 Identify and assess studies and research already conducted or in progress Peter Mason tba relevant to the research questions and evaluate their potential effectiveness for industry. This should include the international mining industry, and where applicable, industries external to the coal industry

6 Identify current legislative and standards requirements relevant to the Stephen tba research questions Tonegato

Investigate and identify strategies and systems utilised inside and outside the Australian coal mining industry which may have potential for direct implementation or be improved to implement in Australia (identify industries with similar issues and analyse controls for further investigation)

7 Identify current legislative and standards requirements relevant to the Seamus Devlin tba research questions

Figure 7 Agreed Action Plan

Rowan & Associates Pty Ltd Private & Confidential

Section 41 SHMS Effectiveness Audit: Isaac Plains Mine Site 17

This page has been left intentionally blank

Rowan & Associates Pty Ltd Private & Confidential ACARP Project Number C19010 Emergency Response: Mine Entry Data Management – Extension Part B Final Report

Darren Brady* Ray Smith David Cliff Anthony Bennett Andre deKock Steve Tonegato Peter Mason Seamus Devlin Geoff Nugent

DATE OF ISSUE: 6th July 2015

* Contact Author: [email protected]

Survivability Testing‐ ACARP project C19010 ‐Emergency Response: Mine Entry Data Management (Extension)

Introduction The aim of this extension to the project was to conduct testing and identify existing and future strategies and hardware which have the potential to protect underground infrastructure such as tube bundle sample lines and data/communication cables delivering samples and information required by decision makers during or after an incident at an underground coal mine. The need for this information has consistently been identified by the researchers throughout the main ACARP project C19010 Emergency Response; Mine Entry Data Management. Since this project started New South Wales has introduced new legislation and the Work Health and Safety (Mines) Regulation 2014, has specific requirements for post incident monitoring. The mine operator of an underground coal mine must now ensure that arrangements are developed and implemented for the monitoring, so far as is reasonably practicable, of the atmosphere of the mine following an explosion or fire that leads to the withdrawal of persons from, and the cutting of the supply of power to all or part of the mine. In developing and implementing the arrangements the mine operator must ensure amongst other issues, that consideration is given to the design of the post incident monitoring system to increase the likelihood of it being able to continue operating after an incident.

This legislative change should promote change (that could flow onto Queensland practice) from what up until now appears to have mine communication and monitoring systems designed, located and implemented to address primarily the proactive risk management of principal and major hazards but give little consideration to a systems purpose, requirements or functionality post an incident i.e. systems are currently designed for “peace time mining”.

From workshops run as part of the project it was identified that there was a need to do some preliminary testing relating to survivability of underground gas monitoring and data/communication infrastructure. There is evidence that after the initial explosion at Pike River that some means of voice communication was still possible, and although communication and power supply was compromised, the methane sensor in the surface fan évasé, when reconnected to power was still capable of measurement. Following the Moura No. 2 explosion tube bundle lines although compromised were still sampling from underground. With critical information already identified as being required to make informed decisions, it is imperative that underground coal mines have available to them ways to best protect those systems that provide this data.

No practical testing or research specific to the survivability potential of these types of installations implemented in Australian operations could be identified. Testing focused on very simple and inexpensive means with the potential to improve the survivability to show that either they warrant implementation, further testing or enhancement. Methods tested included positioning, securing and shielding. Testing was conducted in the 30m propagation tube at Simtars, Redbank. Although small scale the testing provided a cost effective means in what due to the scale, was typically worse case conditions.

Page 2 of 162 Summary of Test Results Testing showed that by implementing some simple techniques the chances of survivability for monitoring and communication systems could be improved following an explosion. A key factor to improving the likelihood of the ongoing operation of gas monitoring and communications systems post event is the installation standard followed and maintained. Areas identified for consideration to maximise the potential included:

 Real time sensors can be protected and remain operational when secured to a shield plate (connecting cable must also be well secured and shield plate can provide protection for cable as well). This technique warrants further investigation.  Hooks not an effective way of securing tube, cable or real time gas detectors.  Running tubes and cables with no sags and secured at 1m spacing reduces likelihood of damage/movement.  Cables connected to real time sensors (and communication hardware) to be secured at multiple locations close to hardware to avoid compromise from pulling at connections.  Damage to tubes can occur when sags allowed and tube “whipped” and strikes hard/sharp surfaces.  Cable ties (minimum width 5mm) offer an effective means of securing individual cables and tubes (wider cable ties or other supports preferable for bundled tubes).  Wrapping cable or tube with insulation tape at fixing points offers additional protection to damage from the tube pulling through the fitting when exposed to over pressure. Coverage area should extend beyond contact with cable tie as tube was observed to move through cable tie distances up to ~20cm.  As an alternative to insulation tape at fix points which may increase installation times significantly, investigation into availability/suitability of slip on insulating and protective piece is warranted.  Standard 5/8inch outer diameter tube is more likely to kink than the standard 1/2 inch outer diameter tube which has a thicker wall thickness.  Variation observed in kinking properties of same dimension tubes but different suppliers. Kinking properties should be considered when purchasing tube.  Where possible core tube bundle should be used to offer additional protections, particularly against kinking.  Exposure to flames/heat during explosion for such a short time, minimal impacts observed to tube and cable.  Metal joiners may retain heat and melt through the tube after the explosion flames pass (observed once in testing). Using nylon fittings or wrapping joiners in insulation tape can minimise potential for this to occur.  Once kinked, tubes generally don’t return to a condition that allows sample flow.  Unsecured coils of tube likely to kink. Wrapping such coils around a fixed diameter at least equal to (but preferably greater than) the minimum bend radius would offer protection. Development of suitable tube bundle storage reels warranted.  Tubes more affected during tests than cables.

A summary of each of the tests conducted is included in the table below (specific details for each test are included as Appendices).

Page 3 of 162 Test No.1

Date: 7th April 2014 Fuel: 600l of 8.9% methane in air Maximum Static Pressure: 134kPa (Position 1) Dynamic Pressure: 65kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General Cage not fixed to inside of prop tube just installed against transducer crossbar. During the explosion the cage impacted on the transducer bar and was crushed inwards by approximately 40‐50mm.

1 Single 5/8"od Run length of cage attached No observable damage, all cable pink tube approx. every 5m with 5mm ties OK. bundle tube cable ties, pulled taught ‐no sags. Remainder extended into transition cone and connected to wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube.

Page 4 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 2 Electrical cable Run length of cage attached All but one cable tie OK. Last cable 1/2" OD blue approx. every 5m with 5mm tie at the tube/transition cone outer sheath cable ties, pulled taught ‐no sags. interface compromised. with 4 Remainder extended into shielded transition cone and connected to conductors wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube.

3 Electrical cable Run length of cage attached No observable damage, all cable 1/2" OD blue approx. every 5m with 5mm ties OK. outer sheath cable ties, pulled taught ‐no sags. with 4 Remainder extended into shielded transition cone and connected to conductors wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube. 4 Electrical cable Run length of cage attached No observable damage, all cable 1/2" OD blue approx. every 1m with 5mm ties OK. outer sheath cable ties, pulled taught ‐no sags. with 4 Remainder extended into shielded transition cone and connected to conductors wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube. 5 Electrical cable Run length of cage attached No observable damage, all cable 1/2" OD blue approx. every 1m with 5mm ties OK. outer sheath cable ties, pulled taught ‐no sags. with 4 Remainder extended into shielded transition cone and connected to conductors wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube.

Page 5 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 6 single 5/8"od Run length of cage attached No observable damage, all cable pink tube approx. every 1m with 5mm ties OK. bundle tube cable ties, pulled taught ‐no sags. Remainder extended into transition cone and connected to wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube.

7 single 5/8"od Run length of cage suspended Tube still suspended only at hook pink tube approx. every 1m from cage with positions 7,8, 10 and 16 bundle tube hooks (21 in total) cabled tied to cage. Remaining tube extended into transition cone and connected to wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube.

8 single 5/8"od Run length of cage suspended Tube still suspended only at hook pink tube approx. every 5m from cage with positions 2 and 4 bundle tube hooks (6 in total) connected to cage with 5mm cable ties. Remaining tube extended into transition cone and connected to wire rope web in 4 places with 5mm cable ties, excess on floor outside of tube. Looking from opposite end

Page 6 of 162 Test No.2

Date: 16th April 2014 Fuel: ? methane in air Maximum Static Pressure: 527kPa (Position 3) Dynamic Pressure: 284kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General Cage was designed with initial intention to use just one fuel zone, but lack of damage/disturbance in first test led to change to two. The cage interfered with the balloon sealing the fuel zone and as such additional methane was injected in an attempt to maintain the methane concentration in the fuel zone, but this was just providing a larger source of fuel, hence the volume of gas involved in the explosion and the methane concentration are unknown. Cage tied to crossbar at exhaust end with wire and cable ties

A1 Single 5/8"od Run length of cage attached Only first cable tie at ignition end of pink tube approx. every 2.5m with 5mm propagation tube intact, all others bundle tube cable ties, maximum sags at 2.5m centres broken. Cable ties between cable ties. Remainder in transition cone OK. Tube kinked extended into transition cone in transition cone and unlikely that and connected to wire rope web sample could be drawn through. in 4 places with 5mm cable ties, Some damage (including kinking) to excess on floor outside of tube. tube from movement through cable Note not attached at start of ties. cage but taped to cable (Element A2).

Page 7 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element A2 Electrical cable Run length of cage attached Only second cable tie from ignition 1/2" OD blue approx. every 2.5m with 5mm end of propagation tube intact, all outer sheath cable ties, maximum sags others at 2.5m centres broken. with 4 between cable ties. Remainder Cable ties in in transition cone OK. shielded extended into transition cone Cable appeared OK although some conductors and connected to wire rope web of the outer insulation heat in 4 places with 5mm cable ties, affected and slightly damaged excess on floor outside of tube. where secured (unlikely to have Note not attached at start of affected cable). cage but taped to cable (Element A1).

B1 Single 5/8"od Run length of cage attached First two cable ties at ignition end pink tube approx every 1m with 5mm were broken all other cable ties at bundle tube cable ties, pulled taught ‐no sags. 1m spacing were intact. No obvious Remainder pulled back into prop damage to tube at 1m fixing points tube and connected with 5mm as seen for other elements where cable ties at 2.5m intervals with tube moved through cable tie. All maximum sag and a swagelok cable ties at 2.5m spacing on return stainless steel union fitted run of tube were broken. The tube approximately at halfway point was disconnected at the Swagelok of excess tube pulled back in. union, heat appears to have been major contributor rather than just mechanical stresses. Tube kinked on return run and unlikely that sample could be drawn through.

Page 8 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element B2 Electrical cable Run length of cage attached First two cable ties at ignition end 1/2" blue approx. every 1m with 5mm were broken all other cable ties at outer sheath cable ties, pulled taught ‐no sags. 1m spacing were intact. Cable ties with 4 Remainder extended into in transition cone OK. Cable shielded transition cone and connected to appeared OK. No damage obvious conductors wire rope web in 4 places with at ties as seen for other elements 5mm cable ties, excess on floor where cable moved through cable at side of tube. tie. C1 Single 5/8"od Run length of cage attached Only the second cable tie in the pink tube approx. every 2m with 5mm cage at 2m from the ignition end bundle tube cable ties, maximum sags survived all others in cage broken. between cable ties. Remainder Cable ties in transition cone OK but extended into transition cone tube kinked and unlikely that and connected to wire rope web sample could be drawn through. in 4 places with 5mm cable ties, Not compromised but outer tube excess on floor outside of tube. damage at fixing points where tube moved through cable tie.

C2 Single 5/8"od Run length of cage attached Cable ties for first 5m only ones that pink tube approx. every 5m with 5mm survived in propagation tube, all bundle tube cable ties, maximum sags others broken. Cable ties in between cable ties. Remainder transition cone survived but tube extended into transition cone kinked and unlikely that sample and connected to wire rope web could be drawn through. Not in 4 places with 5mm cable ties, compromised but outer tube excess on floor at side of tube. damage at fixing points where tube moved through cable tie.

Page 9 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element D1 Electrical cable Run length of cage attached Only cable tie at 2m intact, all 1/2" OD blue approx every 2m with 5mm others in propagation tube broken. outer sheath cable ties, maximum sags Cable ties in transition cone with 4 between cable ties. Remainder survived and loops still attached. shielded extended into transition cone Cable appeared ok but some conductors where 6 loops were fixed to damage to outer insulation from sensor frame and cable pulling through cable tie. connected to simulated sensor as described below.

D2 Electrical cable Run length of cage attached First and second cable ties from the 1/2" OD blue approx every 5m with 5mm ignition end survived, 3rd and 4th outer sheath cable ties, maximum sags broken last cable tie connecting with 4 between cable ties. Remainder cable to cage survived. Cable ties in shielded extended into transition cone transition cone survived. Cable conductors and connected to wire rope web appeared OK although some of the in 4 places with 5mm cable ties, outer insulation heat affected and excess on floor at side of tube. slightly damaged where secured in cage.

Shield plate Shield plate fitted in transition Bottom cable tie holding simulated cone with plastic sample sensor to shield plate broken top container (to simulate gas one intact but container slipped sensor) attached using 5mm through and fell to floor but sample cable ties. Cable (Element D1) container remained intact. Cable attached to container by way of remained connected through gland. cable gland fitted through lid. Container filled with stone dust.

Page 10 of 162 Test No.3

Date: 29th May 2014 Fuel: 1200l 9.2% methane in air Maximum Static Pressure: 316kPa (Position 3) Dynamic Pressure: 169kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General Looking from opposite end Cage shortened to 18.9m for this test and all subsequent tests so as to avoid interference with balloon seal. Cage bolted to inside of prop tube at exhaust end.

A1 Single 5/8"od Run length of cage attached All cable tied connections intact pink tube approx. every 2.2m with 5mm however more slack evident in bundle tube cable ties, maximum sags tubing after cable tie securing tube between cable ties. Remainder at ~5.4m. extended into transition cone and connected to wire rope with 5mm cable ties.

Page 11 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element A2 Electrical cable Run length of cage attached All cable tied connections intact. 1/2" OD blue approx. every 2.5m with 5mm outer sheath cable ties, maximum sags with 4 between cable ties. Remainder shielded extended into transition cone conductors and connected to wire rope web with 5mm cable ties, finishing with 5 loops of cable complete with end plug fixed to top flange bolt hole with 5mm cable tie.

B1 Single 5/8"od Run length of cage attached Only second last cable tie of main pink tube approx. every 2.5m with 5mm tube run broken all other cable tied bundle tube cable ties, maximum sags connections intact. No issues with between cable ties. Remainder union or tube at join. tube pulled back into prop tube and connected with 5mm cable ties at 2.5m intervals with maximum sag and a Swagelok stainless steel union fitted approximately at halfway point of excess tube pulled back in. B2 Electrical cable Run length of cage attached All cable tied connections intact. 1/2" OD blue approx. every 2.5m with 5mm outer sheath cable ties, maximum sags with 4 between cable ties. Excess cable shielded returned into cage and fixed at conductors same spacing using 5mm cable ties with plug fitted at end.

Page 12 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element C1 Single 5/8"od Run length of cage attached All cable tied connections intact. pink tube approx. every 5m with 5mm Insulation tape over Swagelok union bundle tube cable ties, maximum sags heat affected but not melted ‐ looks between cable ties. Remainder like it has been heat shrinked. tube pulled back into prop tube and connected with 5mm cable ties at 2.5m intervals with maximum sag and a Swagelok stainless steel union wrapped in 2 layers of insulation tape, fitted approximately at halfway point of excess tube pulled back in.

C2 Single 5/8"od Run length of cage attached Cable tie at 6.3m from ignition end pink tube approx. every 2.1m with 5mm of cage broken and cable tie fixing bundle tube cable ties, maximum sags return run of tube 2.75m from between cable ties. Remainder exhaust end broken, all other cable tube pulled back into prop tube tie connections intact. Plastic and connected with 5mm cable enclosure no longer covering union ties at 2.1m intervals with (but union and tube ok). maximum sag and a Swagelok stainless steel union fitted approximately at halfway point of excess tube pulled back in and covered with plastic enclosure.

Page 13 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element D1 Electrical cable Run length of cage attached Cable tie attaching cable to start of 1/2" OD blue approx. every 2m with 5mm cage broken and cable end blown outer sheath cable ties, maximum sags into cage. Cable tie attaching 6 with 4 between cable ties. Remainder loops to sensor strap broken and shielded extended into transition cone cable loops ejected from transition conductors where 6 loops were suspended cone but remained connected to from sensor strap with 5mm sensor via gland. Some minor cable tie and then fitted through damage to outer insulation from cable gland to non‐ operational movement through fixing points. gas detector fitted to shield plate described below.

Non‐ Gas detector bolted to shield No observable difference to sensor operational plate which was bolted to condition other than cable through gas sensor transition cone. gland looked slightly strained. fitted to shield plate connected to cable (Element D1) via cable gland.

Page 14 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element D2 Electrical cable Run length of cage attached Three of the five cable ties in the 1/2" OD blue approx. every 5m with 5mm cage were broken and the cable end outer sheath cable ties, maximum sags was blown into the cage. Six loops with 4 between cable ties. Remainder suspended from wire rope web OK. shielded extended into transition cone conductors where 6 loops were suspended from wire rope web with a 5mm cable tie.

E1 Single black Run length of cage attached Three of eight cable ties in cage cable 6.5mm approx. every 2.5m with 5mm broken, all other cable ties OK OD 3 core. cable ties, maximum sags (including loop in transition cone). between cable ties. Remainder extended into transition cone where 10 loops were suspended from the sensor frame with a 5mm cable tie.

Page 15 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element E2 Single black Run length of cage attached Sample jar and cable ejected from cable 6.5mm approx. every 5m with 5mm tube and onto concrete pad. Cable OD 3 core. cable ties, maximum sags released from cable gland, jar between cable ties. Remainder smashed but lid intact. All cable ties extended into transition cone in cage OK. Less slack in in cable at where 8 loops were laid on the end of tube as cable pulled as jar floor . The end of the cable was and cable ejected from transition attached through a cable gland cone. to a 500ml sample jar filled with stone dust to simulate a gas

detector. The jar was fixed to the transducer crossbar with 2x 5mm cable ties.

E3 Single black Run length of cage attached All OK, no cable ties broken and cable 6.5mm approx. every 2.5m with 5mm cable remained taught. OD 3 core. cable ties, no sags between cable ties. Remaining 1m of cable extended into transition cone and left loose.

Page 16 of 162 Test No.4

Date: 17th July 2014 Fuel: 1200l 8.6% methane in air (plus residual coal dust from unrelated testing) Max Static Press: 279kPa (Pos. 3) Dynamic Press: 166kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General Cage bolted to inside of prop tube at exhaust end.

1 Single 5/8"OD Run almost length of cage (1m Cable tie fixing exhaust end of tube pink tube short of end) attached approx. to cage broken, all other cable tie bundle tube every 3m with 5mm cable ties connections intact. Evidence of with 2 layers of insulation tape damage to outside of tube from around tube at each cable tie striking reo cage. No damage at location. Tube allowed to sag to unions. Minor damage to outer approximate cage centre line. tube at 3rd last cable tie from Stainless steel Swagelock unions interaction with cable tie. fitted at bottom of sag at ~3m and 6m.

Page 17 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 2 Single 5/8"OD Run almost length of cage (3m 3 of 7 cable tie sites showed minor pink tube short of end) attached approx. damage. No damage at unions. bundle tube every 3m with 5mm cable ties (no insulation tape at cable tie location). Tube allowed to sag to approximate cage centre line. Brass Swagelock unions fitted at bottom of sag at ~3m and 6m.

3 Single 1/2"OD Run length of cage attached Only second last cable tie of main clear tube approx. every 3m with 5mm tube run broken all other cable tied bundle tube cable ties (no insulation tape), connections intact. No problem with allowing sag to approximately tube or unions at joins. Tube cage centre line. Two nylon showed signs of minor damage to Swagelok unions used to join outer surface where fixed with tube at ~3m and 6m. Tube cable ties. Flame arrestor and end continued into transition cone of line filter OK. and attached to wire rope at one point with 5mm cable tie. Sick end of line filter and and flame arrestor fitted to end of tube with Swagelok union.

Page 18 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 4 Single 5/8"OD Run length of cage attached First taped brass union showed pink tube approx. every 3m with 5mm minor heat effects second showed bundle tube cable ties with 2 layers of greater effects. Last two cable ties insulation tape around tube at broken and damage to tube from each cable tie location within striking reo cage apparent. cage. Tube allowed to sag to approximate cage centre line. Brass Swagelock unions fitted at bottom of sag at ~3m and 6m and wrapped in 2 layers of insulation tape. Tube continued into transition cone and attached to wire rope at three points with 5mm cable ties. 5 Electrical cable Run length of cage attached All OK. 1/2" OD blue approx. every 3m with 5mm outer sheath cable ties with 2 layers of with 4 insulation tape around cable at shielded each cable tie location within conductors cage. Tube allowed to sag to approximate cage centre line. Cable continued into transition cone and 6 loops attached to wire rope and connected to non protected non operational gas detector described below. Electrical junction box used to join cable at midpoint between fixing points and cable tied to cage at ~6m from explosion end of cage.

Page 19 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element Non protected Hung by cable attached to wire Video footage shows detector non‐ rope at side of transition cone swung with blast but no observable operational with 3x5mm cable ties. physical damage to sensor. A cable gas detector tie was broken and the cable and detector orientated differently post ignition.

6 Electrical cable Cable run last 3m of cage only Cable disturbed and some cable ties 1/2" OD blue attached using 5mm cable ties at broken but cable still secured. outer sheath start of cable and end of cage with 4 with sag to cage centre line. shielded Cable continued into transition conductors cone and 8 loops fixed to gas detector shield plate with 5mm cable ties. See photo below

Protected non‐ Gas detector bolted to shield No damage observed. operational plate, bolted to transition cone. gas detector

Page 20 of 162 Test No.5

Date: 3rd September 2014 Fuel: 1200l 8.4% methane in air plus 150g of <250µm coal dust Max Static Pressure: 190kPa (Pos. 2) Dynamic Pressure: 101kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General Cage bolted to inside of prop tube at exhaust end.

1 5 Core Run length of cage attached Cable tie fixing end of tube (ignition sheathed approx. every 2m with 8mm end) to cage broken, all other cable 5/8"OD tube cable ties with minor sagging tie connections intact. bundle tube between cable tie fixtures.

Page 21 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 1F, Single 5/8"OD Tubes positioned at ignition end All cable tie connections intact. 2F, pink tube only with 1.3 m run along top Minor scorch marks on insulation 3F bundle tube and bottom and fixed with 5mm tape (2F) no signs of heat transfer with brass cable ties at three locations top from union to tube in 1F, 2F or 3F. unions and bottom. IF brass union tightened as recommended (1.25 turns past hand tight) 2F overtightened (2 turns past hand tight) and wrapped in two layers of insulation tape. 3F overtightened (2 turns past hand tight). 2 Single 5/8"OD Run length of cage attached Minor scoring to tube (no tape) pink tube approx. every 3m with from 2nd cable tie (5mm) at ignition bundle tube alternating 8mm and 5mm cable end. 5mm cable tie fixing loops in ties, allowing minor sag between cage broken as was last 8mm cable cable ties. Two layers of tie. Loops at end of cage were insulation tape wrapped around blown out into the transition cone tube at 2nd and 3rd last fixing and kinked. Larger loops in points. 2 loops of tube fixed at transition cone intact. end of cage with 5mm cable tie and 3 larger loops fixed to wire rope in transition cone.

4 Electrical cable Run length of cage attached Cable intact appears that cable 1/2" OD blue approx. every 3m with 5mm gland has come away from housing outer sheath cable ties, cable allowed to sag. and most connections to sensors with 4 Cable connected to non ripped off. shielded operational gas detector fixed to conductors cage.

Page 22 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element Non‐ Gas detector screwed to "T" Gas detector blown out of cage and operational plate and attached to cage 1.5 m destroyed. Detector case separated non protected from exhaust end with S hook. from "T" plate. "T" plate and hook gas detector. still connected but separated from the cage.

5 Single 5/8"OD Run length of cage attached First and last cable ties broken. pink tube approx. every 1m with 5mm Tube OK. bundle tube cable ties, no sag.

Page 23 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 5A Electrical cable Run length of cage attached First and second cable ties broken 1/2" OD blue approx. every 1m with 5mm at ignition end of cage. Insulation outer sheath cable ties two layers of insulation tape showed effects of heat with 4 tape at every second cable tie, damage where it was touching the shielded no sag. Cable attached to non steel cage (softening or minor conductors operational gas detector melting). Cable intact. (described below) via cable gland with 6 loops of cable attached to shield plate with 5mm cable ties.

Non‐ Gas detector bolted to shield No observable damage. operational plate which was bolted to gas sensor transistion cone. fitted to shield plate and connected to cable (Element 5A) via cable gland.

6 Single 1/2"OD Tube in transition cone only, 6 No damage observed. clear tube loops connected to wire rope bundle tube hanging from top with 5mm and end of line cable tie. End of line filter and general body flame arrestor fitted to end of filter and tube with Swagelok union. flame arrestor.

Page 24 of 162 Test No.6A

Date: 17th December 2014 Fuel: 1200l 9.8% methane in air Maximum Static Pressure: 69kPa (Position 3) Dynamic Pressure: 57kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General No cage

1 Electrical cable 12 loops of cable tied with 1 x 2 x 5mm cable ties fixing loops to 1/2" OD blue 5mm cable tie, fixed to sensor sensor frame broken. Cable ejected outer sheath frame with 1x 5mm cable tie. from transition cone. Sensor and with 4 Cable connected to existing cable plate still fixed to frame and shielded coming from cable gland in appeared OK, cable still connected. conductors detector by way of 3 x 5mm cable ties and three layers of insulation tape. Sensor fixed to stainless steel plate with 2 x M6 screws and plate bolted to sensor frame with 2 x M10 bolts.

Page 25 of 162 Test No.6B

Date: 17th December 2014 Fuel: 1200l 10% methane in air Maximum Static Pressure: 91kPa (Position 3) Dynamic Pressure: 66kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General No cage

1 Electrical cable 12 loops of cable tied with 1 x 2 x 5mm cable ties fixing loops to 1/2" OD blue 5mm cable tie, fixed to sensor sensor frame broken. Cable ejected outer sheath frame with 1x 5mm cable tie. from transition cone. Taped looped with 4 Cable connected to existing cable end still connected to sensor frame. shielded coming from cable gland in Sensor and plate still fixed to frame conductors detector by way of 3 5mm cable and appeared OK, no movement of ties and three layers of insulation cable out of cable gland. tape. Loose end taped into a loop and cable tied to sensor frame with 5mm cable tie.

Sensor fixed to stainless steel plate with 2 x M6 screws and plate bolted to sensor frame with 2 x M10 bolts.

Page 26 of 162 Test No.6C

Date: 17th December 2014 Fuel: 1200l 9.9% methane in air Maximum Static Pressure: 22kPa (Position 1) Dynamic Pressure: 9kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General No cage

1 Electrical cable 12 loops of cable tied with 1 x No damage observed 1/2" OD blue 5mm cable tie, fixed to sensor outer sheath frame with 1x 5mm cable tie. with 4 Cable connected to existing cable shielded coming from cable gland in conductors detector by way of 3 5mm cable ties and three layers of insulation tape. Loose end taped into a loop and cable tied to sensor frame with 5mm cable tie. Sensor fixed to stainless steel plate with 2 x M6 screws and plate bolted to sensor frame with 2 x M10 bolts.

Page 27 of 162 Test No.6D

Date: 17th December 2014 Fuel: 1200l 9.8% methane in air plus 150 g of <250 µm coal dust Max Static Pressure: 398kPa (Position 3) Dynamic Pressure: 188kPa

Element Parameter / Configuration Position Post Explosion Observations Comments No. Element General Cage bolted to inside of prop tube at exhaust end.

1 Single Run length of cage attached Tube kinked and unlikely to allow 5/8"OD pink approx. every 3m with sample flow. Tube moved further tube bundle alternate 5mm and 8mm cable than taped at cable tie fix and tube ties with sagging between some minor damage observed cable ties and extended 5m on outer surface. Damage to into transition cone and laid tube at un-taped fixtures as well. unsecured on floor.

2 5 Core Run length of cage attached No observable damage. sheathed approx. every 2m with 5/8"OD tube alternate 5mm and 8mm cable bundle tube ties with minor sagging between cable tie fixtures.

Page 28 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 3 Electrical Run length of cage attached No observable damage. cable 1/2" approx. every 1m with 5mm OD blue cable ties with no sagging outer sheath between cable tie fixtures. 2 with 4 Layers of insulation tape shielded around cable at every second conductors cable tie. Excess cable laid on floor in transition cone.

4 Single 6 loops loosely coiled and Cable tie broken and tube 1/2"OD attached to inside surface of ejected from transition cone, black transition cone with a 5mm tubing kinked in 2 places and antistatic cable tie and to end of cage unlikely to allow samlpe flow tube bundle tube

5 Single 10 loops loosely coiled and Cable tie broken and tubing 5/8"OD pink attached to inside surface of ejected from transition cone tube bundle transition cone with 5mm tubing kinked in 2 places and tube cable tie. unlikely to allow sample flow.

Page 29 of 162 Element Parameter / Configuration Position Post Explosion Observations Comments No. Element 6 Single Tube in transition cone only, Tube ok, flame arrestor 1/2"OD clear 11 loops suspended from obviously impacted on transition tube bundle sensor frame with a 5mm cone and bent slightly out of tube and end cable tie. End of line filter and shape but not detrimental to of line flame arrestor fitted to end of operation. general body tube with Swagelok union laid filter and on floor. flame arrestor.

7 Electrical Cable wired to detector No observable damage. 1% cable 1/2" through standard detector methane applied before and after OD blue cable gland and cable fixed to explosion with no difference to outer sheath shield plate with 3 x 5mm detector measurement. with 4 cable ties and fixed to sensor shielded bracket with 2 x 5mm cable conductors ties. connected to methane detector powered and operating.

Page 30 of 162 Methodology Although somewhat limiting the scale of test, all testing was conducted in Simtars propagation tube which did offer a cost effective and convenient means of testing. Although acknowledged as a possible major contributor to tube and cable damage no tests were conducted to observe damage from debris projected by explosion forces. Simtars propagation tube (Figure 1) is 26.66m in length with a nominal internal diameter of 483mm. One of the main objectives of the testing was to see if installation methods influenced survivability of tube and cable. A spiral cage shown in Figure 2 (8mm diameter smooth bar) approximately 0.2m pitch, with six full length longitudinal bars (12mm diameter deformed bar) equally spaced was sourced with maximum dimensions that still allowed it to be inserted into the propagation tube. The original length was 20.1 m and outer diameter approximately 450mm.

Using a structure of this type allowed for fixing of tube bundle lines and communication cables outside of the propagation tube, ensuring efficiency and reliable fixing techniques and practices. It also allowed detailed examination, evaluation and recording of any damage once withdrawn following exposure to explosion. The routine parameter monitoring of the propagation tube (pressure and optical sensors) was utilised during testing to determine pressures infrastructure was exposed to and extent of flame exposure. Figure 3 shows the locations of the sensors used during testing. It must be noted that when inserted in the propagation tube the cage interfered with some of the sensors; particularly the flame sensors. Pressures measured for each of the tests at each sensor are included in Appendix 7.

Originally a square cage had been proposed, however due to the dimensions of the propagation tube this would have meant the cage was significantly smaller than a round cage. There were also unknowns surrounding the influence pressure waves travelling through the cage into the space between the cage and propagation tube would have. For these reasons it was decided a round cage would be better suited to achieving the project objectives.

One section of tube containing approximately 600 litres of an explosive methane air mixture (~9.5% methane) was initially considered sufficient for testing. Previous testing in the propagation tube had shown that by creating turbulent flow (by use of turbulence rings in the tube) overpressures were increased. The cage itself was identified as a likely means of increasing turbulence hence the decision that the minimum volume of fuel would be sufficient for testing. This also allowed the maximum length of cage to be used.

Because of the lack of any observable damage in the first test (other than the ineffectiveness of hooks as supports) a decision was made to increase the number of tube sections filled with an explosive mixture to two (1200l) for remaining tests. What was not predicted was that with this setup the cage would interfere with the balloon sealing the fuel zone from the rest of the propagation tube. Following issues related to this being observed in the second test, the cage was shortened to 18.9m. The cage was also secured to the propagation tube subsequent to the first test, after damage was observed from the cage striking the transducer crossbar at the exhaust end of the propagation tube.

Although only gas explosions were planned, there was residual coal dust from unrelated testing in the propagation tube for the third test. Following this and the observation that it resulted in no significant additional damage, coal dust was purposely added to the final two tests.

To maximise efficiency and to ensure that comparisons are made with consistent experimental conditions, multiple variables were tested each explosion. Tube bundle tube used for testing was

Page 31 of 162 previously unused tube, but all electrical cable tested was used cable recovered from a mine and donated for testing.

Test parameters were not fixed prior to the commencement of testing but determined prior to each test based on results from previous test.

Page 32 of 162

Figure 1: Simtars propagation tube

Page 33 of 162

Figure 2: Test cage

Page 34 of 162

Figure 3: Sensor locations

Page 35 of 162 Discussion of Results The results of testing are covered in detail in Appendices 1 – 6 to this report. To avoid repetition and yet maintain completeness for reporting test results, photographs used to demonstrate observations and results are limited to inclusion in the Appendices.

The most significant results were arguably from the second test. The reason being the much higher pressures generated during this test. These pressures were “accidental” due to additional methane being added when the targeted concentration could not be maintained. It was subsequently determined that the cage was compromising the balloon seal used to confine methane to the fuel section of the propagation tube and gas was escaping to the cage section of the tube. This extra methane would have contributed to the fuel source for the explosion. The damage seen in this test was greater than for any other. This additional methane could not be replicated in subsequent tests due to a lack of control over the process and the related safety concerns. It did highlight how tubes and cables can withstand such conditions if adequately installed and equally how they can be compromised if not.

Mechanical stresses rather than heat from the explosion were found to be the major contributor to any damage observed during testing. No significant “burning” was observed for either tube or cable, even when tubes were installed to run vertically down the centre line at the front of the cage where they were definitely exposed to flames. This lack of burning can be attributed to the minimal amount of time that the tubes and cables were exposed to the heat because of the speed the explosion travels at. The one case where heat appears to be the main contributor to damage was in the second test (most extreme conditions) when a stainless steel tube joiner appears to have retained the heat from the explosion which was then conducted to the tube and most likely in combination with stresses, resulted in the tube being severed on both sides of the join. This type of damage could not be replicated in further testing, even when the joiners were placed as close as possible to the fuel section in which the explosion initiates. Regardless of failure to replicate the problem, wrapping any metal joiners with insulation tape to minimise heat transfer is recommended as a precautionary measure to minimise the possibility of such an occurrence. The effort and financial outlay required to achieve this added protection which may result in continued monitoring when needed most are minimal. Use of nylon unions is another option.

One of the most significant impacts of the physical stresses caused by the explosion pressures was specific to sampling tubes. It is often thought that in an explosion tubes will be cut or damaged in such a way that leakage becomes a problem and the tube is no longer sampling from the intended location. This may be the case with the involvement of flying debris (which wasn’t included in this testing) however other than the damage at the joiner already described, no tubes were damaged to the point they would leak. The most common way that tubes were compromised was kinking, where due to the explosion forces, loose or poorly secured tube was propelled and bent such that it crimped enough to completely block flow. Once kinked, tubes tended to stay kinked and didn’t return to a condition that would allow flow. This is of outmost significance as it is often considered and somewhat relied upon that even though compromised in an explosion, a tube bundle system still provides samples from underground even though they may not be coming from the intended location. This information can still be used by decision makers, particularly if results indicate an unsafe environment. If the tube is kinked and blocked to sample flow, this information won’t be available.

Although both the commonly used tube sizes, (1/2 inch and 5/8 inch outer diameter) tubes kinked at times during testing, manual handling of the tube confirms (as to be expected) tube with a thinner wall thickness is easier to kink. The problem becomes somewhat exaggerated when the tube with the

Page 36 of 162 thinner wall also has a larger internal diameter. This is the case with the standard 5/8 inch outer diameter tube which has a wall thickness of 1/16 inch and an internal diameter of 1/2 inch, compared with the 1/2 inch outer diameter tube which has a wall thickness of 5/64 inch and an internal diameter of 11/32 inch. An advantage of the 5/8 inch tube is that it allows tubes of longer lengths to be used and can have shorter sample draw times than the 1/2 inch tube, however this needs to be weighed up against the increased likelihood of the tube kinking and restricting flow (not just following an incident). Use where possible, of core tube minimises the likelihood of individual tubes kinking because of the increased strength of the bundled sheathed tubes.

Anywhere slack tube is present (regardless of dimensions) needs to be secured and protected otherwise there is a high chance it will be compromised either by impacting other surfaces or becoming kinked. The tube can be made stronger by securing loops together with multiple cable ties and fixing securely to a fixed structure. If the loops are introduced intentionally to overcome strain a custom designed reel that prevents bending past maximum bend radius would assist. Investigation and development of a suitable reel is warranted.

Kinking is not just a problem relating to excess tube. Testing showed that the tubes that weren’t effectively secured become unattached during the explosion and often kinked as result of this movement. Hooks were found to be an unreliable means of securing tubes and cables. Cable ties proved to be an effective means of securing tubes and cables so long as fix points were frequent enough. Testing showed that survivability and ongoing functionality of tube was increased by minimising distance between fixing points and also minimising any sag.

Best results were observed for tube (and cable) where the fixings were at 1m spacing. The tube and cable fixed at 1m spacing resisted damage even in the second test where the much higher pressures were generated. In cases where sag was allowed there was increased movement of tube through the fix points and damage sustained from “whipping” action when exposed to explosion pressures. The movement through the cable tie also scored the outer surface of the tube. Tube integrity (not as important for cable as it is already insulated) can be improved by protecting tube where cable ties (or any fixture) are used. Movement of the tube past the fixing was at times more than 20cm so insulation needs to extend beyond just the contact area with the cable tie. With multiple layers of insulation not only is the tape a barrier but means that when the cable tie is secured, although tight at contact point, if movement occurs the diameter of the cable tie is greater than that of the tube and less damage to the tube occurs.

Testing showed at times that the fixing point could also be where the tube kinked (depending on the how much slack was available in the tube). This could possibly be aided by the tube becoming warm and therefore softer. Use of the insulation tape could also assist in preventing this from happening. As the recommended distance between fix points is only 1m this will result in a lot of taping and possibly significantly increase installation times. Investigation of a suitable slip on insulating and protective piece is warranted.

Although no definite comment could be made on comparing 5mm and 8mm thick cable ties from testing, conversations with mine site personnel indicate that the even under normal conditions thicker ties are better as the load is distributed more and less damage of cable ties “biting” into tube is observed.

Preventing movement of cable is also important. If cable moves it puts physical stress on connections including joins and connections to hardware which may be compromised and therefore

Page 37 of 162 communications or information are no longer available. When loops of cable are displaced by pressure waves they generate inertia which when they come to stop when the slack is used up, pulls on what they are connected to. Therefore like sample tube excess cable needs to be secured.

This prevention of movement is particularly important around the connection to the hardware where cables should be secured in multiple locations. The hardware itself must also be firmly secured. Testing showed that if not secured not only does the hardware (such as a gas sensor) get propelled but when it reaches the end of the length of the cable it is attached to, the cable can be ripped from the instrument.

Testing showed that with appropriate protection, gas sensors can remain operational following exposure to explosion pressures. The shield plate trialled provided adequate protection to the sensors fitted to it, with no physical damage ever observed. For the trial using the functioning methane sensor no change in response was observed when challenged with a known concentration of gas pre and post exposure to explosion pressures. Not only did the shield plate provide protection for the sensor, it provided a way of securely mounting the sensor and fixing the cable attached to it.

Page 38 of 162

Appendix 1 – Test 1

7th April 2014 The primary aim of Test 1 was to test the proposed methodology, cage strength and containment of all materials within the propagation tube. Experience with the propagation tube has shown that by creating turbulent flow (by use of turbulence rings in the tube) overpressures are increased. The cage itself was identified as a likely means of increasing turbulence, hence the original decision for only one section of the tube to be filled with explosive gas.

Although Test 1 was designed primarily to test the methodology there were some parameters tested. These were related to fixing techniques for cables and tube bundle typically used within the industry.

Details and exact positioning of Test 1 setup is shown in Figure 4 - Figure 12. Four lengths of pink 5/8 inch outer diameter low density polyethylene tube with a wall thickness of 1/16 inch and internal diameter of 1/2 inch were attached by various means to the cage for this initial test. Two lengths were attached with 5mm cable ties, one approximately every 5m (Element 1) and the other every 1m (Element 6), both pulled tight with no sags. Excess length was pulled through into the transition cone and secured with 5mm cable ties to a wire rope web that had been put in place. The other two tubes were run the length of the cage but suspended by hooks that were cable tied to the cage. For one tube, hooks were positioned every 1m (Element 7) and the other every 5m (Element 8) with 21 and 6 hooks respectively. Again excess tube was pulled into transition cone and secured in four places to wire rope web with 5mm cable ties.

Four lengths of 1/2 inch outer diameter electrical cable containing four shielded conductor with a blue outer insulation sheath were also attached to the cage for testing. Two cables were run down each side of the cage for the entire length, one on the bottom of the cage and one just above the halfway point. All cables were pulled taught and secured with 5mm cable ties. Cables on one side were secured at 5m spacing (Element 2 upper cable, Element 3 lower cable) and at 1m spacing on the other side (Element 5 upper cable, Element 4 lower cable). The excess of all four cables was extended into the transition cone and connected to the wire rope web in four places with 5mm cable ties, excess on floor outside of tube. Photographs of the setup are shown in Figure 13 and Figure 14. As can be seen in Figure 13 the upper cable (on left side) secured at 5m spacing tended to sag even though it had been pulled taught. Similarly the tube suspended using hooks at 5m spacing (also left side in Figure 13) also tended to sag. The tube secured at 5m spacing but with cable ties did not sag as much as the cable, most likely because the tube is much lighter.

The maximum measured static pressure for this test was 134kPa with a dynamic pressure at the end of the tube measured as 65kPa. The cage was not secured to the propagation tube for the test and impacted on a cross bar at the end of the tube. This resulted in noticeable damage as the cage was crushed inwards by approximately 40-50mm (Figure 15). For all subsequent tests the cage was secured to the propagation tube.

The main observation from this test was that hooks are not an effective way of securing tube bundle (as seen in Figure 16 and Figure 17) or by inference cables. Even for the tube with hooks at 1m spacing the tube was only suspended at four of the 21 hooks following the explosion. The tube with 5m spaced hooks remained suspended at only two hook positions. There was no observable damage

Page 39 of 162 or movement for the tubes secured with cable ties at either spacing, nor in the transition cone for any of the four tubes.

Another key (predicted) observation was that even though at least some of the tube at the ignition end was exposed to flames from the explosion it did not show any adverse effects as seen in Figure 18. This is most likely because as reported by Foster-Miller (2009) the flame is moving so quickly that the objects in its path are only exposed to the elevated temperatures for such a short period (milliseconds) that the temperature rise of the object can be minimal.

For the four cables installed only one cable tie was damaged (Figure 19). This was for the last cable tie at the tube/transition cone interface for the upper cable (Element 2) that was secured at 5m intervals. Of the tubes and cables secured with cable ties this was the one that although pulled taught for installation showed the most sagging.

Page 40 of 162

Figure 4: General Setup Test 1

Page 41 of 162

Figure 5: Test 1‐ Element 1

Page 42 of 162

Figure 6: Test 1‐ Element 2

Page 43 of 162

Figure 7: Test 1‐ Element 3

Page 44 of 162

Figure 8: Test 1‐ Element 4

Page 45 of 162

Figure 9: Test 1‐ Element 5

Page 46 of 162

Figure 10: Test 1‐ Element 6

Page 47 of 162

Figure 11: Test 1‐ Element 7

Page 48 of 162

Figure 12: Test 1‐ Element 8

Page 49 of 162

Figure 13: Test 1 cage configuration

Page 50 of 162

Figure 14: Test 1 transition cone configuration

Page 51 of 162

Figure 15: Image on left: Cage prior to test; Image on right: Cage following first test

Page 52 of 162

Figure 16: Cage configuration view from exhaust end ‐post explosion, still in propagation tube

Figure 17: Cage configuration view from exhaust end‐post explosion, withdrawn from propagation tube

Page 53 of 162

Figure 18: Tube and cable ignition end of cage post exposure to explosion

Figure 19: Broken cable tie on cable

Page 54 of 162 Appendix 2 – Test 2

16th April 2014 Because of the lack of any observable damage in the first test (other than the ineffectiveness of hooks as supports and one broken cable tie) a decision was made to increase the number of tube sections filled with an explosive mixture to two (1200l). What was not predicted was that with this setup the cage would interfere with the balloon sealing the fuel zone from the rest of the propagation tube. When methane injection started the concentration could not be maintained and it was thought there was a leak so additional methane was injected. The leak although unknown at the time, was through the balloon seal and additional methane was being injected into the propagation tube adding to the amount of fuel ignited. The problem is illustrated in Figure 20. No measure of the amount of fuel (other than greater than 1200l) or the concentration is available for this test but pressure measurements indicate it was by far the “largest” explosion generated during these tests. The maximum static pressure was 527kPa at position 3 sensor and the dynamic pressure was 284Pa.

Details and exact positioning of Test 2 setup is shown in Figure 21 - Figure 29. Four pink 5/8 inch outer diameter tubes were again used for this test. The first was run the length of the cage attached approximately every 2.5m with 5mm cable ties with maximum sag allowed between securing points (Element A1). The remaining tube was extended into the transition cone and secured to the wire rope web in four places with 5mm cable ties. The second tube was attached every 1m with 5mm cable ties pulled tight with no sags. The excess tube was pulled back into the cage and secured at 2.5m spacing with 5mm cable ties, with maximum sag between fix points allowed (Element B1). A stainless steel two ferrule (compression type) Swagelok tube union was fitted approximately at the halfway point of the tube returning back into the cage These joiners are used to connect separate lengths of tube although often based on cost brass fittings are used. The third tube was run the same as the first except it was secured at 2m spacing (Element C1) and the fourth tube the same but secured at 5m spacing (Element C2).

Test 2 also included four lengths of 1/2 inch outer diameter electrical cable containing four shielded conductors with a blue outer insulation sheath that were also attached to the cage. One cable (Element A2) was run the length of the cage attached approximately every 2.5m with 5mm cable ties to the upper section of the cage. Maximum sag between fixing points was allowed. Another cable (Element B2) was run along the opposite side of the cage, fixed at approximately 1m spacing with 5mm cable ties, at the middle height of the cage with no sag allowed. Another cable (Element D1) was fixed above Element B2 using 5mm cable ties at approximately 2m spacing with maximum sag allowed. Excess tube was extended into the transition cone and six loops fixed to the sensor frame and the cable fitted via a 1/2 inch cable gland to a 500ml plastic sample jar (to simulate a gas sensor) filled with stone dust (to add weight and for easy observation of any damage). The sample jar was fixed with 2 x 5mm cable ties to a shield plate purposely designed and built to offer protection to gas sensors. This assembly is shown in Figure 30. The fourth cable (Element D2) was fixed approximately every 5m to the cage above Element A2 and maximum sag allowed between fixing points. The excess tube for Elements A2, B2 and D2 was extended into the transition cone and connected to the wire rope web in four places with 5mm cable ties and the remainder coiled on the floor outside of the tube.

Overall setup for Test 2 is shown in Figure 31 and Figure 32.

Page 55 of 162

Figure 20: Explanation of balloon sealing problem for Test 2

Page 56 of 162

Figure 21: General Setup Test 2

Page 57 of 162

Figure 22: Test 2‐ Element A1

Page 58 of 162

Figure 23: Test 2‐ Element A2

Page 59 of 162

Figure 24: Test 2‐ Element B1

Page 60 of 162

Figure 25: Test 2‐ Element B2

Page 61 of 162

Figure 26: Test 2‐ Element C1

Page 62 of 162

Figure 27: Test 2‐ Element C2

Page 63 of 162

Figure 28: Test 2‐ Element D1

Page 64 of 162

Figure 29: Test 2‐ Element D2

Page 65 of 162

Figure 30: Element D1 ‐simulated gas sensor protected by shield plate (pre ignition)

Page 66 of 162

Figure 31: Test 2 cage configuration pre ignition

Figure 32: Test 2 transition cone configuration- pre ignition

Page 67 of 162 This test produced arguably the most significant results, mainly as a result of the additional gas used. Unfortunately due to lack of control of exact methane volume injected and safety concerns the conditions generated in this test could not be repeated.

A significant observation was the damage to the tube at the stainless steel union with both ends shown in shown in Figure 33. The damage sustained compromised the join in the tube completely. It was thought that tubes may separate at such joins but the mechanism of the damage was not predicted. The damage did not appear to be just from mechanical stresses; instead heat appeared to play a significant part. The fitting was located approximately 4m from the exhaust end of the cage. Observation of the damage led to the assumption that the stainless steel fitting absorbed and retained the heat from the passing explosion/heated post explosion gases, and in conjunction with the likely movement and stress on the tube from the explosion forces essentially cut through the tube like “a hot knife through butter”. Discussions with fitting suppliers and industry experts indicated that this type of damage would be more likely with brass fittings due to better heat absorbance and conductivity. Brass fittings are more commonly used by mines because the additional cost for stainless fittings is generally not justified. The damage observed obviously would significantly compromise sampling from any intended location. Following these discussions testing of various fitting types and protection mechanisms was added to the test program.

Figure 33: Damage to tube at union

For the tubes and cables secured at spacing greater than 1m and allowed to sag, almost all of the cable tie fittings were broken. Only the first two cable ties for the tube and cable fixed at 1m spacing with no sag were broken, with no observable damage to the tube or cable. All the cable ties on the return run of this tube back into the cage at 2.5m spacing and allowed to sag were broken. This highlighted the advantage of having tube secured at minimum spacing with minimal sag. The report from Foster- Millar (2009) included that rigid attachment of cables to roof or upper rib with attachments spaced a maximum of 1.5m apart provides significant protection for cables from blast pressures and the evidence from testing supports this. All of the cable ties in the transition cone remained intact.

A unique problem for tube compared with cables is that if they kink sample flow through the tube is reduced or even completely restricted. Such kinking was evident in this second test. The tube run back into the cage and fixed at 2.5m spacing was kinked such that flow would be unlikely as seen in Figure 34. Secure fixings are key to preventing kinking. Figure 35 shows the kinking that occurred to tube within the cage when cables ties failed and it became unsecured. Similar kinking was seen for the other tubes where they entered the transition cone but not where they were secured with cable ties (that all remained intact) to the wire rope in the transition cone and tube here showed no signs of damage. The configuration at the interface to the transition cone prior to ignition and the differences post ignition can be seen in Figure 36.

Page 68 of 162

Figure 34: Kinked tube restricting sample flow

Page 69 of 162

Figure 35: Kinking of unsecured tube in cage

Figure 36: Tube at end of cage prior to ignition on left and post ignition on right

Figure 36 clearly shows issues with tube kinking that would prevent sample draw. Another issue identified with the tube in this test was that the tubes not secured at 1m spacing showed significant movement from original fixing point, and during the process some damage to the outside of the tube resulted most likely as it was moving past cable tie and rubbing against the cage (or just the cable tie). Any damage as a result of the cage could be expected underground where tube and cables are fixed to mesh. The damage may be made easier if the tube is softened by the temperatures generated by the explosion. Examples of this damage can be seen in Figure 37 and Figure 38. The damage observed didn’t appear to compromise the integrity of the tube in this case to the extent that it would leak, but

Page 70 of 162 may have influenced its rigidity/strength as there was regular evidence of kinking in this process as seen in Figure 39. Although not as significant as for the tube, similar damage to the outer insulation was observed for the electrical cable as seen in Figure 40.

Figure 37: Damage to tube outer surface at fix point

Figure 38: Damage to outer surface of tube and cable

Page 71 of 162

Figure 39: Damage and kinking to tube at fix point

Figure 40: Damage to electrical cable outer insulation

The protection offered by the shield plate to the simulated gas sensor was also tested in Test 2. Although not broken the simulated sensor was pulled from the shield plate and dropped to the floor of

Page 72 of 162 the transistion cone as seen in Figure 41. The bottom cable tie holding the simulated sensor to the shield plate was broken as seen in Figure 42. As can be seen in Figure 43 the cable remained connected via the cable gland. It is most likely that the explosion forces acted on the cable and pulled the simulated sensor, breaking the bottom cable tie and pulling it from the remaining top cable tie. Although the sensor in a real scenario may have remained operational it would not have been sampling from the most representative location and those using the results would be unaware of where it was sampling from. This could have serious implications on decision making.

Figure 41: Test 2 transition cone post ignition

Figure 42: Shield plate post ignition

Page 73 of 162

Figure 43: Simulated sensor post ignition

Testing showed that the shield plate could offer protection but the method of fixing the sensor to it was critical. The importance of good quality cable glands was also highlighted, without which the cable may have been stripped from the simulated sensor. This becomes even more significant for sensors fixed more securely. If the cable is stripped from the sensor no information will be available to those making decisions on the surface.

Page 74 of 162 Appendix 3 – Test 3

29th May 2014 To allow the use of two sections of air methane mix (1200l) the cage was shortened from 20.1m to 18.9m for this test (and subsequent tests). This meant that the cage did not interfere with the balloon seal which it did in Test 2 resulting in the additional unknown volume of methane. It was accepted that overpressures greater than those generated in Test 2 due to the additional (uncontrolled) methane were not going to be possible in the remaining tests. 1200l of 9.2% methane in air was ignited with 316kPa the maximum static pressure measured and 169kPa the measured dynamic pressure, both significantly less than that measured for Test 2. A main focus of Test 3 was issues with tube unions as observed in Test 2.

Details and exact positioning of Test 3 setup is shown in Figure 44 - Figure 55. Four pink 5/8 inch outer diameter tubes were used in Test 3. The first (Element A1) was run the length of the cage attached approximately every 2.2m with 5mm cable ties with maximum sag allowed between securing points. The remaining tube was extended into the transition cone and secured to the wire rope web with 5mm cable ties. The second tube (Element B1) was run the length of the cage attached approximately every 2.5m with 5mm cable ties with maximum sags between cable ties. The excess tube was pulled back into the cage and secured at the same spacing (2.5m) again with maximum sags. A stainless steel Swagelok union was used to join tube at approximately the halfway point of the excess tube pulled back in. The third tube (Element C1) was run the length of the cage attached approximately every 5m with 5mm cable ties with maximum sags between cable ties. The excess tube was pulled back into the cage and secured at 2.5m spacing again with maximum sags. A stainless steel Swagelok union wrapped in two layers of insulation tape was used to join tube at approximately the halfway point of the excess tube pulled back in. The fourth tube (Element C2) was run the length of the cage attached approximately every 2.1m with 5mm cable ties with maximum sags between cable ties. The excess tube was pulled back into the cage and secured at 2.1m spacing again with maximum sags. A stainless steel Swagelok union was used to join tube at approximately the halfway point of the excess tube pulled back in. The fitting was covered with a plastic enclosure. The testing arrangements for the unions are shown in Figure 56.

Test 3 also included four lengths of 1/2 inch outer diameter electrical cable containing four shielded conductors with a blue outer insulation sheath. One cable (Element A2) was run the length of the cage attached approximately every 2.5m with 5mm cable ties to the upper section of the cage. Maximum sag between fixing points was allowed. The excess cable was extended into the transition cone and connected to the wire rope web with 5mm cable ties, finishing with 5 loops of cable complete with end plug fixed to top flange bolt hole with 5mm cable tie. The second cable (Element B2) was run along the opposite side of the cage, fixed at approximately 1m spacing with 5mm cable ties, at the middle height of the cage with maximum sag allowed. The excess cable with a plug fitted to the end was returned into the cage and fixed at the same spacing (2.5m) using 5mm cable ties. The third cable (Element D1) was fixed above Element B2 using 5mm cable ties at approximately 2m spacing with maximum sag allowed. Excess tube was extended into the transition cone and six loops fixed to the sensor strap with a 5mm cable tie. The end of the cable was fitted to a non-operational gas detector fixed to a shield plate as shown in Figure 57. The shield plate was bolted to the transition cone in four places. The fourth cable (Element D2) was run the length of the cage above Element A2 and fixed approximately every 5m using 5mm cable ties with maximum sag allowed between fixing points.

Page 75 of 162 The excess cable extended into the transition cone where six loops were suspended from the wire rope web with a 5mm cable tie.

Three lengths of single black 6.5mm outer diameter, three core cable were also used in Test 3 (note this cable is lighter than the 1/2 inch OD blue cable). The first cable was run the length of the cage attached with 5mm cable ties approximately every 2.5m with maximum sags allowed between cable ties. Remaining cable was extended into the transition cone where 10 loops were suspended from the sensor frame with a 5mm cable tie. The second of these cables was run the length of the cage and fixed approximately every 5m with 5mm cable ties and maximum sags allowed between fixings. The remaining cable was extended into the transition cone and eight loops were laid on the floor. The end of this cable was attached through a cable gland to a 500ml sample jar filled with stone dust to simulate a gas detector. The jar was fixed to the transducer crossbar with two 5mm cable ties (shown in Figure 58). The third black cable was run the length of the cage and fixed approximately every 2.5m with 5mm cable ties with cable pulled taught and no sags between fixing points. The remaining 1m of cable was extended into the transition cone and left loose.

Overall setup for Test 3 is shown in Figure 59 - Figure 61.

Page 76 of 162

Figure 44: General setup Test 3

Page 77 of 162

Figure 45: Test 3 Element ‐ A1

Page 78 of 162

Figure 46: Test 3 Element ‐ A2

Page 79 of 162

Figure 47: Test 3 Element ‐ B1

Page 80 of 162

Figure 48: Test 3 Element ‐ B2

Page 81 of 162

Figure 49: Test 3 Element ‐ C1

Page 82 of 162

Figure 50: Test 3 ‐ Element C2

Page 83 of 162

Figure 51: Test 3‐ Element D1

Page 84 of 162

Figure 52: Test 3 ‐ Element D2

Page 85 of 162

Figure 53: Test 3 ‐ Element E1

Page 86 of 162

Figure 54: Test 3 ‐ Element E2

Page 87 of 162

Figure 55: Test 3 ‐ Element E3

Page 88 of 162

Figure 56: Test 3 tube union configuration

Page 89 of 162

Figure 57: Shield plate protecting connected but non‐operational gas detector

Page 90 of 162

Figure 58: Simulated gas sensor fitted to transducer crossbar

Page 91 of 162

Figure 59: Test 3 cage configuration pre ignition ‐ view from exhaust end

Page 92 of 162

Figure 60: Test 3 transition cone configuration pre ignition

Page 93 of 162

Figure 61: Non‐operational and simulated sensor configuration in transition cone

The lower pressures generated in this test compared to those of Test 2 resulted in significantly less damage to fixings and tube. Some damage was however still observed. Tube (Element A1) run the length of the cage and secured approximately every 2.2m with 5mm cable ties and maximum sags moved, and more slack was evident after ~5.4m but cable ties were all intact and tube undamaged. Only one cable tie was broken for tube secured every 2.5m and maximum sag allowed (Element B1) and no cable ties were broken for a tube fixed every 5m with maximum sag allowed (Element C1). Element C1 was run along the cage close to the bottom of the cage which minimised the “free hanging” sag which may be a reason no cable ties were damaged. Two cable ties (at opposite ends of cage) were broken for a tube secured every 2.1m with maximum sag allowed (Element C2).

Following the ignition all unions and tubes at unions remained intact. The plastic cover used for Element C2 was however blown away from the union. The insulation tape that was wrapped around the union for Element C1 did show signs of being heat affected with a “heat seal/shrink” appearance

Page 94 of 162 around the union as seen in Figure 62. The inability to recreate the conditions of Test 2 is likely to have been a factor in the lack of damage observed.

There were some observations of post ignition damage for the 1/2 inch blue electrical cable installations. The cable fixed every 2.5m with maximum sags and extended into the transition cone and five loops complete with end plug fixed to top flange bolt hole with a 5mm cable tie (Element A2) remained intact following the ignition. The other cable (Element B2) secured at 2.5m spacing with maximum sag and excess cable returned into cage also showed no damage. The 1/2 inch blue cable attached approximately every 2m with maximum sag and extended into transition cone with six loops suspended by one 5mm cable tie from the sensor strap and connected to a non-operational gas detector (Element D1) showed some signs of damage. The cable tie attaching the cable to the ignition end of the cage was broken as was the cable tie attaching the six loops to the sensor strap. The looped cable was ejected from the transition cone as shown in Figure 63 but remained connected to the sensor via the cable gland. This highlights the importance of having the sensor well secured. If the sensor had not been as well secured it is possible that it would have been pulled from its mounting and ejected as well. No observable difference to the sensor condition was noted other than the cable through the gland looked slightly strained as shown in Figure 64. Some minor damage to the outer insulation from movement through fixing points was also observed. The cable attached close to the top of the cage at 5m spacing with maximum sags (Element D2) had three of the five cable ties in the cage broken and the cable end was blown into the cage. The six loops suspended from the wire rope remained suspended but showed signs of disturbance as evident in Figure 65.

This was the first test using the black 6.5mm outer diameter three core cable. The cable attached close to top of the cage at 2.5m spacing and maximum sag (Element E1) had three of the eight cable ties in the cage broken. The remaining cable extended into the transition cone with ten loops attached to the sensor frame remained intact. It is likely that because the cable was positioned behind the shield plate, that it was provided protection from the blast pressures, hence its survival. The cable attached at a similar height but on the other side of the cage (Element E2) fixed at 5m spacing with maximum sag between securing points had no cable ties in the cage damaged by the ignition. The eight loops of the excess of this cable laid on the floor were ejected from the transition cone (Figure 66). The simulated sensor the cable was connected to, was also ejected and smashed against a block wall some metres from the end of the transition cone (Figure 67). The cable was pulled from the gland fitted to the lid of the jar simulating the sensor. The cable in the cage was pulled and less slack/sag noted towards the end of the cage. The third black cable (E3) attached every 2.5m with no sag between fix points and remaining 1m left loose in transition cone, showed no signs of damage.

Page 95 of 162

Figure 62: Test 3 union wrapped in insulation tape- top pre ignition, bottom post ignition

Figure 63: Sensor cable: Left: pre ignition Right: post ignition

Page 96 of 162

Figure 64: Cable connection to non‐operational sensor post ignition

Figure 65: Element D2 ‐Pre ignition left post ignition right

Page 97 of 162

Figure 66: Cable ejected from transition cone

Figure 67: Smashed simulated sensor

Page 98 of 162 Appendix 4 – Test 4

17th July 2014 Again two sections of tube were filled with an air methane mix for Test 4. 1200l of 8.6% methane in air was ignited generating a maximum measured static pressure of 279kPa at sensor position 3 and a dynamic pressure of 166kPa. These were slightly lower but very similar to the pressures from Test 3. Previous propagation testing outside this project had been conducted between tests 3 and 4 and coal dust was used in those tests. It would appear that some coal dust remained in the propagation tube and although no obvious signs of being involved in the explosion it may have ignited subsequently, evidenced by the generation of smoke more than ten seconds after the ignition and developing in intensity to the point of being rather thick as seen in Figure 68 about 30 seconds after the ignition but subsequently reduced, presumably when all of the residual coal dust was consumed. This had not been seen in previous tests and at first it was thought that one of the items being tested had caught alight. Post ignition observations revealed that parameters tested were covered in soot and much dirtier than other tests. Care must be taken not to mistake this (in included photographs) as items being singed by the explosion. Testing continued to include examination of the susceptibility of tube unions. Discussions with providers of the unions (Swagelok) previously used in testing indicated that use of nylon unions would be result in less heat transfer to the tube so would be of value to include in testing. Swagelok provided nylon unions for 1/2 inch outer diameter tube for testing, hence inclusion of 1/2 inch outer diameter tube in this test.

Details and exact positioning of Test 4 setup is shown in Figure 69- Figure 75. Four tubes were used in various configurations; three were 5/8inch outer diameter pink low density polyethylene and one was 1/2 inch outer diameter 11/32 inch internal diameter clear low density polyethylene clear tube with a wall thickness of 5/64 inch. The first 5/8 inch tube (Element 1) was run to within 1m of the end of the tube and attached every 3m with 5mm cable ties with two layers of insulation tape around the outside of the tube at each cable tie location (see Figure 76). The tube was allowed to sag to approximately the centre line of the cage and stainless steel Swagelok unions were used to join the tube at approximately 3m and 6m from the ignition end at the bottom of the sag. The second 5/8 inch tube (Element 2) run to within 3m of the end of the cage was installed in the same way as Element 1, but with two brass unions used instead of stainless steel and no insulation tape was wrapped around the outside of the tube prior to being cable tied. The third 5/8 inch tube (Element 4) was secured to the cage in the same manner as the first tube the only difference being that the brass unions were used and wrapped in two layers of insulation tape and the tube continued into the transition cone and was attached to wire rope at three points with 5mm cable ties. The 1/2 inch clear tube (Element 3) was run the length of the cage and attached approximately every 3m with 5mm cable ties with no insulation tape around the tube. The tube was allowed to sag to the cage centre line. Nylon tube unions were fitted in a similar position to the stainless steel and brass unions on the other tubes as seen in Figure 77. This tube was continued into the transition cone and attached to the wire rope at one point with a 5mm cable tie. An end of line general body filter and flame arrestor provided by Sick Pty Ltd was fitted to the end of this tube (Figure 78).

Test 4 also included two lengths of 1/2 inch outer diameter electrical cable containing four shielded conductors with a blue outer insulation sheath. One cable (Element 5) was run the length of the cage attached approximately every 3m with 5mm cable ties with two layers of insulation tape around the cable at each cable tie location. The tube was allowed to sag to approximately the centre line of the cage between fixing points. An electrical junction box was used to join cable at the midpoint between fixing points and cable tied to the cage (Figure 79) at approximately 6m from explosion end of cage.

Page 99 of 162 The excess cable was extended into the transition cone and six loops attached the wire rope and connected to a non-protected non-operational gas detector secured by the cable attached to the wire rope at the side of the transition cone with three 5mm cable ties. The second cable (Element 6) was only run the last 3m of the cage and attached at the start of the cable and the end of the cage with 5mm cable ties and allowed to sag to the approximate cage centre line. The cable was continued into the transition cone and eight loops fixed to the gas detector shield plate with 5mm cable ties. The cable was connected to the gas non-operational detector fixed to the shield plate.

Photographs of the overall setup for Test 4 are shown in Figure 76 to Figure 81.

Figure 68: Thick smoke issuing from propagation tube following explosion

Page 100 of 162

Figure 69: General setup Test 4

Page 101 of 162

Figure 70: Test 4 ‐ Element 1

Page 102 of 162

Figure 71: Test 4 ‐ Element 2

Page 103 of 162

Figure 72: Test 4 ‐ Element 3

Page 104 of 162

Figure 73: Test 4 ‐ Element 4

Page 105 of 162

Figure 74: Test 4 ‐ Element 5

Page 106 of 162

Figure 75: Test 4 ‐ Element 6

Page 107 of 162

Figure 76: Example of tube wrapped with insulation tape at fixing point (top) compared with standard fix point (bottom)

Figure 77: Test 4 unions

Figure 78: End of line filter and flame arrestor in transition cone

Page 108 of 162

Figure 79: Cable junction box

Figure 80: Test 4 cage setup

Page 109 of 162

Figure 81: Test 4 transition cone setup

Element 1, the 5/8 inch outer diameter tube attached approximately every three metres and allowed to sag to the cage centre point had only the cable tie fixing the exhaust end of the tube broken. There was evidence of damage to the outside of the tube from striking the reo cage, an example of which is seen in Figure 82. Although damage was caused by the reo cage, similar damage could be expected from contact with roof/rib mesh. This type of damage was not observed in previous tests. Again this supports securing tube taught to ensure minimal movement of tube and the posibility of being damaged and integrity compromised from “whipping action” if exposed to pressure wave. Figure 83 shows how the insulation tape wrapped around the tube is damaged rather than the tube itself. If using insulation tape, it is recommended that coverage area extend beyond contact with cable tie as tube was observed to move through cable tie distances up to ~20cm at times. No damage to tube was observed at the stainless steel unions used to join the tube in two places.

Three of the seven cable tie sites for the 5/8 inch outer diameter pink tube (Element 2) secured approximately every 3m showed minor damage from interaction between the tube and cable tie/cage. No damage to the tube was observed at the brass unions.

The third 5/8 inch outer diameter pink tube (Element 4) secured approximately every 3m, had the last two cable ties broken and damage to the tube from striking the the reo cage was apparent. It is possible that by having the unions fitted to the tubes, the weight they added and the way they were positioned added to the “whipping” action resulting in damage not previously observed. Although

Page 110 of 162 there was no damage to the tube at the brass unions, the insulation tape used on both showed minor heat effects, more for the union closer to the exhaust end as seen in Figure 84.

The 1/2 inch outer diameter clear tube fixed approximately every 3m (Element 3) also showed minor damage to the outer surface of the tube not protected by insulation tape as seen in Figure 85. Only the second last cable tie in the cage was broken. There was no damage to the tube at nylon unions, nor the unions themselves. The tube joins were uncompromised. Video footage taken looking into the transition cone during the test shows that the end of line filter and flame arrestor hanging in the cone were flung and left swinging from the pressure wave and most likely collided with the transition cone but were not damaged.

Figure 82: Damage to outer surface of tube from striking cage

Figure 83: Damage to insulation tape rather than outside of tube

Page 111 of 162

Figure 84: Taped brass unions- top ignition end, bottom exhaust end

Figure 85: Damage to outer surface of tube at fixing point

The electrical cable attached approximately every 3m along the length of the cage (Element 5) with a junction box joining the cable about six metres from the end of the cage remained intact. The six loops of cable connected to the wire rope in the transition cone also remained in place post ignition. The unprotected non-operational gas detector the cable was observed swinging in video footage collected of the test but apart from a securing cable tie being broken and minor displacement from original position and twisting and displacement of connecting cable no physical damage was obvious (see Figure 86).

The cable ties for the cable (Element 6) secured to the last 3m of the cage remained intact. There was damage to some of the cable ties securing the eight loops of cable to the shield plate, but despite being somewhat disturbed they remained secured and the cable remained connected to the non-operational gas detector fitted to the shield plate (see Figure 87). It is likely that the shield plate offered protection to the cable as well. There were no signs of damage to this detector.

Page 112 of 162

Figure 86: Cable and unprotected detector pre (left) and post (right) ignition

Page 113 of 162

Figure 87: Cable fixed to shield plate pre (left) and post (right) ignition

Page 114 of 162 Appendix 5 – Test 5

3rd September 2014 For the fifth test 1200l (two sections) of 8.4% methane in air mix was ignited with 50g of <250µm coal dust in the first three tube sections after the two gas fuel zones. The coal dust was added following observations of the influence residual coal dust had on Test 4. Again the contribution of the coal dust appeared to more burning after the explosion than being involved in the explosion itself. The maximum static pressure of 190kPa was measured at position 2 sensor and the dynamic pressure was measured as 101kPa.

Details and exact positioning of Test 5 setup is shown in Figure 88 - Figure 95. A five tube (5/8 inch outer diameter) sheathed core tube bundle was included in the fifth test, run the length of the cage and attached approximately every 2m with alternating 8mm and 5mm cable ties to the top of the cage with minor sagging between fixings (Element 1).

In an effort to replicate the damage to tube observed in Test 2 at the union, three single 5/8 inch tubes were positioned at the ignition end of the cage and run approximately 1.3m along the top and bottom of the cage and fixed using 5mm cable ties at three locations top and bottom. Having the unions in this position ensured that they would be exposed to flames and stressed by the over pressure generated. One brass union (considered most likely to damage tube) was used to join each tube at the midpoint of the cage. One union was tightened as recommended (1.25 turns past hand tight) one was overtightened (2 turns past hand tight) and wrapped in two layers of insulation tape, and the third was also overtightened but not wrapped. It was thought that an overtightened fitting would be most likely to damage to the tube. Figure 96 shows the configuration of the brass unions prior to ignition.

Another single 5/8 inch pink tube (Element 2) was run the length of the cage and attached approximately every 3m with alternating 8mm and 5mm cable ties with minor sags between fixing points. Two layers of insulation tape were wrapped around the outer surface of the tube at the second and third last fixing points. Two loops of tube were fixed at the end of the cage with a 5mm cable tie (Figure 97) and three larger loops were fixed to a hanging wire rope in the transition cone (Figure 98). A single 5/8 inch tube (Element 5) was also run the length of the cage attached approximately every 1m with 5mm cable ties pulled tight with no sags. An additional length of single 1/2 inch clear tube (Element 6) fitted with an end of line filter and flame arrestor was secured in the transition cone. The tube was coiled into six loops and connected to wire rope, hanging from the top of the transition cone, with a 5mm cable tie.

A length of 1/2" outer diameter blue outer sheath electrical cable with 4 shielded conductors (Element 4) was run the length of the cage attached approximately every 3m with 5mm cable ties and allowed to sag between fix points. Four loops of cable were attached to the cage prior to the cable being connected to a non-operational gas detector screwed to a simple “T” plate and suspended on an “S” hook from the top of the cage 1.5m from the exhaust end of the cage as shown in Figure 99. Another 1/2 inch outer diameter electrical cable (Element 5A) was run the length of the cage and attached approximately every 1m with 5mm cable ties over two layers of insulation tape at each fix point. Six loops of cable were attached to the shield plate with 5mm cable ties. The cable was then connected to a non-operational gas detector fixed to the shield plate.

Photographs of the general setup for Test 5 are shown in Figure 100 - Figure 102.

Page 115 of 162

Figure 88: General setup Test 5

Page 116 of 162

Figure 89: Test 5‐Element 1

Page 117 of 162

Figure 90: Test 5‐Element 1F, 2F and 3F

Page 118 of 162

Figure 91: Test 5‐Element 2

Page 119 of 162

Figure 92: Test 5‐Element 4

Page 120 of 162

Figure 93: Test 5‐Element 5

Page 121 of 162

Figure 94: Test 5‐Element 5A

Page 122 of 162

Figure 95: Test 5‐Element 6

Page 123 of 162

Figure 96: Test 5 brass fitting configuration

Figure 97: Two loops of 5/8 tube attached to cage

Page 124 of 162

Figure 98: Three larger loops in transition cone

Figure 99: Non‐operational gas detector fitted to cage

Page 125 of 162

Figure 100: Test 5 general setup viewed from ignition end of cage

Figure 101: Test 5 general setup viewed from exhaust end of cage

Page 126 of 162

Figure 102: Test 5 general setup transition cone

Post ignition, the cable tie (8mm) at the ignition end of the cage fixing the five core tube bundle (Element 1) was broken but all other cable ties were intact. There were no signs of heat transfer from the unions to the tube for any of the three unions used (Elements 1F, 2F and 3F) but minor scorch marks were visible on insulation tape, all connections remained intact as seen in Figure 103. No damage to cable ties securing these tubes was sustained and all three tubes remained in place. The tube fixed at 3m spacing (Element 2) showed some minor scoring to the outer surface at the second cable tie (5mm tie, no insulation tape) at the ignition end. The 5mm cable ties securing the loops at the end of the cage were broken and the tube blown out into the transition cone and kinked such that sample flow would not be possible as seen in Figure 104. The last 8mm cable tie was also broken. The larger loops in the transition cone remained intact. The first and last cable tie securing the 5/8 inch outer diameter pink tube, fixed at 1m spacing (Element 5) were broken but the tube remained undamaged. The looped 1/2 inch clear tube (Element 6) with end of line filter and flame arrestor was undamaged.

Figure 103: Brass connections Test 5 - post ignition

Page 127 of 162

Figure 104: Kinked tube

The electrical cable (Element 4) attached to the non-operational gas detector fixed to the cage was pulled from the detector and most of the connections to the detector ripped off as seen in Figure 105. The cable itself appeared to be intact although the loops were detached from the cage and the end of the cable ejected into the transition cone as seen in Figure 106. The hook securing the “T” plate was distorted and disconnected and although bolted to the “T” plate, through the detector body, the detector came away from the plate (Figure 107) smashed into many pieces (Figure 108) which were ejected into and out of the transition cone.

Figure 105: Cable connections ripped from detector

Page 128 of 162

Figure 106: Cable and detector parts ejected into transition cone

Figure 107: Hook and "T" piece used to secure detector to cage (undamaged hook on left for comparison)

Page 129 of 162

Figure 108: Pieces collected from smashed detector

The electrical cable (Element 5A) run the length of cage and attached approximately every 1m with insulation tape wrapped around the cable at fixing points, had the first and second cable ties broken at the ignition end of the cage. The insulation tape showed signs of heat damage (softening or minor melting) where it had been touching the steel cage as seen in Figure 109. The cable however remained intact including the six loops attached to the shield plate in the transition cone and there was no observable damage to the detector fixed to the shield plate (see Figure 110).

Figure 109: Heat damage to insulation tape

Page 130 of 162

Figure 110: Transition cone post ignition

,

Page 131 of 162 Appendix 6 – Test 6

17th December 2014 For the final test (Test 6) multiple tests were conducted on the day. The first three tests were very simple and conducted without the cage in the propagation tube. Essentially they were just looking at various configurations of cable and sensor positioning. 1200l of methane in air was ignited for each of these three tests. The pressures generated were significantly less than for those of previous tests (and the fourth and final test that day that included the cage in the propagation tube). This highlights the increased pressures generated due to turbulence, which as predicted resulted from the cage being positioned within the propagation tube. Specific details of these three tests are included in Figure 111 - Figure 114.

Test 6A (Figure 115) used 1200l of 9.8% methane in air and generated a maximum static pressure of 69kPa and a dynamic pressure of 57kPa. Twelve loops of electrical cable were fixed with one 5mm cable tie to the sensor frame as seen in Figure 116. This cable was joined by way of cable ties and insulation tape to an existing cable connected to a non-operational gas detector. The gas detector was fixed to a stainless steel plate with 2 x M6 screws and the plate bolted to the sensor frame with 2 x M10 bolts as seen in Figure 117.

The cable was ejected from the transition cone (Figure 118) post ignition and cable ties broken. The cable was however still connected to the detector with signs that the cable had been pulled and slightly stressed at the cable gland. The detector itself appeared unaffected and was still fixed to the mounting plate and secured to the sensor frame.

Page 132 of 162

Figure 111: General setup for Test 6A, B and C

Page 133 of 162

Figure 112: Setup for Test 6A

Page 134 of 162

Figure 113: Setup for Test 6B

Page 135 of 162

Figure 114: Setup for Test 6C

Page 136 of 162

Figure 115: Test 6A configuration

Figure 116: Cable secured to sensor frame

Page 137 of 162

Figure 117: Detector mounting

Figure 118: Transition cone post ignition

Page 138 of 162 For Test 6B 1200l of 10% methane in air was ignited generating a maximum static pressure of 91kPa and a dynamic pressure of 66kPa. Configuration of this test was very similar to that of 6A, the main differences being the orientation of the cable fixed to the sensor frame and an additional fix point for the cable. The loose end of the cable was folded over and taped to form a loop and cable tied to the sensor frame. The twelve loops were also cable tied to the sensor frame (see Figure 119). For Test 6B the cable was orientated as shown in Figure 120.

Figure 119: Cable secured to sensor frame

Figure 120: Test 6B configuration

Page 139 of 162 Similar to Test 6A, the cable ties were broken after the ignition and the cable was ejected from the transition cone, but remained attached to the unaffected detector and the additional looped end fixing point as seen in Figure 121.

Figure 121: Transition cone post ignition

The third and final test conducted as part of Test 6 without the cage, 6C, tested a slight variation on the setup for Test 6B. An additional fix point for the cable was added between the loops and the detector as highlighted in red in Figure 122. Post ignition there was no damage observed, highlighting the need and advantage in securing cables effectively.

Page 140 of 162

Figure 122: Cable secured to sensor frame

The main test in Test 6 (Test 6D) included use of the cage. Two sections (1200l) were filled with 9.8% methane in air plus 50g of <250µm coal dust in the first three tube sections after the two gas fuel zones. Again the contribution of the coal dust appeared to more of a burning after the explosion than being involved in the explosion itself. The Maximum static pressure of 398kPa was measured at position 3 sensor and the dynamic pressure was measured as 188kPa.

Details and exact positioning of Test 6 setup is shown in Figure 123 - Figure 130. A single pink 5/8 inch outer diameter tube (Element 1)was run the length of cage attached approximately every 3m with alternating 5mm and 8mm cable ties and allowed to sag between fixings. Two layers of insulation tape were wrapped around the outer surface of the tube where the second and third last cable ties secured the tube. The end 5m of the tube extended into the transition cone and was laid unsecured on the floor. A five 5/8 inch outer diameter tube core sheathed tube bundle (Element 2) was configured in the same way as for Test 5; run the length of the cage and attached approximately every 2m with alternating 8mm and 5mm cable ties to the top of the cage with minor sagging between fixings. Six loops of loosely coiled single 1/2inch outer diameter black antistatic tube bundle tube (Element 4) was attached to the inside surface of the transition cone and the end of the cage with 5mm cable ties. Single pink 5/8 inch outer diameter tube was configured in a similar manner but with 10 loosely coiled loops (Element 5). Eleven loops of single 1/2 inch outer diameter clear tube (Element 6) was suspended from the sensor frame in the transition cone with a 5mm cable tie and an end of line filter and flame arrestor fitted to the end of the tube and left lying on floor of cone.

A 1/2 inch outer diameter blue sheathed electrical cable (Element 3) was run the length of the cage and attached approximately every 1m with 5mm cable ties with no sagging between fixtures. Two layers of insulation tape were applied around the cable at every second cable tie site. Excess tube was laid on the floor in the transition cone.

Page 141 of 162 A main focus of Test 6D was the inclusion of a powered operating methane detector (Element 7). The cable was wired to the detector through the standard cable gland. The connecting cable was fixed to the shield plate with three 5mm cable ties and fixed to the sensor bracket with two 5mm cable ties (see Figure 131). The cable was run through a hole in the top of the transition cone and connected to an appropriate power supply. Prior to ignition, 1% methane was applied to the detector which returned a response of 1%.

Configuration of Test 6D in transition cone is shown in Figure 132.

Page 142 of 162

Figure 123: General setup Test 6D

Page 143 of 162

Figure 124: Test 6D‐ Element 1

Page 144 of 162

Figure 125: Test 6D ‐ Element 2

Page 145 of 162

Figure 126: Test 6D ‐ Element 3

Page 146 of 162

Figure 127: Test 6D ‐ Element 4

Page 147 of 162

Figure 128: Test 6D ‐ Element 5

Page 148 of 162

Figure 129: Test 6D ‐ Element 6

Page 149 of 162

Figure 130: Test 6D ‐ Element 7

Page 150 of 162

Figure 131: Operational detector and cable

Figure 132: Test 6 transition cone configuration - pre ignition

Page 151 of 162 Following the ignition the single 5/8 inch outer diameter pink tube fixed at 3m spacing (Element 1) showed signs of movement and where the third last cable tie was secured to insulation tape the tube moved further through the cable tie than the tape extended and experienced outer surface damage as seen in Figure 133. Similar movement was observed for the second last tie although that tie was broken and insulation tape scored. The tube left lying on the floor of the transition cone was kinked and unlikely to allow sample flow.

Figure 133: Damage to outer tube surface

No damage was observed to the five tube sheathed core bundle (Element 2). The 1/2 inch outer diameter black tube (Element 4) became unattached and was blown out of the transition cone and kinked such that it was unlikely to allow sample flow as seen in Figure 134

Figure 134: Kinked 1/2 inch outer diameter black tube

Page 152 of 162 Similarly the pink 5/8 inch outer diameter tube looped and hung in the transition cone (Element 5) was ejected and kinked in several places to the extent sample flow would not be possible as seen in Figure 135.

Figure 135: Kinked 5/8 inch pink tube

The clear 1/2 inch outer diameter tube attached to the sensor frame, remained intact with the only damage to the assembly being from impact of the flame arrestor on the transition cone and bending out of shape as seen in Figure 136. The transition cone post ignition is shown in Figure 137.

Figure 136: Bent flame arrestor post ignition

Page 153 of 162

Figure 137: Transition cone - post ignition

Page 154 of 162 It was observed during handling (and appears to be the case from test 6D) that the clear 1/2 inch outer diameter tube was the most resistant to kinking. This is to be expected when compared to the 5/8 inch outer diameter tube as it has a thinner wall thickness and greater internal diameter. However it could be expected that based on dimensions alone that the black antistatic tube (also 1/2 inch outer diameter and same wall thickness) would have the same resistance to kinking. It was noted that the black antistatic tube had more inconsistencies in tube dimensions than the clear tube (different suppliers). The outer diameter for the clear tube was consistent with outer diameter consistently measuring 1/2 inch and inner diameter 11/32 inch. However the black antistatic 1/2 inch outer diameter tube from showed variation (even to the eye) with diameters not concentric. Measurements over a 1 metre length returned outer diameter variation between 0.490 and 0.502 inches and varying wall thicknesses. In one measurement, one side had a wall thickness of 0.062 inches and the opposite side had a thickness of 0.075 inches. It is not known what influence this variation has on strength and resistance to kinking.

A simple test was conducted with similar lengths of all three tubes. All tubes were bent to the same angle at the same time. The pink 5/8 inch outer diameter tube was the first to kink, followed by the 1/2 inch outer diameter black antistatic tube. Both of these tubes reached a point where they went from non-kinked to totally kinked almost instantly. This was not the case for the 1/2 inch outer diameter clear tube which as can be seen in the lower image of Figure 138 slowly closed over before kinking. Although only limited testing was done the clear tube was by far the last to kink.

One of the most significant results of the project was that when protected by the shield plate and connecting cable adequately secured a gas detector can remain operational following exposure to blast pressures from an explosion. The methane detector (Element 7) remained operational following the ignition. The detector appeared to have moved slightly from its mounting but the connecting cable remained secured and undamaged following the ignition (see Figure 139). When 1% methane was applied to the sensor a reading of 1% was returned.

Page 155 of 162

Figure 138: tube "kink" angle

Page 156 of 162

Figure 139: Gas detector operational post ignition

Page 157 of 162 Appendix 7 – Measured pressures

Page 158 of 162

Page 159 of 162

Page 160 of 162

Page 161 of 162

Page 162 of 162