UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Atmospheric fuzzy risk assessment of confined spaces at mine reclamation sites Mohammadi, Ladan 2011

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2011_fall_mohammadi_ladan.pdf [ 7.77MB ]
Metadata
JSON: 1.0072023.json
JSON-LD: 1.0072023+ld.json
RDF/XML (Pretty): 1.0072023.xml
RDF/JSON: 1.0072023+rdf.json
Turtle: 1.0072023+rdf-turtle.txt
N-Triples: 1.0072023+rdf-ntriples.txt
Original Record: 1.0072023 +original-record.json
Full Text
1.0072023.txt
Citation
1.0072023.ris

Full Text

  Atmospheric Fuzzy Risk  Assessment of Confined Spaces  at Mine Reclamation Sites  by  Ladan Mohammadi  B.Sc., University of Tehran, 2002 M.Sc., University of Tehran, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE  REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy in The Faculty of Graduate Studies (Mining Engineering) The University of British Columbia (Vancouver) July, 2011  © Ladan Mohammadi 2011ii   Abstract  In 2006, a tragic accident took place at the Sullivan mine in Kimberley, British Columbia. Four people died as the result of their entry into an oxygen-depleted sampling station located at the toe of a waste dump. The dump had been in active use for over 50 years and the sampling shed for about 5 years without any problem. The accident was reported as being unprecedented in the history of mining. The accident shows that reclamation sites can be an atmospheric danger only recognizable if a risk assessment is carried out on a regular basis for many years after closure. It is important to conduct regular assessments since there are physical, chemical and environmental factors that affect oxygen-depletion in waste dumps that change over time. In this thesis, an Atmospheric Fuzzy Risk Assessment (AFRA) tool was devised to recognize confined space dangers at sulfide waste dumps undergoing reclamation. The tool is a fuzzy expert system to transfer knowledge on atmospheric hazards. Modeling the complex environment of a waste dump where internal and external factors change temporally and spatially using conventional mathematical tools is a difficult task. Therefore, a technique based on fuzzy logic and weighted inferencing was applied since this method relies on a heuristic approach that allow for case–based reasoning. AFRA can help mining engineers and other safety professionals to recognize this type of danger while developing a confined space inventory at any site.  The second goal of this research has been to create an application for hand-held pocket PCs and/or Smart phones that can be used by first-responders to provide answers about a possible confined space situation to help them decide to enter or not into that space.  iii  Preface  Temperature and pressure data as well as measured/heuristic values of properties of the Number One Shaft Waste Dump at the Sullivan Mine site in Kimberley, B.C. were provided by Teck Resources, the company responsible for the technical investigation of the confined space accident at Sullivan mine. In parallel with the work of the appointed Technical Advisory Panel, my research was conducted in the Centre for Environmental Research in Minerals, Metals, and Materials (CERM3) at the Norman B. Keevil Institute of Mining Engineering at the University of British Columbia. The research was funded by WorkSafeBC and the Workers Compensation Boards of Manitoba and Nova Scotia, independently of Teck Resources and the final outcome was shared with the Technical Advisory Panel for review and comment. The outcome will also be shared with B.C. Ministry of Mines and Energy for practical use by mining engineers and managers.  Some parts of Chapter 2 and 3 were published in the following paper. This includes Parts of Sections 2.1, 2.2.1, 2.3, 2.3.1.1, 2.3.5. Some parts of Sections 3.1, 3.2, 3.3.1, a summary of the accidents in Sections 3.4, a some ideas in Section 5.3.3 and 5.3.4 and a some parts of Appendix E.1. Mohammadi, L., Meech, J.A, 2008, Implementing Atmospheric Risk Assessment in Mine Reclamation, 23rd International Conference on Solid Waste Technology and Management, Philadelphia, PA, USA, March, pp.12.  I reviewed the literature on confined space accidents and regulations and built the preliminary structure of the Atmospheric Fuzzy Risk Assessment (AFRA) tool and wrote the draft of the first paper. Dr. Meech finalized the papers and gave me ideas about reviewing more focused literature in order to narrow the topic. He also helped to gather confined space fatalities data from the literature and online resources.   iv The Atmospheric Confined Space Manual introduced in Appendix C was written to help the users of AFRA with recognizing atmospheric confined space hazards. I gathered all relevant information in an attempt to consolidate elements from other manuals (WorkSafeBC, NIOSH, and OSHA). To further enhance the manual, I reviewed different confined space accidents at mines and other industrial sites.  A summary of atmospheric related dangers at mine reclamation sites in Appendix C is reviewed and summarized in the following paper. I wrote the paper and Dr. Meech edited and finalized it. The paper includes some of the ideas and sentences in Sections 1.1, 2.5, 3.3.1, 3.3.2, 3.5, 4.1, 5.2, 5.3.7, 7.2 in this thesis, as well as parts of the Sections C.8.5, C.8.13, C.8.9, C.9., C.10, E.2 in the Appendcies. Mohammadi, L. , Meech, J.A., 2011. Atmospheric Occupational Health and Safety Issues at Mine Reclamation Sites, Proceedings of the International Conference on Environmental Pollution and Remediation (accepted), Ottawa, Ontario, Canada, 17-19 August 2011.  Parts of the Sections 5.1, 5.2, 5.3, 5.4, 5.7, parts of the conclusion in Chapter 7, parts of the Sections E.1 and E.2 as well as some of the tables in Appendix F were published in a report to WorkSafeBC that was subjected to two anonymous peer reviews by the agency and fed back to me. The report is published online at the WorkSafeBC website. I gathered and interpreted all data and information in the literature related to waste dump environments and applied them to create, verify, and validate the Atmospheric Fuzzy Risk Assessment tool. Dr. Meech revised the tool and provided feedback to make it more user-friendly and understandable. I wrote the report while Dr. Meech reviewed and edited it and suggested ways to improve the report and make it more presentable.  Meech, J.A., Mohammadi, L., January 2011, Confined Space Atmospheric Risk Assessment, Focus on Tomorrow, WorkSafeBC, pp. 81.  Some parts of Chapter 6 have been submitted as a refereed paper:  Mohammadi, L., Meech, J.A., 2011. AFRA – Heuristic Expert System to Assess the Atmospheric Risk of Waste Dumps, Water, Air, and Soil Pollution, (submitted), pp.20.   v This paper describes the complete structure of AFRA and was written by myself and revised by Dr. Meech. I wrote the following paper for Proceedings of the Tailings and Mine Waste 2011 in Vancouver, BC. This paper is also about structure of AFRA and discusses some of the test results.  Mohammadi, L., Meech, J.A., 2011. Atmospheric Fuzzy Risk Assessment of Confined Space Entry at Mine Reclamation Sites, Proceedings Tailings and Mine Waste 2011 Vancouver, BC, November 6 to 9 (submitted), pp 12.   Since the aim of the thesis was to distribute knowledge from the literature to mining industry personnel, I prepared several posters and presented them at several conferences and events: Mohammadi, L., Meech, J.A., 2008, Risk Assessment of Oxygen-Depletion and Hazardous Gas Emissions at Mine Reclamation Sites, BC Environmental and Occupational Health Research Network (BCEOHRN) AGM and Scientific Exchange, Vancouver, B.C., November.  Mohammadi, L., Meech, J.A., 2009, Atmospheric Fuzzy Risk Assessment of Mine Reclamation Sites, BC Innovation Council Gala Evening, Vancouver, B.C., October.  Mohammadi, L., Meech, J.A., 2009, Atmospheric Fuzzy Risk Assessment at Mine Reclamation Sites, CIM Student Night – Vancouver Branch, November.  I have orally presented my work on two major occasions outside of UBC: Mohammadi, L., and Meech J.A., 2010. Fuzzy Risk Assessment of Atmospheric Confined Space Dangers at Mine Reclamation Sites. presented at the Canadian Institute of Mining, Metallurgy and Petroleum AGM Conference and Exhibition, Vancouver May 9-12, pp.22.  Mohammadi, L., and Meech, J.A., 2009, Fuzzy Risk Assessment of Confined Spaces at Mine Reclamation Sites. presented to the Chief Inspector of Mines and other mine safety personnel at the Ministry of Energy, Mines and Petroleum Resources, Victoria, May 5, pp.27.   vi Table of Contents  Abstract ......................................................................................................................................... ii Preface.......................................................................................................................................... iii Table of Contents ........................................................................................................................ vi List of Tables................................................................................................................................ ix List of Figures ............................................................................................................................... x List of Acronyms ........................................................................................................................ xii List of Symbols .......................................................................................................................... xiv Acknowledgements .................................................................................................................... xv Dedication ................................................................................................................................. xvii Chapter 1: Introduction ................................................................................................................ 1 1.1. Statement of the Problem .............................................................................................................. 1 1.2. Objectives and Contributions of the Work ................................................................................... 3 1.2.1. Motivations and Objectives ......................................................................................................................... 3 1.3. Thesis Overview .............................................................................................................................. 5 Chapter 2: Confined Space Definition, Fatalities and Regulations ........................................... 7 2.1. Overview .......................................................................................................................................... 7 2.2. Review of Confined Space Fatalities ............................................................................................ 8 2.2.1. In the United States ..................................................................................................................................... 8 2.2.2. In Canada ................................................................................................................................................. 17 2.2.3. In British Columbia .................................................................................................................................... 18 2.2.4. Multiple Fatalities ...................................................................................................................................... 20 2.3. Confined Space Definitions and Characteristics ....................................................................... 22 2.3.1. Occupational Safety and Health Administration (OSHA) ........................................................................... 23 2.3.2. National Institute for Occupational Safety and Health (NIOSH) ................................................................ 25 2.3.3. Definition of a Confined Space by CCOHS and WorkSafeBC .................................................................. 25 2.3.4. WorkSafeBC ............................................................................................................................................. 26 2.3.5. B.C. Mines Act .......................................................................................................................................... 27 2.4. Shortcomings of Confined Space Regulations ......................................................................... 29 2.5. Summary ........................................................................................................................................ 30 Chapter 3: Sullivan Mine Accident ............................................................................................ 32 3.1. Overview ........................................................................................................................................ 32 3.2. Sullivan Mine Tragedy .................................................................................................................. 33 3.3. Controlling Factors ....................................................................................................................... 38 3.3.1. Direct Factors ............................................................................................................................................ 38 3.3.2. Indirect Factors - Role of Oxygen-Depleted Water in Consuming Oxygen ............................................... 42 3.4. Similar Confined Space Accidents around the World............................................................... 52 3.4.1. Earth Dam at Carsington, U.K. .................................................................................................................. 52 3.4.2. Breathing Water Well in Alberta ................................................................................................................ 53 3.4.3. Breathing Coal Mines ................................................................................................................................ 54 3.4.4. Asphyxiation from Oil-Contaminated Soil .................................................................................................. 58 3.4.5. Soil Penetration by CO Gas ...................................................................................................................... 58 3.5. Summary ........................................................................................................................................ 61 Table of Contents vii Chapter 4: Analysis of Factors that Contributed to the Sullivan Mine Accident .................... 62 4.1. Overview ........................................................................................................................................ 62 4.2. Effect of Atmospheric Conditions on Confined Space Risk .................................................... 62 4.2.1. Development of Theory ............................................................................................................................. 63 4.2.2. Results of the Theory ................................................................................................................................ 67 4.3. Summary ........................................................................................................................................ 82 Chapter 5: Atmospheric Fuzzy Risk Assessment for Confined Spaces2 ................................ 84 5.1. Overview ........................................................................................................................................ 84 5.2. Research Problem ........................................................................................................................ 85 5.3. Methodology .................................................................................................................................. 87 5.3.1. Fuzzy Logic Rulebase ............................................................................................................................... 87 5.3.2. Other Confined Space Systems ................................................................................................................ 88 5.3.3. Atmospheric Fuzzy Risk Assessment (AFRA) .......................................................................................... 90 5.3.4. Transition from an Enclosed Structure to a Confined Space ..................................................................... 96 5.3.5. Fuzzy Rules Used Within AFRA................................................................................................................ 97 5.3.6. Detailed Risk Assessment Architecture .................................................................................................. 103 5.3.7. Gas Emission .......................................................................................................................................... 106 5.3.8. Permeability ............................................................................................................................................ 108 5.3.9. Gas Generation (Oxygen-Depletion) ....................................................................................................... 109 5.3.10. Dump Reactivity ................................................................................................................................. 112 5.3.11. Cover Effectiveness ........................................................................................................................... 114 5.3.12. Internal Temperature .......................................................................................................................... 117 5.3.13. Gas Confinement ............................................................................................................................... 120 5.3.14. Human Exposure ................................................................................................................................ 122 5.4. Simplified Representation of the Rules in AFRA ..................................................................... 122 5.5. Validation and Verification of the AFRA Model ....................................................................... 123 5.6. Risk Assessment under Different Climate Conditions ........................................................... 128 5.7. Cyclic Behaviour of a Reactive Waste Dump ........................................................................... 130 5.8. General Atmospheric Fuzzy Risk Assessment ........................................................................ 139 5.9. Summary ...................................................................................................................................... 140 Chapter 6: Compact Version of Confined Space Risk Assessment ...................................... 141 Chapter 7: Concluding Chapter ............................................................................................... 146 7.1. Conclusion ................................................................................................................................... 146 7.1.1. Contributions ........................................................................................................................................... 150 7.2. Recommendations ...................................................................................................................... 151 7.3. Future Work ................................................................................................................................. 153 Bibliography .............................................................................................................................. 155 Appendix A Statistics on Confined Space Fatalities .............................................................. 170 Appendix B Calculating Oxygen-depletion in the Sump by Oxygen-Deficient Water .......... 176 B.1. No Leakage of Air in to the Shed and Sump. ................................................................................ 176 B.2. Considering Gas Flow from the Pipe into the Sump .................................................................... 177 Appendix C Atmospheric Confined Space Manual (Gasses from Mines and Soils) ............ 178 C.1. Atmospheric Confined Space Manual ............................................................................................ 178 C.1.1. Intent ........................................................................................................................................................... 178 Table of Contents viii C.1.2. Application .................................................................................................................................................. 178 C.1.3. Definitions ................................................................................................................................................... 178 C.1.4. Reasons for Inclusion of This Protocol ........................................................................................................ 181 C.2. Plant and Equipment Requirement ................................................................................................ 182 C.3. Procedural Requirements ................................................................................................................ 183 C.4. People Requirements ....................................................................................................................... 184 C.5. Assessment in the Work Place ....................................................................................................... 187 C.6. Entry Permit ...................................................................................................................................... 187 C.7. Processes that Cause Oxygen  Depletion ....................................................................................... 191 C.8. Common Materials that Cause Asphyxiation, Poisoning or Explosions in Confined Spaces  192 C.8.1. Acetylene .................................................................................................................................................... 192 C.8.2. Ammonia Fume (NH3) ................................................................................................................................. 192 C.8.3. Argon (Ar) ................................................................................................................................................... 192 C.8.4. Cadmium .................................................................................................................................................... 192 C.8.5. Carbon Dioxide (CO2) ................................................................................................................................. 192 C.8.6. Carbon Monoxide (CO) ............................................................................................................................... 196 C.8.7. Chlorine (Cl2) .............................................................................................................................................. 196 C.8.8. Hydrogen (H2) ............................................................................................................................................. 196 C.8.9. Hydrogen Sulfide (H2S) ............................................................................................................................... 196 C.8.10. Nitrogen (N2) ............................................................................................................................................. 200 C.8.11. Nitrogen Dioxide (NO2) ............................................................................................................................. 201 C.8.12. Nitrogen Oxide (NO) ................................................................................................................................. 202 C.8.13. Methane (CH4) .......................................................................................................................................... 202 C.8.14. Sulfur Dioxide (SO2) .................................................................................................................................. 204 C.8.15. Other Contaminants in Confined Spaces .................................................................................................. 205 C.9. Guideline for Blasting Fume Migration Control to Residences ................................................... 206 C.10. Gas Migration and Pathways in Reclaimed Mines ...................................................................... 209 C.11. Dangerous Limits of Explosive and Toxic Gasses ..................................................................... 212 Appendix D Hazardous Gasses from ARD in Sulfide/ Coal Mines ........................................ 214 Appendix E Literature Knowledge Acquisition for Gas Generation and Emission .............. 222 E.1. Convective and Diffusive Gas Flow in Waste Dumps................................................................... 222 E.2. Gas Generation and Emission ........................................................................................................ 227 E.3. Intermediate Factors ........................................................................................................................ 231 E.3.1. Permeability ................................................................................................................................................ 231 E.3.2. Reactivity .................................................................................................................................................... 236 E.3.3. Cover .......................................................................................................................................................... 241 E.3.4. Internal Temperature ................................................................................................................................... 244 Appendix F Input Ranges of Reference Dump Properties for AFRA. .................................... 247 Appendix G Examples of Predictions by AFRA CE ................................................................ 256 List of Tables  ix  List of Tables  Table 2-1. Confined space fatalities from 1980-89 reported by NTOF ................................... 13 Table 2-2. Asphyxiation or poisoning deaths from 1984-86 reported by OSHA IMIS ............. 14 Table 2-3. Number of deaths caused by atmospheric hazards in a confined space .............. 14 Table 2-4. Death from poisoning, inhalation, ingestion in Canada ......................................... 17 Table 2-5. Multiple fatalities in confined space accidents by NTOF ....................................... 20 Table 4-1. Fuzzy rules to predict gas velocity from free energy difference ............................ 78 Table 5-1. Weights to calculate DoB in "High" gas emission at dump bottom...................... 107 Table 5-2. Weights to calculate DoB in "High" permeability ................................................. 108 Table 5-3. Factors affecting "High" gas generation (oxygen-depletion within the dump) ..... 111 Table 5-4. Factors affecting waste rock dump reactivity. ..................................................... 113 Table 5-5. Factors affecting cover effectiveness to control oxygen transfer ........................ 116 Table 5-6. Factors affecting internal temperature. ............................................................... 118 Table 5-7. Initial degree of belief in confinement for different enclosed structures ............... 120 Table 5-8. Estimated and measured internal temperatures in reference and test dumps. ... 124 Table 5-9. Summary of reference and test dump risk assessment ...................................... 125 Table 5-10. Atmospheric Risk of reference and test dumps for different scenarios.. ........... 127 Table 5-11. Estimated risk for different climates during the summer ................................... 129 Table 6-1. Factors affecting the DoB in the presence of an atmospheric hazard. ................ 145 Table 6-2. Factors affecting the DoB in not entering the space without SCBA. ................... 145 Table A-1. Confined-space accident deaths from 1980 to 2009 in the United States .......... 170 Table A-2. Asphyxiation in different industries in 1984 in 47 states of US,1984-1986 ......... 171 Table A-3. Death from inhalation and oxygen-deficiency in a confined spaces in mining .... 171 Table A-4. Work-related fatalities in Manitoba, 1995-2005 .................................................. 172 Table A-5. Number of confined space related fatalities in British Columbia ......................... 173 Table A-6. Oxygen deficiency fatalities in BC 1997-2009 in all sectors ............................... 175 Table C-1. Risk assessment table in the workplace. ........................................................... 187 Table C-2. Theoretical volume of carbon monoxide from commercial explosives ................ 207 Table C-3. Lower and higher limits for explosive gasses ..................................................... 212 Table C-4. Adverse health effects of different concentrations of toxic gasses and O2 ......... 212 Table D-1. Pore gas in soil above mineral deposits............................................................. 221 Table F-1. Selected confined structure (sump is this case) properties ................................ 247 Table F-2. Selected pathway (pipe is this case) properties. ................................................ 247 Table F-3. Estimation of undetermined permeability. .......................................................... 248 Table F-4. Estimation of cover effectiveness. ...................................................................... 249 Table F-5. Estimation of dump reactivity ............................................................................. 250 Table F-6. Estimation of gas velocity .................................................................................. 251 Table F-7. Estimation of gas generation ............................................................................. 252 Table F-8. Estimation of gas emission ................................................................................ 253 Table F-9. Risk Assessment for warmest seasonal period in time....................................... 254 Table F-10. Estimation of undetermined internal temperature ............................................. 255   List of Figures  x  List of Figures  Figure 2-1. Confined-space accident deaths in the United States 1980 to 2006...................... 9 Figure 2-2. Industrial confined-space death rates in the U.S. from 1980 to 1989 .................. 15 Figure 2-3. Death from inhalation in confined spaces in mining industry in U.S..................... 16 Figure 2-4. Confined space related fatalities in the province of British Columbia ................... 19 Figure 3-1. No. 1 Shaft Waste Dump at the Sullivan Mine prior to covering. ......................... 34 Figure 3-2. Seepage collection ditch in September 2005 prior to covering with glacial till ..... 35 Figure 3-3 Sullivan Mine No. 1 Shaft Waste Dump after sealing with glacial till..................... 36 Figure 3-4. Correlation of air velocity in the pipe with atmospheric temperature .................... 40 Figure 3-5. Time series of air velocity and atmospheric temperature .................................... 40 Figure 3-6. Schematic Diagram of the sampling shed at Number One Shaft waste dump. .... 41 Figure 3-7. Gas velocity, oxygen and carbon dioxide concentrations .................................... 42 Figure 3-8. The sump at the bottom of the sampling shed. ................................................... 43 Figure 3-9. Oxygen equilibrium in water as a function of oxygen content in air ..................... 44 Figure 3-10. Oxygen concentration of sump air in contact with oxygen-depleted water ......... 46 Figure 3-11. Oxygen-depletion of the air in the sump by oxygen-deficient water and air ....... 49 Figure 3-12. Gas velocity and temperature variation at time of the accident. ........................ 49 Figure 3-13. Time to consume the oxygen to 18% in the Sullivan mine sampling shed. ........ 50 Figure 3-14. Time to consume oxygen to 18% in a arbitrarily shed with A:V ratio of 4 m-1. .... 51 Figure 3-15. Safe/unsafe situations based on different air velocities and water flow rates. ... 51 Figure 4-1. (a) single-inlet single-outlet steady flow system – exergy flow (g2032);  ..................... 65 Figure 4-1. (b) Flow of energy to move gas in and out of a control volume. .......................... 65 Figure 4-2. Locations of the boreholes at No.1 Shaft Waste Dump ....................................... 68 Figure 4-3. Correlation of gas velocity in the pipe with ∆f. ..................................................... 69 Figure 4-4. Correlation of gas velocity in the pipe with ∆f (Summer, Fall and Winter, 2007) .. 70 Figure 4-5. Correlation of gas velocity in the pipe with ∆f (Spring 2007). ............................... 71 Figure 4-6. Representation of volume change in No. 1 shaft waste dump ............................. 71 Figure 4-7. Diurnal changes in g4666PO g3398 PIg4667, g4666TO g3398 TIg4667 and gas velocity ................................... 72 Figure 4-8. Relationship between pressure difference and temperature difference. .............. 74 Figure 4-9. Effect of seasonal changes on internal temperature of boreholes ....................... 80 Figure 5-1. Stages for the atmospheric risk assessment ....................................................... 92 Figure 5-2. Input data for reference waste dumps ................................................................. 95 Figure 5-3. Fuzzy sets defining levels of risk for a confined space. ....................................... 96 Figure 5-4. Fuzzy Associated Memory Map for Confined Space Risk ................................... 99 Figure 5-5. Overall Flowchart of AFRA ............................................................................... 100 Figure 5-6. Two-rule fuzzy expert system (inferencing and defuzzification) ......................... 102 Figure 5-7. Structural architecture of AFRA. ....................................................................... 103 Figure 5-8. Diagram showing different adaptation factors. .................................................. 104 Figure 5-9. Fuzzy representation of oxygen level. ............................................................... 109 Figure 5-10. Erosion channel at a rehabilitation site ............................................................ 120 Figure 5-11. Idealized risk and internal dump temperature over time  ................................. 134 Figure 5-12. Sensitivity analysis for various internal temperatures at No.1 Shaft Dump ...... 138 Figure 6-1. Overall Flowchart of AFRA CE .......................................................................... 144 Figure C-1. Nitrogen cycle in soil ........................................................................................ 201 Figure E-1. Oxidation rate at a sulfide waste dump (depth and time after construction) ...... 225 Figure E-2. Depiction of the three reaction zones in a sulfide waste dump .......................... 225 Figure E-3. Temperature profile in borehole 1A in No. 1 waste dump ................................. 226 List of Figures  xi Figure E-4. Oxidation rate declines when moving from the edges of the waste dump ......... 229 Figure E-5. Oxygen levels and temperatures in borehole 36 ............................................... 229 Figure G-1. AFRA CE Output for an Initial DoB in an atmospheric hazard of 0% (NO)  ...... 256 Figure G-2. AFRA CE Output for an Initial DoB in an atmospheric hazard of 30% .............. 257 Figure G-3. AFRA CE Output for an Initial DoB in an atmospheric hazard of 50%. ............. 258 Figure G-4. AFRA CE Output for an Initial DoB in an atmospheric hazard of 80%. ............. 259 Figure G-5. AFRA CE Output for an Initial DoB in an atmospheric hazard of 99%. ............. 260   xii List of Acronyms  AFRA Atmospheric Fuzzy Risk Assessment  AFRA CE Atmospheric Fuzzy Risk Assessment Compact Edition ARD Acid Rock Drainage BCEOHRN B.C Environmental and Occupational Health Research Network U.S. BLS U.S. Bureau of Labour Statistics CCOHS Canadian Center for Occupational Health and Safety is a Canadian federal   government agency which is responsible to prevent work-related injury and   illness. CDC Centers for Disease Control DoB  Degree of Belief FACE Fatality Assessment and Control Evaluation - a NIOSH program. It was called   Fatal Accident Circumstances and Epidemiology before 1992 (Suruda et al,  1993).  IMIS  Integrated Management Information System in OSHA IDLH Immediately Dangerous to Life and Health LEL                 Lower explosive limit of a particular contaminant.  HEL                Higher explosive limit of a particular contaminant.  LFL                 If a flammable gas concentration falls below its Lower Flammable Level it will                         burn. For example, the LFL of hydrogen gas in air is 4%. This range depends                         on to temperature, pressure and oxygen or chlorine percentage.                         (http://www.ccohs.ca/oshanswers/chemicals/compressed/compress.html?print).   MVT  Mississippi Valley Type   NIOSH National Institute for Occupational Safety and Health. NIOSH is part of the U.S.  Department of Health and Human Services and acts as a research agency   focused on education, and training in occupational safety and health. NIOSH  research is to identify health and safety risks and measure them in working places to reduce danger NTOF National Traumatic Occupational Fatality surveillance system developed by   NIOSH to collate data on traumatic work-related fatalities (Suruda et al, 1993).  OHS Occupational Health and Safety consultancy in U.K that provides a wide   spectrum of policies, guidelines, and activities focused on promoting and   xiii  managing health and safety and asbestos-issues in the work place as well as   occupational hygiene   OSHA Occupational Safety and Health Administration is a federal agency in the U.S.  Department of Labour, which has been active since 1971. According to U.S. DoL, OSHA publishes and enforces safety and health regulations for most businesses and industries in the U.S. to prevent work-related injuries, illnesses, and deaths.   SCBA   Self-Contained Breathing Apparatus   TLV-TWA        Threshold Limit Values -Time- Weighted Average is a time weighted average                          concentration of the toxic gas for 40 hours work week (8 hours a day), which  will cause no adverse health effect on nearly all workers after being exposed                          repeatedly during this time (Eltschlager et al., 2001(a)).  TLV                 Threshold Limit Value is the concentration that should not exceed during all                         exposures (Eltschlager et al., 2001(b)).  UFL                If a flammable gas concentration falls above its Upper Flammable Level it will                          burn. For example, the UFL of hydrogen gas in air is 75%. This range is                         dependent to temperature, pressure and oxygen or chlorine percentage.                         (http://www.ccohs.ca/oshanswers/chemicals/compressed/compress.html?print).   UN  Undetermined  WCB  Workers Compensation Board of Manitoba  WHMIS  Workplace Hazardous Materials Information System  WorkSafeBC Workers Compensation Board in British Columbia                      xiv List of Symbols  Cg3364 g2926,g2911g2932g2917= Heat capacity of the air or gas = about 1.099 [J·g-1K-1],  ∆f = Free energy of the unit-mass [J·g-1],  Ψ = Flow exergy of a unit-mass of the gas [J·g-1],  gz = Potential energy of the unit-mass [m2·s-2],  h = Enthalpy of the unit-mass of the gas [J·g-1],  P = Pressure [Pa],  qk  = Heat transfer to the unit-mass of the gas [J·g-1],  R = 0.287 [J·g-1K-1],  s = Entropy of the unit-mass [J·g-1K-1],  T = Temperature [K],  V2 2  = Kinetic energy of the unit-mass [m 2 ·s-2],  v = Unit volume, v= RT1P1  [m 2 ·g-1],  Vg4652g4652g1318= Velocity [m·s-1],   w  = Work interactions per unit mass of the working gas on the outside [J·g-1]  x = Exergy of the unit-mass [J·g-1],  z  = Height [m],      xv Acknowledgements  My special gratitude goes to my supervisor, Dr. John A. Meech, who taught me how to think effectively and question concepts deeply in order to solve complex problems. He helped me gain the confidence necessary to conduct my research independently and showed me how to effectively present my work. He fostered my view of problems broadly and practically while patiently helping and protecting me through the different stages of the research. Without Dr. Meech’s novel ideas and initiative coupled with his strong belief for the necessity of this work, this research would have never been started nor completed. I thank him for offering me a new way of living by providing me with the opportunity to study at UBC and changing my life for the better. I would like to warmly thank the kindness, support, and care of Dr. Leslie Lavkulich during my studies as well as the useful advice of Dr. Marcello Veiga. I offer my enduring gratitude to my adorable parents, Mr. Khosrow Mohammadi and Mrs. Behnaz Akhavan, who have supported me all through the years of my education both emotionally and financially, especially my father who has always inspired me to continue beyond my very best. I owe particular thanks to my husband and soon to be naturopathic doctor, Hamid Reza Tajbakhsh who has patiently supported me during both good and tough times of my Ph.D. studies and managed my stress with all of the naturopathic tools he could use, without whom this work would have been much more difficult to complete. My greatest thanks goes to my friend, Ali AbdulHussein, for his heart warming help in my early years of study. My younger brother, Amirahmad Mohammadi, who I could always rely on for discussing problems of my thesis, also deserves my sincere gratitude. My thanks also go out to my older sister, Katayun Mohammadi, who encouraged me to focus on my studies by always mentioning the value of education when we were younger. I also thank all other faculty and staff at UBC who have helped me through this journey especially in my early years in Canada. Dr. Ali Asghar  xvi Hassanipak at University of Tehran also deserves my gratitude since he encouraged me to not only start my masters’ degree but also to apply to the University of British Columbia in pursuit of my PhD.  I would like to thank WorkSafeBC (Workers' Compensation Board of British Columbia), and the Workers’ Compensation Board of Nova Scotia for their financial support during this project. I also appreciate the reviews and advice from two anonymous external experts at WorkSafeBC on the final project report. The reviewers' points of view helped the project in terms of making it more understandable and presentable. My gratitude also extends to the members of the Technical Advisory Panel on the Sullivan Mine Incident formed by Teck Resources (formerly Teck-Cominco) for their input and comments: Walter Kuit and Bruce Dawson from Teck Resources; Ricci Berdusco, (resigned 2008), Phil Pascuzzi (resigned 2008), Al Hoffman, Kim Bellefontaine (since 2008), and Diane Howe (since 2008) from B.C. M.E.M.P.R.; Clem Pelletier from Rescan Environmental (resigned 2008); Dr. Andy Robertson from Robertson GeoConsultants; Dr. Ward Wilson and Dr. John Meech from the University of British Columbia; Mike O'Kane and Mark Phillip from O'Kane Consultants; Daryl Hockley from SRK Consulting.  Finally, I would also like to thank all the researchers who have done investigations on waste dumps since without their papers and insights I would not have been able to build and validate my model and make a contribution to the field of mine waste management and engineering.      xvii Dedication    This thesis is dedicated to the victims who lost their lives at the Kimberly accident between May 15 and 17, 2006.  • Douglas Erickson, 48, a contractor from nearby Ha Ha Creek. • Robert Newcombe, 49, a Teck-Cominco employee who lived in Cranbrook. • Kim Weitzel, 44, a paramedic from Kimberley. • Shawn Currier, 21, a paramedic who lived in Cranbrook.    1 Chapter 1  1.  Introduction  1.1. Statement of the Problem In May of 2006, a tragic accident took place at the Sullivan mine in Kimberley, British Columbia. Over a period of about 36 hours, four people died as a result of each of them entering into a sampling station located at the toe of the Number One Shaft Waste Dump (http://thetyee.ca/News/2007/07/09/MineDeaths/; Sullivan Mine Incident Technical Panel, 2010; http://www.bcas.ca/EN/main/news/newsArchive/4274/report-confirms-unprecedented- incident.html). Acid Rock Drainage (ARD) effluent from the waste dump had caused oxygen- depletion to occur within this structure transforming it from a simple "shed" to an extremely dangerous confined space. The direct reason for these fatalities was the concentration of oxygen-depleted gas emissions from the waste dump from a buried pipe into the sump at the bottom of the sampling shed.  Pollution from mine wastes may be emitted into different environments (the hydrosphere, the lithosphere, and the atmosphere). Risk assessment in waste dump reclamation programs is generally done to protect the environment and to develop construction rules to handle issues such as acidic water drainage collection, erosion, slope stability, and liquefaction (Kaczynski, 1986; Amanti et al., 1996; Chaulya et al, 1993; Chaulya et al, 2000; Engels et al., 2007; Kreft-Burman et al., 2007). With the exception of spontaneous combustion, dangerous atmospheric releases from a waste dump have not been regularly assessed or even recognized as a realistic need. It is now apparent that a new risk assessment method should also be applied to identify human occupational-health dangers at reclamation sites especially from atmospheric emissions (Mohammadi and Meech, 2008). It seems prudent that future permitting processes should include such a risk assessment approach. Currently, there is no Chapter 1 - Introduction 2 risk assessment tool available to use with reclamation sites or, for that matter, to apply a standardized procedure to other types of confined spaces with a possible atmospheric hazard. Neither is there a requirement on the part of operators to conduct such an analysis. Recognizing the presence of a hazard and evaluating the development of dangerous conditions at a site are essential first steps in the prevention of such accidents (Mohammadi and Meech, 2011).  Accidents of this kind occur because of a lack of knowledge transfer from experts on a variety of inter-related and inter-connected topics into the actual workplace. One way to accomplish this knowledge transfer is to create a software system to educate and train people about the key issues surrounding confined space hazards at mine reclamation sites. Such a system can be used routinely to perform a risk assessment at a site to discover hazards early and introduce mitigation practices. The consequence in the case of exposure to a confined space with an atmospheric hazard is death of any person who enters the space in an unauthorized or unsafe manner (Mohammadi and Meech, 2011). The degree of danger varies on an hourly, seasonal, and decadal basis, so recognition of the hazard is difficult to establish in situations where such dramatic changes in risk take place. Someone may occupy the space one evening without any problem while the next day the temperature has risen and the space is now deadly.  A major component of this research has been the creation of an Expert System tool to provide knowledge transfer from specialists to industry to workplaces. Regulating mine managers and workers to use this knowledge transfer tool while they design and implement mine closure practices can prevent a "blind" reliance on "past safe performance" of their industry (or the site) – which is one of the major reasons for failing to identify a confined space situation.  Atmospheric risk assessment is perhaps as important a task as is the environmental risk assessment of ARD impact on the aquatic environment and such analyses should be carried Chapter 1 - Introduction 3 out on a regular basis after mine closure. It is emphasized that both environmental and atmospheric risk analysis should be carried out in parallel to find site specific solutions that are best for both aspects. Regular atmospheric risk assessment is important because unknown future activities may be carried out at the site that can change the risk level. Certain provisions and rehabilitation activities at a closed mine may increase atmospheric risk. Over time, changes occur within the dump as sulfide minerals continue to oxidize consuming oxygen from pore gas, generating acid, and producing carbon dioxide from reaction of the acid with carbonate-type minerals. More importantly, temperature changes within the dump due to these reactions lead to periods of danger followed by dormant behaviour and then followed again by danger over decades of time. 1.2. Objectives and Contributions of the Work 1.2.1. Motivations and Objectives The specific objectives of the research have been to: 1. Investigate the causes of the accident that occurred at the site. 2. Study related accidents and current confined space regulations. 3. Identify potential gasses in waste dump systems (e.g. oxygen-depleted air, carbon          dioxide).  4. Identify principal factors that may control possible gas generation (GENERATION). 5. Gather information from experts, technical reports/papers and literature pertaining          to the waste dumps to identify environmental, operational, design and structural         factors that can cause hazardous gasses to be released from the waste dump         (EMISSION).   6. Investigate gas behaviour after emission (dispersion, trapped and/or vented).         (CONFINEMENT).   7.     Investigate the exposure of people to confined space (HUMAN EXPOSURE)  8. Build an atmospheric fuzzy risk assessment tool. 9. Evaluate and validate the effectiveness of the model. Chapter 1 - Introduction 4 10. Develop a tool for first-responders to help them avoid harm to themselves while performing a rescue in an enclosed structure with a potential atmospheric hazard.  All these steps in the research have been taken to support a tool that can help prevent future accidents. The methodology is considered applicable to other industrial sites however the focus in this research has been on the specific issues of mine reclamation.  The major objective is to design software to conduct an atmospheric risk assessment to assist in performing the analysis with an emphasis on identifying potential hazards to human health from poor atmospheric conditions. The software focuses mainly on oxygen depletion and the generation of carbon dioxide, but it also includes a "list of possible gasses in confined spaces" that covers toxic or flammable gas (or dust) generation at other places besides mine reclamation sites. The risk is assessed for waste dumps with sulfide contents from 0.5 to 5% and does not account for highly reactive waste dumps that may spontaneously combust.  In this thesis the values of the three elements - gas generation, gas emission and gas confinement - results in the likelihood of a hazard, which in combination with the value of the element of human exposure, will result in a final risk assessment. Since exposure of humans to a confined space with an atmospheric hazard leads to death in seconds or minutes, the likelihood of the consequence of exposure is 100 percent.  A compact version of this risk assessment tool has also been created to operate on a hand-held device to help rescue personnel–firemen, police, paramedics, and others to decide on entering or not a "permit-required" confined space even when signage is not in place. This thesis presents details of the development of AFRA - Atmospheric Fuzzy Risk Assessment Tool. A heuristic technique based on fuzzy logic was considered a preferred approach since this method mimics a dialog between humans to achieve successful transfer of information. Such systems can be tuned to be cautious or to allow risky behaviour – in this case, of course, the former is clearly preferred. Linguistic terminology is output rather than numerical values which can accelerate the understanding of the model by novice users. A Chapter 1 - Introduction 5 fuzzy logic rule-base allows for case–based reasoning using an If-Then rule structure. Although numerical methods are being developed with respect to the complex interactions between solid material, water, and air within a waste dump, these methods are in the early stages and considerable assumptions must be made to apply the mathematics appropriately. Waste dumps are extremely heterogeneous masses that show different behaviours spatially, temporally, and geographically. The nature of a problem that is difficult to predict renders these mathematical approaches to situations where dumps are heavily instrumented with input data being collected on an hourly basis over months and years. Such methods are both impractical and infeasible in most cases. Fuzzy logic rule-bases on the other hand, use "approximation arithmetic" of an unknown but correct model – the mathematics is subsumed within the system itself. The method is a precursor to the concept of "Computing with Words" in which perceptions are as important as measurements (Zadeh, 1965; Zadeh, 2002).  1.3. Thesis Overview An outline of the structure of this thesis is as follows: Chapter 2 presents an overview of data on confined space fatalities and regulations around the world. It compares confined space related fatalities in U.S., Canada, and the province of British Columbia. It discusses and compares different confined space regulations and points out their limitations and deficiencies. Chapter 3 describes the Sullivan Mine accident. It compares the accident with other similar atmospheric related confined space accidents. In Chapter 4 the factors that contributed to the Sullivan mine tragedy are discussed and analysed. This chapter prepares the reader to understand the phases that were considered in developing AFRA.    Chapter 5 is devoted to the development of a detailed and general Atmospheric Fuzzy Risk Assessment (AFRA) tool. In this chapter processes that contribute to oxygen-depleted gas within the dump and its emission to the outside atmosphere will be described in detail. Chapter 1 - Introduction 6 The verification and validation of AFRA are also described. Details of the development of a compact version of the tool for a Pocket PC are included in Chapter 6.  Appendix C includes an analysis of possible atmospheric gasses that can be hazardous at mine sites. A review is given of hazardous gasses that can be generated from minerals, soils, or any other material or activity at a mine site. An Atmospheric Confined Space Manual for Gasses from Mines and Soils is presented in Appendix C. This manual is an independent document on its own that can be printed and used. The document is designed to help mine engineers and managers understand and recognize the presence of an atmospheric problem at their work place. A new way to describe a confined space hazard is suggested based on the short-comings of existing descriptions.  Each chapter ends with a summary to conclude on the key findings, while the last chapter provides final conclusions and summarizes the contributions, as well as offering recommendations for future work in this field of study.   7  Chapter 2  2. Confined Space Definition, Fatalities and Regulations  2.1. Overview Confined space accidents expose humans to conditions that can cause immediate death. These accidents typically involve asphyxiation due to emission of a hazardous gas into an unventilated enclosed space. The air may contain too little oxygen; or, it may contain poisonous, flammable, combustible or explosive agents; or, contaminants (e.g., fume, dust, or mist) that pose an immediate threat to life or interfere with a person's ability to escape. The gas may be from contaminated soil or other processes, or it may be air that is oxygen- depleted. Confined space accidents may also involve engulfment in bulk materials (soil, grain, snow, etc.) that flow into the space to cover the victims. Rapid influx of water may lead to drowning (Mohammadi and Meech, 2008 – Note many of these hazards are contained within the Confined Space Entry Program - A Reference Manual, 2005 and in OSHA, 2004). The statistics for confined space accidents in the U.S. during the 1980s collected by NIOSH show that of 670 fatalities over 55% were due to atmospheric conditions with about one third caused by engulfment. The remaining deaths were from falls, fires, or drowning (Suruda et al., 1994).  Of all hazards, a confined space atmospheric hazard is the most difficult for an individual to recognize especially when the issue is oxygen-depletion. There is nothing visual to provide warning; there is no odor and no noise. Death is quick and the victim rarely is aware of the danger as bad air enters the lungs to deplete oxygen from the blood resulting in Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 8 unconsciousness within 10 to 40 seconds. It is a painless death, but in all cases, it can be prevented by following proper safety procedures as outlined in Appendix C.  It is estimated that more than 1.5 million workers actually enter a confined space annually (Confined Space Policy, 2006). The Occupational Safety and Health Administration (OSHA) suggest that 85% of confined space accidents can be prevented if proper safety precautions at job sites are implemented. Of course, each site poses a serious problem for exposed workers and their employers (Confined Space Policy, 2006) and so, management of confined spaces by personnel and managers is an extremely important task. But, in order to follow proper procedures, one must first identify that a confined space hazard exists. Not all confined structures are hazardous–an ordinary room with the windows and doors shut is a confining space, but with only a few occupants, it is not a problem as air seeps in under doors and through heating ducts. An outhouse is a potentially dangerous confined space through toxic gas emissions (hydrogen sulfide and methane) from the pit below. Most outhouses are designed with good ventilation and so, such facilities are not classified as dangerous and no deaths have yet been reported for this type of structure. But consider an outhouse covered in snow which insulates the structure and blocks vents preventing air from entering. Under these conditions, even an outhouse could become dangerous (Mohammadi and Meech, 2008). 2.2.  Review of Confined Space Fatalities  2.2.1. In the United States Statistics on confined space accidents have been well-documented over the past quarter century in the United States for the period 1980 to 1989 by NIOSH and from 1992 to 2009 by the U.S (Mohammadi and Meech, 2008). Bureau of Labour Statistics (BLS) (Suruda et al., 1994; U.S. BLS, 2010). Figure 2-1 shows the numbers of confined space related fatalities due to inhalation of toxic gasses or oxygen-deficient air or from oxygen-deficiency caused by Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 9 cave-ins or collapse of materials since 1980 (Suruda et al., 1994; U.S. BLS, 2010). The data in Figure 2-1 show a steadily improving trend as the number of confined space accidents per year has significantly declined (Mohammadi and Meech, 2008). Table A-1 in Appendix A gives more details on these data which includes the number of fatalities related to each cause. It is difficult to compare confined space related fatalities from different sources. This thesis compares work-related confined space fatalities, but even in this case, some industries may have been excluded from the system used to gather the information and/or some groups may not have been counted in the surveillance. The type of exposure that causes each accident varies from one surveillance system to another. To make the data more understandable, the types of work related fatalities, type of exposure, and source of gathering the data needs to be clarified.     Figure 2-1. Confined-space accident deaths in the U.S. 1980 to 2006 (Mohammadi and Meech,                      2008) - (Data before 1989 from NIOSH; Data after 1992 from U.S BLS (U.S. BLS,                      2010)). (Data in both NIOSH (by CDC) and BLS are public domain and do not need                      specific permission) 0 20 40 60 80 100 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 Nu m be r o f F at al iti es Year Confined Space Industrial Fatalities (U.S) after NIOSH and BLS Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 10 Figure 2-1 summarizes data on confined space fatalities from 1980 to 1989 (Suruda et al., 1994) based on the National Traumatic Occupational Fatality (NTOF) surveillance system developed by NIOSH to collate data on traumatic work-related fatalities. The NTOF database is a census of traumatic work-related fatalities based on all U.S. death certificates in which the death or injury that caused the death occurred at a work place (Suruda et al, 1993). NTOF contains information from the death certificate that states where and how a death occurred at any workplace. The certificates indicate the cause of death and provide details of the death as well as the opinion of the certifying coroner, medical examiner, or physician. The types of hazards considered by NIOSH were investigated by Suruda et al., (1994) and included asphyxiation, poisoning, and drowning. A fatality was included only if the accident took place in a confined place such as a vat, a pit, a silo, a tank or a bin. If the site of the accident was not indicated in cases of gas poisoning (methane, hydrogen sulfide, etc.), it was still considered as a confined space accident.  Fatalities from engulfment in bulk materials were included but those due to a roof-fall or mine-caving were excluded. Death by explosion, trench cave-in, electrical shock, temperature extremes, barrier failures, radiation or physical hazards such as being hit by moving equipment and machinery or due to a slip or fall were excluded. NIOSH admits that its data on the total number of confined space fatalities is probably lower than the true value, since it is difficult to recognize all deaths in a confined space from the information available in all death certificates (Suruda et al., 1994). There is no integrated source of data or study for atmospheric-related confined space fatalities prior to 1982. However, according to the California Department of Labour Statistics and Research, only 21 of 1,011 (2%) work-related fatalities for 1981-1982 involved confined spaces (Suruda et al., 1994) while NIOSH reported 143 confined-related fatalities during 1976-1977 (average of 72). So it appears that the average yearly fatalities in the U.S. have dropped from 1977 to 1986 by about 20% (Suruda and Agnew, 1989).  Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 11 Data from 1992 were gathered by the BLS Census of Fatal Occupational Injuries (CFOI). BLS CFOI gathered accurate data of fatal work injuries in all 50 US states and the District of Columbia since 1992. Data from this source were extracted for all types of industrial accident deaths caused by atmospheric reasons for 3 categories of exposure in confined spaces: 1. Depletion of oxygen in enclosed, restricted, or confined spaces; 2. Inhalation in enclosed, restricted, or confined spaces;  3. Depletion of oxygen from cave-in or collapsed materials. Although confined space related fatalities were not itemized in all categories for all years since 1992, all categories were present from 1992 to 2003 while for 2004 to 2006, the last category is absent and for 2007 to 2009, only the second category is itemized. These data do not include drowning in confined spaces or trench cave-ins as neither of these causes was itemized. So, with the exception of drowning, the BLS data are considered compatible with those of NIOSH. Both sources provide good indications of work-related confined space fatalities in all states (U.S. BLS, 2010).This system uses various sources to identify fatal worker injuries. For each work related fatality - occupation and other worker characteristics, equipment involved, and circumstances of the event - are obtained from source records, such as death certificates, workers' compensation reports, federal and state agency administrative reports and claims, reports to various regulatory agencies, medical examiner reports, or police reports, news and other non-governmental reports. The CFOI data considers the fatalities related to paid-workers that were injured during the conduct of their employment. It also counts volunteer and unpaid family workers who perform the same duties as paid workers. The data include all fatal work injuries for workers covered by OSHA or other federal or state agencies. Certain industries such as mining and highway, water, rail, and air transportation, are not covered by OSHA because they are regulated by other Federal agencies, such as the Mine Safety and Health Administration and various agencies within the Department of Transportation. Those accidents outside the scope of regulatory coverage Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 12 such as self-employed and unpaid family workers are not covered by any Federal or State agencies. (U.S. BLS, 2010; Suruda and Agnew, 1989). Between 1980 and 1989, according to NTOF, 305 of 1218 (25%) asphyxiation-related fatalities took place in confined spaces (of these 305 fatalities, 31 were by obstruction of the respiratory tract and the other 274 involved mechanical suffocation); confined spaces were also involved in 274 of 1018 (27%) poisoning-related deaths and 91 of 947 (10%) drowning.Table 2-1 categorizes confined space work-related fatalities during these years by cause of accident. Of the 670 fatalities, over 55% were due to atmospheric conditions with about a third of the deaths caused by engulfment. The remainder were from drowning or other causes (Suruda et al., 1994). Clearly, atmospheric hazards are the primary causes of death in confined space accidents (Suruda et al., 1994). There are few studies that categorize atmospheric hazard related fatalities based on type of harmful agent. Suruda and Agnew (1989) have extracted detailed causes of atmospheric-related fatalities for the period between 1984 and 1986 from the OSHA Integrated Management Information System (IMIS).  According to this work, 233 deaths were caused by asphyxiation or poisoning from the total of 4,756 deaths in 47 U.S. states between 1984 and 1986. Of the 233 fatalities presented in Table 2-2, 146 occurred in confined spaces from asphyxiation by oxygen- deficient air or poisoning by gasses or chemicals. It is not apparent how many fatalities which occurred in confined spaces were related to each agent. 27 (11%) of the deaths were related to oxygen-deficiency which most likely occurred in an enclosed area. This number is more than half of the fatalities caused by all other simple asphyxiants in Table 2-2 (Suruda and Agnew, 1989). OSHA data shows fewer fatalities than NIOSH (based on NTOF) during 1984- 1986, 146 compared to 207 since the OSHA database is not as complete as that of NIOSH and do not include all confined space fatalities. OSHA reports include all work related fatalities in the United States for which OSHA has jurisdiction (only three states were omitted from this data: California, Washington, and Michigan). The OSHA database excludes mining, Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 13 transportation, maritime, federal employees, agricultural workers and other sectors regulated by other agencies. For example, the number of total fatalities from 1984-1986 was 4,755 in the OSHA database while NIOSH reports 6,258 work-related deaths during these years. A study in Pennsylvania in 1983-1984 and one in Colorado in 1982-1984 indicate that OSHA only reports about 60% of all occupational traumatic deaths within the industrial sectors for which it has jurisdiction (Suruda and Agnew, 1989). Suruda et al., (1994) have also extracted the cause of atmospheric related fatalities in confined spaces between 1980-1989 from NTOF data – see Table 2-3. Their data show ~17% of the fatalities in this period were related to oxygen-depletion. This is higher than the number of deaths caused by other specified gasses.  During 1984 to 1986, 15 deaths in the U.S. occurred because of using the right respirator wrongly or using the wrong respirator. Of this number, 2 workers died in a confined space due to the use of half-face cartridge respirators in an oxygen-deficient atmosphere instead of a Self-Contained Breathing Apparatus (SCBA). Two workers and one rescuer died when the SCBA ran out of air. One worker died when the air-supplied respirator hose became loose during sandblasting. The 9 other fatalities were related to improper supply of poor air, i.e., contaminated with CO from compressor motor exhaust, nitrogen or argon (Suruda and Agnew, 1989). The proper self contained breathing apparatus or air-supplied respirator should have been used in these cases.  Table 2-1. Confined space fatalities from 1980-89 reported by NTOF (Suruda et al., 1994).                   (Data are public domain in CDC and does not need specific permission)  Cause of accident % of Confined Space Fatalities Number of Fatalities Atmospheric   56 373 Engulfment in Bulk Material   34 227 Other   10   70 Total 100 670     Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 14 Table 2-2. Asphyxiation or poisoning deaths from 1984-86 reported by OSHA IMIS in 47 states                     in U.S. (number in front of each agent is the number of victims) "Adapted by                      permission from BMJ Publishing Group Limited. [from Asphyxiation and Poisoning                     at Work in the United States 1984-86, Suruda, A.J., Agnew, J., v 46, 541-546, 2011] ".  Cause of Accident Agents No. of Fatalities Simple asphyxiant Propane (1), Methane (10), Argon (7), Nitrogen (18), Carbon dioxide (12) 48 Toxic gasses Ammonia (3), Carbon monoxide (25), Chlorine (3), Hydrogen sulfide (30), Nitrous oxide (4) 65 Oxygen-deficient atmosphere Substance not reported (lack of oxygen) 27 Mechanical Asphyxiation - 42 Solvents Chlorodifuoromethane (1), Coal tar pitch volatiles (1), Dichlorodifluoromethane(2), Diesel fuel (1), Gasoline (3), Isopropanol (1), Methyl chloride (1), Methyl chloroform (4), Methylene chloride (8), Naphtha (3), Perchloroethylene (1), Toluene (2), Trichlorotrifluoroethane (F-113) (4), Trichloroethylene (2), Trichloroethane (1) 35 Undefined or  incomplete report  7 Other Dimethyl sulfate (1), Maleic amhydride (1), Sulphuryl fluoride (1), Cyanide (1), Hydrogen fluoride (4), Phosphorus (yellow) (1) 9 Total  233   Table 2-3. Number of deaths caused by atmospheric hazards in a confined space noted on                    death certificate identified by NTOF, 1980-1989 ((Data are public domain in CDC and                    does not need specific permission, Suruda et al., 1994).  Type of      atmospheric hazard % of Atmospheric Fatalities Number of Fatalities H2S 13.7 51 Methane 10.2 38 Inert gasses 8.6 32 Carbon monoxide 6.7 25 Sewer gasses 6.7 25 Oxygen-deficiency 16.6 62 Other gasses 16.6 62 Unknown atmosphere 20.9 78 Total 100.0 373       Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 15   Figure 2-2. Industrial confined-space death rates in the U.S. from 1980 to 1989 (Table is a public                      domain in CDC and does not need specific permission, Suruda et al., 1994)  Figure 2-2 shows the rate of industrial-related confined-space deaths in the U.S. from 1980 to 1989. Of all industries, mining-oil and gas has a significantly higher rate per 1000 workers than do other sectors (Suruda et al., 1994). No industry is free of these tragic occurrences with events taking place even in the rather commonplace fields of public administration, retail trade, service sector, and real-estate, insurance, and finance (Mohammadi and Meech, 2008) – see Table A-2 for more details. Data on atmospheric- related confined space accidents in mining are less complete than data for other workplaces. Figure 2-3 shows that 40 fatalities occurred in confined spaces in mining since 1992, of which 3 death in 2002 were due to oxygen-depleted air in an enclosed area (U.S. BLS, 2010) – see Table A-3 in Appendix A. In the BLS data, death by inhalation includes asphyxiation, strangulation, or suffocation involving chemicals or poisonings and toxic effects (except drugs, alcohol, or medicine) which 0 1 2 3 4 5 6 7 Fa ta lit ie s/ 10 00 ,0 00  w or ke rsChapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 16 are all categorized as traumatic injuries. For inhalation-related fatalities, the agents and details of the accidents are not given. The fact that these kinds of accidents are continuing to happen in mining causes us to question the current effectiveness of general confined space mining regulations.    Figure 2-3. Death from inhalation and oxygen-deficiency in confined spaces in the U.S. mining                      industry (U.S. BLS, 2010). Data in BLS is public domain and do not need specific                      permission)  According to the Mine Safety and Health Administration (MSHA) in 2003, the major atmospheric-related deaths were 33 cases of black lung and 7 cases of silicosis in surface mining (U.S. Department of Health and Human Services, 2005). The majority of all fatalities occurred amongst sand and gravel operators with the remainder taking place in surface coal mines. Although the reported fatalities are related to chronic atmospheric hazards that occur after many years of exposure to poor atmospheric conditions (dust and gasses), the data by BLS on mining related fatalities (in Figure 2-3) shows the significance of the need for atmospheric risk assessment. According to the NIOSH Surface Mining Fact Sheets, about 85% of mining industry employees work in surface mining making this task more important. 0 1 2 3 4 5 6 7 8 1992 1994 1996 1998 2000 2002 2004 2006 2008 Number of Fatalities Year Confined Spaces Fatalities in the U.S. Mining Industry after NIOSH and BLS Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 17 2.2.2. In Canada Statistics Canada provides data on deaths caused by inhalation of toxic and poisonous gasses. Threats to breathing in this database are classified as accidental or intentional and the type of death is unspecified. Accidental inhalation-related deaths are shown in Table 2-4, however it is not possible to tell which of these deaths were caused by a confined space. Although the report specifies the number of deaths from "being confined in a low-oxygen environment" the data does not seem reliable – for example in 2003 only a single death is reported for this cause despite the fact that British Colombia data show that five persons died in that year because of oxygen-deficiency in a confined space. This indicates little consistency in confined space fatality data for Canada. The Workers Compensation Board of Manitoba (WCB, 2006), indicates that 2% of the total work-related deaths occurred in confined spaces over the period 1995 to 2005. It is important to note that the total number of fatalities by these causes has not dropped in recent years unlike in the United States. See Table A-4 in Appendix A. Table 2-4. Death from poisoning, inhalation, ingestion and exposure to substances that cause                   obstruction of respiratory tract in 2000-07 in Canada (Adapted from the Statistics                    Canada CANSIM database http://cansim2.statcan.gc.ca, CANSIM Table 102-0540,                    August 3, 2011).   External Cause of Death  2000  2001  2002  2003  2004  2005  2006  2007   Death by being Confined in or Trapped in a Low Oxygen Environment (w81)   5  4  6  1  4  4  5  4  Death from Inhalation or Ingestion of objects obstructing the respiratory tract (w80)  153  160  186  202  163  182  177  223  Threat to breathing due to cave-in, falling earth and other substances (W77)  8  6  6  9  5  6  7  6  Accidental poisoning by and exposure to   Organic solvents and halogenated hydrocarbons and their vapours (X46)  7  6  5  4  8  5  4  4  Other gasses and vapours (X47)  74  44  56  49  53  32  29  63  Pesticides (X48)  1  1  0  1  0  1  0  0  Unspecified chemicals and noxious substances (X49)  15  16  18  21  19  27  16  24  Total  263  237  277  287  252  257  238  324  Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 18 2.2.3. In British Columbia Examining work-related fatalities between 1995 and 2008 in B.C, based on number of claims accepted for fatal benefits by Subsector and the year in which it was accepted, confined space fatalities have been separated out of the data (WorkSafeBC, 2010(a)) - see Table A-5 in Appendix A. Figure 2-4 shows confined-space related fatalities during these years. Despite the high quality of the data, it must be noted that the "four fatalities" of the Sullivan mine are reported under completely different industrial categories – none of which shows that the incident was an accident that occurred simultaneously at a "mine reclamation site". The four fatalities were reported by Sector as (WorkSafeBC, 2010(a)): • A 44-year old paramedic and a 21-year old paramedic died in a sampling shed because of oxygen-deficient air - the Sector was "Provincial Government". • A 50 year old project manager died by oxygen-deficiency in a sampling shed - the Sector was "Canadian Pacific Ltd. (Rail and Mining)". • A 48 year old technical engineer died in a sampling shed by oxygen-deficiency - the Sector was reported as "Technical Services".  This is quite misleading for future investigators in the sense that this does not provide immediate awareness of the potential danger of a confined space at a sulfide waste reclamation site. The problem gives pause for concern about the validity of the separation of data by type of event and industry with respect to "mining industry related", "confined space related" and "mining industry confined space related".  Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 19   Figure 2-4. Confined space related fatalities in the province of British Columbia                      (WorkSafeBC, 2010(a)).  Based on claim counts from Standard Occupational Classification (SOC) by accident type and injury year, between 1997 and 2009, there were 55 fatalities in all sectors due to oxygen- deficiency in B.C. (WorkSafeBC, 2010(b)). This number (only for oxygen-deficiency) is much higher than the total number of fatalities related to all atmospheric-related hazards in B.C which is 30 during this period from WorkSafeBC's past annual reports (WorkSafeBC, 2010(a)) - see Figure 2-4. The cause of oxygen-deficiency is unclear in the SOC claim counts and neither does the database specify if the deficiency occurred in a confined space. Of this number, 23% of the deaths were related to other trades (not specified), with agriculture responsible for 10% and forestry, mining and fishing together responsible for 7%. Unfortunately the percentage of accidents related only to mining is unavailable –Table A-6 in Appendix A. Although the information is unclear, this is the only source of data on atmospheric related fatalities in B.C. 0 1 2 3 4 5 6 7 8 1995 1997 1999 2001 2003 2005 2007 2009 Fa ta lit ie s Year Claims for Atmospheric Related Fatalities in British Columbia Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 20 In B.C, the total number of fatalities by inhalation in a confined space appears to have remained the same in recent years indicating a need to reassess and analyse current procedures to prevent these types of accidents. The detailed reason for each inhalation- related accident is not stored in the WorkSafeBC database (WorksafeBC, 2010(b)), and further contacts with officials at the Agency did not rectify this deficiency. If such records were available, the causes of the accidents should be known for prevention and comparison. 2.2.4. Multiple Fatalities As the statistics on fatalities in confined spaces show, these types of accidents continue to occur at high frequency in many apparently innocuous situations throughout North Amercia. Such tragedies invariably involve multiple deaths as rescuers fail to recognize the danger and succumb to the same hazard as the first victim(s), making multiple fatalities a significant characteristic of confined space accidents. Looking at the NTOF data, Suruda et al., (1994), reported that 72 (12%) of confined space related incidents between 1980 and 1989 were multiple fatalities in the U.S. (585 separate confined space incidents with 670 fatalities) in which most victims died from atmospheric hazards (Suruda et al., 1993).  Table 2-5. Multiple fatalities in confined space accidents by NTOF between 1980 and 1989 (data                     in CDC are public domain and do not need specific permission, Suruda et al., 1994).  Confined Spaces Accidents Fatalities Number Deaths Single 513 513 Double 61 122 Triple 9 27 Quadruple 2 8 Total 585 670  FACE is a project conducted by NIOSH to investigate selected workplace fatalities including confined space related deaths at different sites. The purpose of the study was to understand the effect of different factors on the accident such as management, worker type, Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 21 and working environment. The information from FACE is more beneficial than that from NTOF since it identifies detailed reasons for the accident such as what was missing (if anything) in the confined space entry procedure, i.e., failure to follow procedures or lack of permit system. Of these fatalities, 31 were co-workers and 3 were firefighters (Suruda et al., 1993). Death certificate data by NIOSH identified 11 government agency rescuers who died in confined space accidents between 1980 and 1988. Of the 11 multiple-fatality victims, 3 were identified as police, 2 were firefighters, 1 was a paramedic, and 5 were from other government agencies (Suruda et al., 1993). In 6 of the 62 accidents, SCBA were available for use by the rescuers. In 6 cases, the rescuer died because of inappropriate use of SCBA such as using organic vapour cartridges instead of an air-supplied respirator. All rescuers except for one, who died in an explosion, were overcome by a hazardous atmosphere. Of these 97 victims, 8 were killed by oxygen-deficient air (Suruda et al., 1993). According to OSHA, out of 146 confined space fatalities from 1984 to 1986, 17 (12%) were coworkers or emergency personnel. In 16 cases, the rescue was attempted without SCBA, and in one, the SCBA ran out of air. Two out of 17 rescuers were police officers trying to provide help in a sewage treatment plant (Suruda and Agnew, 1989; Suruda et al., 1993).  According to NTOF it is difficult to recognize the actual rescuers since there were incidents in which the initial victim survived and the rescuers died which may have been reported in NTOF as a single victim fatality. There may also be some accidents in which more than one initial worker was inside the confined space and none of the rescuers died (Suruda et al., 1993). As can be seen the quality of data on multiple fatalities in all of the jurisdictions studied is poor and varies considerably from one census system to another. Unfortunately there is no recent data available on multiple fatalities in confined spaces in U.S and Canada. Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 22 2.3. Confined Space Definitions and Characteristics Failures to identify a hazard as well as incompelete risk assessment are major reasons for confined space accidents (Pettit et al., 1987), but many regulations do not help to identify the presence of such a hazard. Terminology used to define a confined space hazard and describe its characteristics vary considerable and needs to be refined. As pointed out previously, statistics on past confined space accidents are inconsistently recorded in different jurisdictions and industries. It is suggested that the following data elements should be gathered: name and age of victim; location of accident; cause of death; time and date of accident; type of confined space; number of associated victims; industry in which the accident occurred; number of victims who lived; number of rescuers who died. OSHA coined the terms “confined space” and “permit-required confined space” and setup the general requirements to identify and manage confined spaces (Franseen, 1995). While NIOSH, CCOHS, and WorkSafeBC regulations share a common combined form of OSHA’s definitions, differences are apparent. The term "permit-required confined space" refers to a space that meets the definition of a "confined space" AND poses a health or safety hazard. OSHA states that a permit-required confined space has one or more of these characteristics: • Contains or has the potential to contain a hazardous atmosphere; • Contains a material with the potential to engulf someone in the space; • Has an internal configuration that might cause an entrant to be trapped or asphyxiated such as inwardly converging walls or a floor that slopes downward and tapers to a smaller cross section; • Contains any other recognized serious safety or health hazard. A "non-permit confined space" does not contain a hazard or hazardous atmosphere. When change occurs in use or configuration, these spaces should be re-evaluated and may be reclassified as "permit-required" (OSHA, 2004).  If a workplace is "permit required", a written entry permit must be posted and the employer must inform exposed employees of the problem. A written entry permit is a document showing that a hazard assessment has been done. Unfortunately, the definition of each confined space characteristic differs from agency Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 23 to agency and the terminology should be clarified to help understand the sensitivity of each characteristic and how spatial or temporal variations can cause a problem. Defining a "confined space" is important in reducing ambiguity and avoiding misinterpretation. 2.3.1.  Occupational Safety and Health Administration (OSHA) The permit system for confined space entry was established in 1989 by OSHA (Suruda et al, 1993). OSHA regulations attempt to protect workers in confined spaces with an emphasis on identifying and monitoring these spaces as well as training workers (Acer and Bayer, 1990). In 1993, OSHA characterizes a confined space based on size, configuration and use as follows (OSHA, 2004, Pettit and Linn, 1987): "A place…large enough for an employee to enter fully and perform assigned      work, …not designed for continuous occupancy by the employee, and … (with) …   limited or restricted means of entry or exit. These…include underground vaults, tanks, storage bins, pits and diked areas, vessels, silos and other similar areas."  "Limited or restricted means" is defined on the OSHA website (ohsaonline, 2007), but not in the written documents. On the website, "any configuration or other characteristic...that interferes with an entrant's ability to escape or to be rescued in an emergency...can cause limited or restricted means for entry or exit, such as ladders or stairs". A basement with stairs containing pipes or other physical instruments that may hinder escape is classified as a confined space.  Examples of confined spaces vary between agencies. OSHA classifies a trench or excavation greater than 1.52 m deep as a confined space due to cave-in hazards (Suruda et al. 1994), but there is no mention of an atmospheric hazard. This type of structure can limit air circulation and may contain an atmospheric hazard from pore gas that migrates through the trench walls. Bad air can be generated by mineral oxidation or decomposition of organic material in the trench. Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 24 2.3.1.1 Permit–Required Confined Space Characteristics      Choosing the right respirator and venting with proper air is important in dealing with a dangerous confined space. The designation "permit-required" requires signage and people who enter the space must follow safety practices such as using breathing equipment and O2- measuring devices. When a person is in the space, another must remain outside to assist. All rescue personnel–firemen, police, paramedics, and medical staff–must be trained to identify "permit-required" situations even if signage isn't in place. These spaces may have been entered in the past without a problem leading workers to assume no danger exists. They may decide to ignore procedures (see Appendix C).      The term "permit-required" is used in the U.S. to classify a space that must be entered with extreme caution and proper equipment. A confined space must be re-evaluated whenever a change occurs in use, configuration or environment, and if appropriate, reclassified as a "permit-required" space (OSHA, 2004). Unfortunately confined space accidents continue to occur at high frequency and invariably involve multiple deaths as rescuers fail to recognize the danger and succumb to the same hazard as the first victim(s).  OSHA’s definition is prescriptive and does not render as much care in evaluating these spaces as may be necessary. It is suggested that the term “confined space” be substituted for permit-required confined space, and the term “enclosed structure” be used instead of the many differing terms for these types of spaces. The term “confined space” would mean a confined or enclosed structure with a physical, chemical, radiation, inhalation, and/or poisoning hazard in which a permit to enter is always required. An "enclosed structure" on the other hand, does not require a restricted entry permit unless a risk assessment leads to the need to convert it to a "confined space". According to this definition, a room with a closed door and window is an enclosed structure (potentially-safe) while a trench or ditch is an example of a potentially-unsafe enclosed structure (confined space) in which cave-in hazards Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 25 exist or in which toxic, flammable, explosive, or O2-deficient gas may migrate through soil and become confined therein.  2.3.2.  National Institute for Occupational Safety and Health (NIOSH) NIOSH defines a confined space as having one of these elements (Pettit and Linn, 1987):  1- Limited entry and exit: openings limited by size or location; usually small, < 45 cm; difficult to move easily or take-in equipment such as life-saving equipment (respirators); This feature may not be present in all confined spaces - in some open-top spaces such as pits, degreasers, excavations, and ship-holds, the opening may be very large. Access may require ladders, hoists, etc. - escape from such places may be difficult in an emergency. 2- Unfavorable natural ventilation: the atmosphere inside may become different than normal due to lack of free air movement. Chemicals may be processed or decomposing, organic substances are present, hazardous gasses can accumulate, or the air may become O2-deficient (or O2-rich increasing risk of fire or explosion if a source of ignition is present). 3- Not designed for continued worker occupancy: Confined spaces may be designed to store product, enclose materials and processes, or transport products or substances. They are not designed for workers to enter and work. As a result, worker entry into a confined space for inspection, maintenance, repair, cleanup, etc. is often difficult and dangerous. 2.3.3. Definition of a Confined Space by CCOHS and WorkSafeBC A confined space is an enclosed or partially enclosed structure not designed or intended for human occupancy. It has restricted entrance/exit in terms of location, size or means. Confined spaces can become dangerous due to: design, construction, location, atmosphere, materials in the space, work activities, mechanical processes or presence of a hazard (CCOHS, 2002). Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 26 These characteristics are ambiguous. A confined space may not be small and can be located below or above ground. Examples include vats, hoppers, utility vaults, tanks, sewers, pipes, access shafts, truck or rail tank cars, manhole, tunnel, wells, cold storage, silos, ships' holds, sub-cellars, culverts, open ditch, aircraft wings. Ditches and trenches may also be confined spaces when access or egress is limited. Hazards may include (CCOHS, 2002):  "poor air quality: poisonous gas or lack of oxygen, chemical exposures due to skin contact, ingestion or inhalation of 'bad' air, Fire Hazard: due to flammable liquids and gasses and combustible dusts, Process-related hazards such as residual chemicals, release of contents of a supply line, noise, moving parts of equipment, structural hazards, entanglement, slips, falls, radiation, temperature extremes including atmospheric and surface, shifting or collapse of bulk material, barrier failure resulting in a flood or release of free-flowing solid, uncontrolled energy including electrical shock, visibility problems and biological hazards." Because of the nature of the hazard and work, air testing should be ongoing: “Conditions can change while workers are inside the confined space and sometimes, a hazardous atmosphere is created by the work activities in the confined space".  2.3.4. WorkSafeBC Part 9 of the Safety at Work Regulations in WorkSafeBC deals with confined spaces and is based on OHS regulations. The document contains a brief definition of "confined space":  “(it) is closed or partially enclosed, is not designed or intended for continuous human occupancy, has limited or restricted means of entry and exit that may complicate provision of first aid, evacuation, rescue and other emergency responses, and is large enough and…configured (so) worker(s) can enter to perform assigned work”.   The regulation does not provide a description of each property and no example of a confined space is given (OHS Regulation, 2008). However, all the characteristics of confined spaces based on the OHS definition as well as many examples are present in the document. Confined spaces are not divided between "permit required" or "non-permit required". The Reference Manual defines a confined space as a place with all of the following characteristics (Confined Space Entry Program - A Reference Manual, 2005, by permission): Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 27 1- Enclosed or partially enclosed - no need to be small, tight, or fully enclosed; may be large or small and partially enclosed; workers may be able to move freely in the space.  2- Not designed or intended for continuous human occupancy - entry for purposes such as inspection, maintenance, repair, or construction; space not normally ventilated; entered on a regular basis such as for sampling.  3- Limited or restricted means of entry or exit - difficult to evacuate or provide emergency response services; entry point difficult to walk through; access by ladder or stairway with poor access due to “restrictive slope, narrow width, or extreme length”; physical obstructions may make exit difficult. 4- Large enough in such a way that a worker can enter to perform the assigned work; a space that is too small for a worker to enter is not a confined space.   The manual states that a place containing a toxic atmosphere is a confined space, even when entry and exit are unconstrained. A confined space permit can be updated by the supervisor, standby person, or tester. The standby person may alter the permit to update the list of workers inside. The tester may alter the permit to record test results. The entry supervisor who signed the permit may update it if there is a change in the work crew after each shift change, or if another supervisor takes over. In the case of significant change that may affect safe work procedures, only a qualified person can change the work procedures. The entry supervisor changes the permit according to changes made by the qualified person (Confined Space Entry Program - A Reference Manual, 2005, by permission). 2.3.5. B.C. Mines Act Section 3.4.1 of the Mines Act BC regulations deals with Hazardous Atmosphere and Confined Space Work. The regulation begins as follows:  "The manager shall ensure…written procedures are developed and implemented for work in confined spaces where irrespirable, toxic or flammable atmospheres might be encountered."  In this regulation, a confined space is defined as: "An area, other than an underground mine, that includes all of the following: (a) is enclosed or partially enclosed; (b) is not designed or intended for continuous human occupancy; (c) has limited or restricted means for entry or exit; (d) is large enough and so configured that a worker could enter to perform first aid, institute evacuation, rescue or other emergency response service."  Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 28 This regulation does not help an individual recognize a hazard since there is no definition of each characteristic and no examples are provided. There is no mention of the presence of hazard in the confined space as well. The regulation makes no mention of reclamation work. Section 3.4.2  describes safe work procedures with the necessity for atmospheric testing, ventilation, and use of protective equipment in confined spaces being described in Sections 3.4.3, 3.4.5 and 3.4.6 respectively (Mines Act, 2008). The BC Mines Act, 2008 offers rules on reclamation practices in Section 9.13.1, but nothing on confined spaces. Sec. 10.7.11 refers to waste dumps, but the only risks mentioned for control are slope-stability and erosion, although dangerous gasses such as methane, and dusts such as asbestos, silica, and lead have specific rules. General regulations for waste management and for confined spaces do not help identify a hazard at a reclamation site indicating a need for a consistent approach to identify confined space dangers during reclamation. It would be useful if the definition of a permit-required confined space was added into the Mines Act. According to the Mines Act BC, a confined space is:  "a tank, process vessel, underground vault, tunnel, or other enclosure not designed or intended for human occupancy. A person...enters...only if there    (is) work to be done".   The Manitoba mining regulations relate to work in confined spaces and provide examples:  "No worker shall enter and no employer shall cause or permit a worker to enter a tank, pit, sump or other confined space until proper stated confined space procedures are put in place (Operation of Mines Regulation in MB, 1994)".   If these examples were in the BC Mines Act, perhaps identification of the sump at the Sullivan mine as a "permit-required" confined space might have been recognized. The mine plan and reclamation program part of the BC Mines Act deals with acid rock drainage, but does not mention O2-depletion hazards from air and water coming from a sulfide waste dump. Section 10.1.9 states that a plan to predict, prevent, mitigate and manage ARD should match Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 29 the Guidelines for Metal Leaching and ARD at mine sites in British Columbia, but this guideline also does not mention atmospheric hazards.   2.4. Shortcomings of Confined Space Regulations The definition of a confined space, possible characteristics, and examples are incomplete in most mine regulations examined. OSHA’s definition of confined spaces talks about size, configuration and use and then evaluates the space as “permit-required” or “non-permit required” based on atmospheric hazard or physical harm potential. NIOSH places emphasis on "limited entry and exit" and refers to the small size and diameter of the entrance. WorkSafeBC defines a confined space as a large or small space that may not be enclosed on all sides. WorkSafeBC mentions the fact that pits, excavations, and other types of confined spaces near a contaminated site or waste pile containing hazardous material can collect toxic gas depending on the type of waste (Hazards of Confined Spaces, 2004):   "A confined space may be located next to a source of a hazardous contaminant. The contaminant (can) enter the confined space through porous walls, such as those ...found in sewers or trenches, or through difficult-to-seal openings such as conduits"   Some confined space protocols mention the impact of sudden atmospheric changes that can affect the influx of toxic gasses into an enclosed structure. NIOSH discusses the danger of a manhole located within a swampy area (Michaelsen and Park, 1954; Pettit, 1994) in which a sudden drop in barometric pressure caused methane to diffuse into the manhole through its walls. At other times, this space was entered without problem. In section 3.4.3 of the B.C. Mines Act as well as section 9.9 of the WorkSafeBC regulations, testing for danger and the conduct of a hazard assessment are discussed. However, no mention is made of changes in pressure and temperature on the conditions in an enclosed structure (BC Mines Act, 2008): Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 30 "Tests of the atmosphere inside the confined space shall be made at intervals during the work progress to ensure that the quality of air does not deteriorate ...test results shall be recorded as required by section 3.4.3."  Mention of the role of atmospheric changes should also be included in this section of the Act. Since the WorkSafeBC Confined Space Manual contains more details to help with recognition, it is suggested that the BC Mines Act should require mining personnel responsible for waste dump reclamation activities to apply this manual in their work. There is a need for a consistent approach which clearly defines a confined space and contains all information necessary to recognize a potential atmospheric hazard. In this research, a manual called the Atmospheric Confined Space Manual (especially for gasses from mine wastes and soils) has been written to help with such hazard recognition – Appendix C. 2.5. Summary  This chapter has examined and analysed data on confined space accidents and fatalities in North America. While a decline in such events has occurred in the U.S. between 1982 and 1995, there are still significant numbers of confined space accidents occurring each year throughout North America. A consistent method to record such accidents is required to provide accurate and valid data on these situations. Definitions and descriptions of confined spaces differ significantly among the various North American regulatory agencies. To provide clarity, simplified terminology is proposed to rectify the situation. Any "enclosed structure" may become a "confined space" containing a hazard. Only these two terms are necessary. All "confined spaces" then require a permit and can only be entered using the listed procedures and safety equipment. It is very important to recognize that a hazard exists (or may exist) within an enclosed structure and then, implement proper procedures, permits, and signage to ensure unsafe exposure is eliminated. Current B.C. Mining regulations do not address atmospheric issues Chapter 2- Confined Spaces Definition, Fatalities and Regulations Around the Globe 31 with waste dumps undergoing reclamation. There is insufficient definition and description of confined spaces in the B.C. Mines Act. With reclamation sites, climatic temperature and pressure changes control gas emission into an enclosed area at different times of the day  (Mohammadi and Meech, 2011). As such, a confined space can switch from being safe to being unsafe. A confined space may be measured as safe when at other times it is unsafe. The regulations should warn about this transformation (Mohammadi and Meech, 2011). General confined space regulations may be insufficient for confined spaces in different industries. As a result, confined space regulations should be customized for each industry and perhaps a specific risk assessment tool for each type of workplace should accompany the regulations. Details of previous confined space accidents should be included within existing regulations to help prevent reoccurrence of an accident.     1  Many Parts of this Chapter have been published previously. Mohammadi, L., Meech, J.A., 2008, Implementing Atmospheric Risk Assessment in Mine Reclamation, 23rd International Conference on Solid Waste Technology and Management, Philadelphia, PA, USA, pp.12. 32 Chapter 3  3. Sullivan Mine Accident: Similarities and Differences with Other Confined Space Accidents1  3.1. Overview There are many examples of waste management techniques aiming to minimize threats to the environment that have been developed over the years based on experience. Unfortunately, adverse effects of some of these practices create occupational health hazards that are not yet fully-understood or appreciated. With these systems, consideration of a new requirement may be necessary to ensure a safe working environment. Mine waste dumps consist of wastes managed by reclamation policies generally applied after mine closure. The activity is isolated from normal mining operations with the primary goal to make the site as compatible as possible with its surrounding environment (Mohammadi and Meech, 2008). The Sullivan mine accident happened at a reclamation site where the environment was being changed to accommodate innovative designs that had become state-of-the-art in reclaiming waste dumps over the past decade with best practices being applied to  minimize threats to the surrounding environment and ecosystems in a safe and economic way. These waste management techniques unfortunately brought a new and initially unrecognized condition to the list of confined space fatalities (Mohammadi and Meech, 2008). Although the accident was a unique occurrence never before reported for sampling practices at any other mine site, the circumstances parallel those of virtually all atmospheric-related confined space accidents (Mohammadi and Meech, 2008 and 2011). These include, but are not limited to: - The hazard is unrecognized by operators and by the victims; Chapter 3 – Sullivan Mine Accident  33 - The hazard (in this case oxygen-depletion) has no associated odour or colour; - Death is very quick (seconds to minutes); - Multiple deaths take place as rescuers die in a futile attempt to save the first victim(s); - After the fact, the danger is obvious. This suggests that the accident might have been predicted (and prevented) if an a priori atmospheric risk assessment had been performed (Mohammadi and Meech, 2008). If a long- term atmospheric risk assessment process had been required by the B.C Mines Act as the reclamation and revegetation work ensued, it is possible that contractors and mine employees at the site might have recognized that the sampling station had become what is known as a "permit-required confined space". Then the current standards and regulations that apply to confined spaces in the BC Occupational Health and Safety (OHS) and/or in the BC Mines Act could have been applied to prevent this tragedy. This chapter describes the accident in detail and compares it with other similar confined space accidents. 3.2. Sullivan Mine Tragedy The Sullivan Mine was one of the largest lead-zinc mines in the world when it closed in 2001 after 92 years of operation. About 500 people were employed at the mine that year. The No.1 Shaft Waste Dump was created during the 1940s and continued to operate right up to closure – see Figure 3-1. The height from the flat top of the dump to the toe is about 55m. The dump contains about 2.9M tonnes of sulfide rock. The estimated total volume is 1M m3 with about 30% voids. Other mine waste such as domestic garbage, steel, plastic, wood, residual Shotcrete, glacial till, and other debris was also placed in the dump from time to time. After closure, a skeleton-crew remained to manage reclamation activities aimed at restoring the site to a form compatible with the local environment. This work began prior to Chapter 3 – Sullivan Mine Accident  34 closure and is still on-going today. In 1995, a V-notch weir was installed about 100m from the toe to sample effluent and measure flowrate. In 1997, to overcome winter ice build-up on the weir, concrete blocks were placed around it and a shed erected over it – see Figure 3-2.  In 2004, the open ditch was partially covered as the toe was extended forward about 70m to reduce the dump profile in preparation for revegetation and erosion control. In 2005, 1 m of glacial till was placed over the dump surface and the ditch.    Figure 3-1. No. 1 Shaft Waste Dump at the Sullivan Mine prior to regrading and covering                      with glacial till (Reports of the Technical Advisory Committee of the Sullivan                      Mine accident, 2007, by permission Teck Metals Ltd).  Chapter 3 – Sullivan Mine Accident  35   Figure 3-2. Seepage collection ditch in September 2005 prior to covering with glacial till                       (Phillip et al., 2008, by permission).  The accident at the Sullivan Mine occurred when oxygen-depleted air (and water) caused by sulfide oxidation flowed out of the Number One Shaft waste dump through an underground drainage channel and buried pipe that connected directly to a sampling shed - Figure 3-1. Seasonal temperature changes have been found to cause air to flow from the dump into the shed (Phillip et al., 2008). Before the ditch was covered, there was no direct connection between air in the dump and air in the shed. The ditch was open to the air allowing effluent to become re-oxygenated as it flowed along the channel. So, water entering the shed was not oxygen-depleted prior to the ditch being covered – see Figure 3-2. After covering the ditch as part of regrading activities, a well-sealed, underground drain was created, isolated from the atmosphere, creating an unrecognized hydraulic conduit for air and gasses to flow between the dump and the shed. If an atmospheric risk assessment had been done as the work proceeded, contractors and mine employees might have realized that the shed had become a "permit-required" confined space. Then the confined space Chapter 3 – Sullivan Mine Accident  36 regulations in the B.C. Occupational Health and Safety Regulations or the BC Mines Act could have been applied to prevent the accident. The entire dump was also covered with glacial till. This seal reduces water percolation and, when saturated with water, restricts air infiltration slowing oxidation as the pore air remains depleted of oxygen - see Figure 3-3. Appendix D describes the hazardous gasses that accompany ARD from Sulfide or Coal mine sites. The drainage ditch running along the toe was engineered into a drain and then covered by the toe extension to prevent seepage. The sump collected the effluent and diverted it through a buried pipeline to a water-treatment facility. Monthly sampling was done to monitor flow rate and contaminant levels. The sampling shed at the Number One Shaft waste dump was installed to sample Acid-Rock- Drainage seepage from the waste dump. The sampling shed was in use in the fall of 2005 and winter of 2006 on a regular basis up to one week before the tragedy occurred without any incident or indication of a problem (see Figure 3-2).  Figure 3-3 Sullivan Mine No. 1 Shaft Waste Dump after sealing with glacial till                      (Phillip et al., 2008, by permission).  Chapter 3 – Sullivan Mine Accident  37  On May 15 and 17, 2006, four people – a consultant, a mine employee, and two ambulance paramedics – died in the sampling shed. They each lost consciousness and fell into the sump because of the lack of oxygen. The accident is described in detail in the following to highlight the errors in judgement made by rescuers that cost them their lives.  The first victim was Doug Erickson, an environmental consultant hired by Teck-Cominco, the owners of the site. He entered the shed in the late afternoon on May 15th to take a water sample when it is assumed he was overcome. He was not reported missing until the morning of May 17th upon which Bob Newcombe, a mine employee, went to see if Erickson was at the sampling shed. He observed Erickson lying in the sump face-down in the flowing water. Newcombe apparently thought that Erickson had drowned as that is how he reported the situation to the 911-operator when he called on his mobile phone requesting emergency help. He then called mine consultant, Dave van Dierden, and asked him to come to the site gate to admit the ambulance. Despite identifying the shed as a "confined area" as he described it to the 911-operator, he entered it to try to help Erickson, but was overcome. van Dierden arrived at about the same time as the paramedics came on the scene. Kim Weitzel, the senior paramedic, who thought she was responding to a possible drowning accident, entered the shed and began to climb down the ladder into the sump. She remarked that there were two bodies, not one, and apparently realized that an atmospheric danger might be present. She began to ask van Dierden if a "gas" problem existed, but her words died away as she fell unconscious and collapsed into the sump. Her colleague, Shawn Currier, ran to the shed when informed by van Dierden that Weitzel had fallen. Before van Dierden could stop him, Currier entered the shed and also collapsed. van Dierden then called 911 to report these new difficulties and stated that H2S gas might be present. Shortly thereafter, the fire department arrived with appropriate breathing apparatus. They brought out all four people and delivered them to the hospital, but it was too late. Chapter 3 – Sullivan Mine Accident  38 Following a preliminary investigation, The Chief Inspector of Mines for B.C. issued an immediate warning to all mines about possible similar circumstances and ordered other mine effluent sampling sheds to be treated as confined spaces. In October 2006, a report issued by the Chief Inspector identified the accident as being "unprecedented in the history of mining". According to the report "the process that led to the oxygen-depleted atmosphere has not...occurred anywhere else in the world." 3.3.  Controlling Factors 3.3.1. Direct Factors During the summer of 2005, the dump was covered with 1m of glacial till and the slope re- contoured to enhance growth of suitable plant material and reduce erosion. In this way the drainage ditch became covered. A 12 m long, 400 mm diameter pipe directed acidic water from the ditch to the shed – see Figure 3-2.  This change in the effluent collection and monitoring system isolated oxygen-depleted effluent and pore gasses from the atmosphere. Covering the ditch meant air in the shed became directly connected to "bad" air in the dump.  In August 2006, the dump was instrumented under the oversight of a Technical Advisory Panel set-up by Teck Corporation and the B.C. Ministry of Energy, Mines, and Petroleum Resources to monitor respiration of the dump. Data collected included air velocity, temperature, pressure, and gas composition in the pipe, at the end of the pipe, and about waist height in the shed. Site meteorology, cover moisture content, internal temperature, gas composition, and pressure at 16 locations were also monitored.  Samples taken immediately following the accident showed oxygen levels at the bottom of the sump of about 2%, while carbon dioxide was about 7% - dangerous limits for these gasses are given in Table C-7 in Appendix C. Based on data from the Cranbrook airport, the May 13-17, 2006 period includes a sharp increase in maximum daily temperature to about 20°C with a strong decrease in barometric pressure. On May 8, 2006 when the shed was Chapter 3 – Sullivan Mine Accident  39 entered safely, the Cranbrook airport data indicated a rising barometric pressure and a temperature below 11 oC. The cumulative airflow volume measure during the 2006-2007 winter was about 1.2 M m3 – 4 times the estimated void space in the dump.  At different depths around the dump in June 2007, oxygen concentrations of air in the dump ranged from normal (about 21%) to near zero. Carbon dioxide concentrations ranged from near zero to about 5% in most locations, but were measured as high as 21% at one location. The instruments showed that internal temperatures ranged from 5 to 16 oC indicative of sulfide reactions while outside air temperatures ranged from -7 to +26 oC. The gas velocity measured at a point 5 cm inside the exit of the effluent pipe has been reported to be well- correlated with the outside temperature measured at a meteorology station located part way up the slope of the dump (Phillip et al., 2008). As can be seen in Figure 3-4, for temperatures below about 11.5 oC, the direction of flow is into the dump (dump is "inhaling"), and when the temperature rises above about 11.5 oC, the dump begins to "exhale" resulting in oxygen- depleted air flowing through the pipe along with the effluent water into the bottom of the sump (Phillip et al., 2008). Because of this behavior, this dump can be called "a breathing waste dump". The value of 11.5 oC is close to the average internal temperature at the center of dump. Seasonal variations cause the outside temperature to rise above 11.5 oC in summer and to drop below this value in winter. Figure 3-5 shows a time series plot of gas velocity and outside temperature from March to December 2007. The dump exhales during the summer while during the winter it inhales. During the spring and fall, both exhaling and inhaling can occur during some days in the month or during the day or night depending on the outside temperature. Temperatures at a level above the center of the dump are high all year long at about 16°C.  Chapter 3 – Sullivan Mine Accident  40   Figure 3-4. Corelation between gas velocity and outside temperature. (Phillip et al., 2008,                      adapted by permission).                      Equation: y = 2E-06x4 - 2E-05x3 - 0.0018x2 - 0.026x + 0.5332, R2=0.927.   Figure 3-5. Time series of air velocity and atmospheric temperature                      (Phillip et al., 2008, adapted by permission).  Movement of gas into the shed is not diffusion-controlled. Temperature differences within the dump and outside cause air movement by convection – see Figure 3-6. During the winter, gas within the waste dump is warmer than outside (lighter) and rises from the surface of the dump. This process causes the dump to inhale air from its sides and bottom (i.e., from the Exhales Inhales R² = 0.927 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 -18 -8 2 12 22 32 G as  Ve lo c ity  (m /s ) Outside Temperature (0C)Chapter 3 – Sullivan Mine Accident  41 shed). Although the atmosphere of the shed is safe at this point, oxygen-depleted water could still cause a dangerous atmosphere – see Section 3.3.2. During the summer, when outside air is warmer (and lighter), the flow reverses itself. This is when the dump exhales leading to a dangerous atmosphere within the shed.    Figure 3-6. Schematic diagram of the sampling shed at the Number One Shaft Waste Dump.  Figure 3-7 shows gas composition measurements at 2.4 m from the pipe discharge to the ARD collection sump. Gas composition changes with flow direction so when the gas velocity is negative for a short period, the oxygen concentration in the sump falls below 21%. Taking into account the flow of oxygen-depleted air from the pipe into the shed (from 11th to 13th of May 2007), the time for the oxygen level in the sump to become depleted can be as short as ten hours for air being displaced by toxic pore gas and about 2 days for oxygen-depleted water flowing at the typical rate observed at the Number One Shaft waste dump assuming no air leakage through the shed walls and doors (see Section 3.3.2).  Chapter 3 – Sullivan Mine Accident  42    Figure 3-7. Gas velocity, oxygen and carbon dioxide concentrations measured 2.43 m along                      the ARD pipe from the sump (data collected by the Technical Panel                      on the Sullivan Mine accident, Teck Metals Ltd).  Sampling sheds like that used at the Sullivan mine exist at many sites around the world and workers are likely using them without knowledge of this danger. Surface anomalies on top of the dump may also concentrate "bad" air creating a danger during colder months. Dangerous gasses can accumulate in a confined space (Mohammadi and Meech, 2011) such as a sampling shed or a topographical feature such as a hollow or erosion channel. Knowledge about similar atmospheric-related confined-space accidents can contribute to developing the risk assessment tool. 3.3.2. Indirect Factors - Role of Oxygen-Depleted Water in Consuming Oxygen  Air flowing out of the dump is not the only potential hazard-causing emission. Water flowing in the pipe is also oxygen-depleted and can remove oxygen from the air confined in the shed above the sump. Prior to 2005, the open ditch brought drainage from the dump to the shed with the water flowing into the bottom of the sump over a weir (Figure 3-8).   -5 0 5 10 15 20 25 -2 -1 -1 0 1 1 2 Oct-06 Nov-06 Jan-07 Mar-07 Apr-07 Jun-07 Aug-07 Sep-07 Nov-07 Jan-08 O 2 an d CO 2 (% ) Ga s V el oc ity  Time Gas Velocity Oxygen concentration Carbon Dioxide ConcentrationChapter 3   Figure 3-8. The sump at the bottom of                     the ladder (Courtesy o  Because of oxidation of sulfide oxygen-depleted. Covering the drainage ditch resulted in the isolation of effluent from the atmosphere. What previously was oxygenated oxygenated groundwater creating an unrecognized and dangerous situation. depleted water can cause air within the sump and shed to become according to Henry’s Law. Before covering the ditch, this water w atmosphere as it flowed between the dump and the shed, and the dissolved oxygen content could return to equilibrium (or close to it) with air at normal (equilibrium) and physics (kinetics water flowing through backfill material in an underground mine (Bayah et al., 1984). The equilibrium of oxygen in water as a function of oxygen content in the air and temperature is shown in Figure 3 different temperatures and four different oxygen p  – Sullivan Mine Accident  the sampling shed. Note the weir to the bottom f WorkSafeBC).  minerals within the dump, water coming out of the dump is  surface water now was de as exposed to the oxygen-levels. The chemistry ) of these reactions have been described previously for -9. Saturated dissolved oxygen levels ercentages in the atmosphere have 43  -right of  oxygen-depleted - This oxygen- oxygen-depleted  in water at four  been Chapter 3 – Sullivan Mine Accident  44 calculated. If dissolved oxygen in the water is below its saturation value, oxygen transfers from the air into the water (Bayah et al, 1984) according to the following equilibrium reactions:  Ct* = KH Pgt  (3-1) KH = 9.73 x 10-4 exp(1799/T) (3-2)   Pgt = Xgt Pa  (3-3)         where:  Ct*   = Saturated Dissolved oxygen at time t (mg/L)  Pgt  = Partial Pressure of oxygen in the gas phase (KPa) KH    = Henry’s Law Constant (mg/L-KPa) Xgt  = oxygen in the air at time t (%) Pa   = Atmospheric Pressure about 102 KPa T    = Atmospheric Temperature oC        Figure 3-9. Oxygen equilibrium in water as a function of oxygen content in air                     (Created by Furmals in Bayah et al, 1984).        The actual removal of oxygen by oxygen-depleted water is independent of temperature since the removal rate increases with temperature while the driving force of the saturated dissolved-oxygen level decreases with temperature. These two effects cancel each other out over the temperature range of interest leading to a neutral response of the rate of oxygen removal by water to changes in temperature (Bayah et al., 1984). The extent of oxygen- 0 2 4 6 8 10 12 14 16 0 3 6 9 12 15 18 21 O xy ge n  Co n ce n tra tio n  in  W at er  (m g/ L) Oxygen Concentration in Air (%) Equilibrium of Oxygen in Water as a Function of Oxygen Content  and Air Temperature T=5    C T=10  C T=15  C T=20  C o   o   o    o  Chapter 3 – Sullivan Mine Accident  45 depletion of air in the sump depends on the effluent flowrate, its level of oxygen, and the leakage rate of air into the shed from outside (Bayah et al., 1984). If no leakage of air into the shed is assumed, and water has only 1 mgL-1 oxygen and its flow rate is 33 Lmin-1, for the initial oxygen content of sump air of 20.9%, within a period of about 46 hours, oxygen- depleted water will reduce the oxygen content of air in the sump to about 18% (Mohammadi and Meech, 2011), as shown in Figure 3-10. Since the water has not reached the equilibrium value of C* (about 8 mg·L-1 for this concentration of oxygen in air), the dissolution reaction continues. Over time, the Ct (dissolved oxygen in water) will drop to about 0 mgL-1. At this point, the value of Ct* also drops to an amount very close to the value of Ct, and so, the reaction stops. The formulas used to prepare this graph are shown in Equations 3-4 and 3-5 (details in Appendix B.1). Assumptions about conditions and properties of the sump are as follow: Pressure             =  102 kPa Temperature          =   20 oC(293 K) Shed Volume          =   21,600 L Sump Water Volume    =   1,200 L Sump Air Volume      =   4,800 L Sump Interfacial Area        =   2.4 m2 Sump Depth          =   2.5 m  Water Flow Rate       =   33.3 L·min-1 Air Leakage           =   0.0 L·min-1 KL(293K)             =   32.3 cm·hr-1                   g3031g3004g3031g3047 g3404 g1837g3013 g3002g3023 g4666g1829 g1499  g3047 - g1829g3047)  g1837g3013 g3404  g1837g3013 g4666g2870g2877g2871g4667 g46661.024g4667g3021g2879g2870g2877g2871 (3-4)   (3-5)  where:  KL      = Mass Transfer Coefficient A/V  = Interfacial Area of Water / Volume of Water (m-1) Ct       = Dissolved Oxygen in Water at Time t (mg·L-1) Ct*       = Saturated Dissolved Oxygen at Time t (mg·L-1)   Chapter 3 – Sullivan Mine Accident  46   Figure 3-10. Oxygen concentration of sump air in contact with oxygen-depleted water at different                      times (no air leakage into the shed assumed).  Gas blowing out of the pipe will mix with sump air and decrease the oxygen content further. The concentration of contaminated gas in the space increases relative to the total gas in the sump. For this to happen the total volume must increase or some of the gas in the sump must exit through leakage points. In the case where gas enters the sump through the pipe, as it mixes with the air it will push air out of the sump – see Appendix B.2. When gas is flowing from the pipe, this process will cause oxygen-depletion in the sump, unless fresh air leaks into the structure through gaps around the door and holes in the walls. The combined effect of oxygen-depleted water and air coming from the pipe into the sump for the period between 11-May-07 4:00 to 13-May-07 18 (approximate time of accident in 2006) is shown in Figure 3-11. In this graph, whenever gas flows into the sump, the oxygen content in the sump air was updated to the value of the measured oxygen in the pipe. Otherwise, the oxygen level 0 2 4 6 8 10 12 14 16 18 20 22 24 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360O 2 Co n ce n tr at io n  (m g/ L o r % ) Time (Hours) O2- deficient Water Effect on O2- Depletion in Sump Xgt Ct* Ct Ct      = Dissolved O2 in Water in the Sump at Time t (mg·L-1) Ct*  = Saturated Dissolved O2 Concentration atTime t (mg·L-1) Xgt  = Oxygen in Sump Air at time t (%) Problem Very Dangerous After 120 hrs = 14% O2 After 46 hrs = 18 % O2Chapter 3 – Sullivan Mine Accident  47 of the sump in each time interval is a mixture of the oxygen content of the pipe with oxygen content of the sump during the previous time interval of the calculation.  Given the gas velocity and pipe cross sectional area, the gas flow was back-calculated from the measured oxygen level. Gas velocity variations during this period due to changes in outside temperature is shown on an hourly basis in Figure 3-12. Gas velocity is positive from 11 am to 2 am and negative from 2 am to 11 am. Accounting for the flow of oxygen-depleted air from the pipe into the sump, the time for the sump's oxygen level to drop below 18% is less than 10 hours. This shows that the effect of "bad" air blowing into the sump and up into the shed is far more significant than oxygen consumption by the ARD effluent. Nevertheless, oxygen consumption by oxygen-depleted water can be important in other circumstances and should be taken into account especially in situations where the amount of water flow is high relative to the sump volume. In Figure 3-11, during the first 8 hours, gas is not flowing from the pipe. At this point, oxygen is being removed from the sump air only by oxygen-depleted water and as a result the Xgt in the graph shows a slight drop of 0.55% of oxygen. The actual measured oxygen content in the sump at this point remained high at 20.6 % since fresh air can leak in to the shed and sump. This indicates a leakage rate of about 0.19 L/s to compensate for the oxygen-depletion caused by the water. The leakage was enough to maintain the oxygen above 20% and so even a small amount of leakage is enough to compensate for oxygen- depletion by water during the first 8 hours. From the 8 hour to 27 hour time period, the gas velocity becomes negative and gas was blowing out of the pipe in combination with oxygen- depleted water which caused the oxygen level to decline significantly. After this point, for a period of 5 hours, the gas velocity was zero, and fresh air leaking into the sump caused the oxygen to increase back to 20.6%. If air was not leaking into the sump, the oxygen content can not increase by this amount. In the middle of the graph, the oxygen concentration has risen from 2.4 to 10.3%. This increase needs an air leakage flow of 2.7 L/s. In the next hour, Chapter 3 – Sullivan Mine Accident  48 the O2 concentration increased from 10.3 % to 20.0% which corresponds to an air leakage flow of 3.5 L/s. Oxygen removal by water over this two hour period is 0.23%. From hour 28 to hour 29, the gas velocity is zero, yet the oxygen content decreased from 20.5% to 19%. This shows that gas flow from the dump into the pipe has not completely stopped but rather appears to be seeping into the shed at a very low velocity. Air leakage is calculated by the following equation: Volume of Leakage g4666L/hg4667g3404 g3436Xgt‐Xgg4666t‐1g4667XOutsideΔtg3440 .Vsump and shed  (3-6) where: Xgt = Oxygen level in the sump air at time t Xg(t-1) = Oxygen level in the sump air at time t-1 Xoutside = Oxygen content in the air outside the shed VShed = Volume of the sump and shed (26400 L) ∆t = time interval between t-1 and t (one hour)  The air in the shed mixes with air in the sump as the oxygen level in the sump air falls. Calculating the oxygen content in the shed after mixing with gas in the sump for each hour, it was concluded that the shed’s oxygen level is very close to the sump’s oxygen level unless the door is opened. Knowing this, the amount of leakage was calculated for the volume of the sump plus the volume of the shed. In the Sullivan mine accident, workers did not show any symptoms as they opened the door and entered the shed but as each descended the ladder into the sump they collapsed. This took place because when the door is opened, fresh air quickly ventilates the shed making the air in the shed safe to breath. The air in the sump may take a considerably longer time to be ventilated in this way due to the sump configuration. This phenomenon is true for other enclosed structures such as manholes located in an open atmosphere. For such configurations to be ventilated, forced ventilation is needed.     Chapter 3 – Sullivan Mine Accident  49   Figure 3-11. Oxygen-depletion of air in the sump by oxygen-deficient water and air (velocity and                        oxygen content data were provided by technical panel of the Sullivan mine, Teck                         Metals Ltd).    Figure 3-12. Gas velocity and temperature variation approximately one year after accident.                       (time matched to same period as in Figure 3-11)  Although the mechanism of oxygen removal by water is low in the case of the Sullivan site, it is possible that oxygen-depleted water flow rates and shed volumes at other sites could be much more significant (Bayah et al., 1982). The time to consume oxygen down to 18% by deoxygenated water at different flow rates for the shed volume and sump surface area at the Sullivan mine (A:V = 2 m-1) and considering different leakage rates of fresh air into the shed is -0.6 -0.3 0.0 0.3 0.6 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 A ir Ve lo ci ty  (m /s ) O u ts id e Te m pe ra tu re  (o C ) 60 hours Outside Temperature Air Velocity 4 am -11 am 11 am - 2 am 2 am -11 am 11 am -2 am 2 am -11 am 11 am -2 am 11-May-07 4:00 to 13-May-07 18:00 First 8 hrs water removes 0.55 % O2 Leakage rate ~ 0.19 L/s Average leakage rate ~ 2.7 – 3.5 L/s In 14hrs water removes 0.93 %  O2 Leakage rate ~ 0.32 L/s Measured O2  Xgt  = O2 in Sump Air at time t (%)   Ct*  = Saturated Dissolved O2 Concentration at Time t (mg/L)   C0t  = Dissolved O2 in Water in the Sump at Time t (mg/L) Chapter 3 – Sullivan Mine Accident  50 shown in Figure 3-13. In this graph, safe and dangerous regions are separated by a line, each line representing a different air leakage volume as represented by the velocity of air flowing in the pipe. Results are presented in Figure 3-14 for a sampling shed with an A:V ratio equal to 4 m-1. With an air velocity in the pipe of 2 m/s, even small water flows can reduce the oxygen level to 18% in about 2 days. Higher water flows will create even more hazardous conditions. Therefore, the potential for oxygen removal from air by flowing water through a confined space is a significant factor in conducting a proper risk assessment.  Safe and unsafe situations at the Sullivan mine based on different air velocities and water flow rates are depicted in Figure 3-15. The range of operating conditions that occurred at the Sullivan mine are represented by the elliptical region in the graph. For water flow rates below 25 L·min-1 the sump is safe for all but very low air leakage flows. However, for a water flow rate of 77 L·min-1 safe conditions are only achieved if the flow rate of fresh air into the sump is above 0.8 m·s-1 (considering a cross section area of water to air in the sump of 0.126 m2).      Figure 3-13. Time to consume the oxygen to 18% in the Sullivan mine sampling shed, A:V ratio                        of 2 m-1.     0 20 40 60 80 100 120 140 10 100 1000 H o u rs Water Flow Rates (L/min) Time(hrs) to 18% O2 0 0.8 1.2 2 Safe Dangerous Air Velocity (m/s) a = 0.126 m2 2 days to consume oxygen to 18% Chapter 3 – Sullivan Mine Accident  51    Figure 3-14. Time to consume oxygen to 18% in an arbitrary shed with an A:V ratio of 4 m-1.     Figure 3-15. Safe and unsafe situations based on different air velocities and water flow rates.  for deoxygenated water to reduce the oxygen content in air to 18%.     0 20 40 60 80 100 120 140 10 100 1000 H o u rs Water Flow Rates (L/min) Time (hrs) to 18% O2 0 0.4 0.8 1.2 2 Air Velocity (m/s) a = 0.126 m2Safe Dangerous 2 days to consume oxygen to 18% 0 0.2 0.4 0.6 0.8 1 1.2 1 10 100 1000 10000 Ai r Ve lo ci ty  (m /s ) Water Flow Rate (L/min) Safe and Unsafe Conditions at Sullivan Mine Based on Different Air Velocities and Water Flow Rates 50 hrs to Reach 18% Safe UnSafeChapter 3 – Sullivan Mine Accident  52 3.4. Similar Confined Space Accidents around the World 3.4.1. Earth Dam at Carsington, U.K. In 1982, methane-bearing water seepage in a dam at Carsington, U.K. claimed the lives of four graduate students. Methane-consuming bacteria populated the walls in a well leading to the accumulation of oxygen-depleted and carbon dioxide enriched air. Entry into a drain within the dam resulted in loss of life through "asphyxiation by carbon dioxide". It was reported that sulfide minerals in mud-rock in the dam were being oxidized to produce sulfuric acid that reacted with limestone to produce carbon dioxide at a level sufficient to dilute the air significantly. High organic content in the mud-rock, together with a low flow rate of water through the mud rock caused high dissolved methane concentrations (Gooddy and Darling, 2005).  One graduate student entered the toe drain to check for seepage. He was overcome and collapsed. Three other students were outside and each suffered a similar fate in attempting a rescue. The event had widespread implications in the U.K. regarding an employer's duty to ensure workers are properly trained and competent in all areas of work undertaken. The incident has been described in detail in the following references (Baker, 1986; Pearson and Edwards, 1989; Hooker and Bannon, 1993; McAleenan and McAleenan, 1999). A study conducted on dissolved methane in groundwater in the UK demonstrated that the dissolved methane content was orders of magnitude below the theoretical 1600 µg/L value of its Lower Explosive Limit (LEL) after outgassing in a confined space and higher methane concentrations are required before an explosion will occur. Waters with this much dissolved methane do exist in aquifers with reducing conditions, so in some cases, methane exsolved from groundwater in mud rock can introduce an explosive hazard for a confined space. Mud rock because of its low permeability was not the source of natural groundwater discharge. Therefore, dissolved methane in mud rock water was not considered the problem. Rather the problem arose when mud rock is tunnelled for water transfer between catchments, either for Chapter 3 – Sullivan Mine Accident  53 water supply or hydroelectric purposes, or for more general civil engineering reasons. In such cases, mud rock can have an elevated dissolved methane concentration in water because of its high organic level (Gooddy and Darling, 2005).  3.4.2. Breathing Water Well in Alberta In 1999, two children were killed in a well pit in Alberta. They died from lack of oxygen from oxygen-depleted air blowing from a water-well located in a root-cellar on their farm. The farmer discovered his daughter in the cellar and entered to affect a rescue, but he quickly fell unconscious. His son then attempted a rescue and was also overcome. A neighbor called to the scene was able to extricate the farmer who actually recovered, but the two children could not be revived. This situation has been seen at other locations in Alberta and in the western U.S. and has been dubbed "breathing water-well". In studying the accident, pressure, temperature, air flowrate and oxygen, carbon dioxide and methane  levels were measured. The results show significant influence of barometric pressure changes on oxygen levels (Hill, 2002; CGWA Factsheet #5, 2004) from displacement by nitrogen and carbon dioxide. Deoxygenation of air in the well occurs from contact with oxygen-depleted well-water. Moreover, denitrification of commercial nitrate fertilizers in a perched zone above the aquifer generates nitrogen that is picked-up by the ground water. Nitrogen in the water was at a level in equilibrium with gas containing 90% nitrogen. It is also believed that ground water contacts sulfide minerals or organic matter in a semi-saturated, permeable zone. The mineral surfaces and/or organic matter consume dissolved-oxygen and emit carbon dioxide to ground water flowing into the well. A "breathing water well" is one that has been drilled into a partially-saturated aquitard below a perched or low-conductivity zone in which the well casing stops above the water table. In a "breathing water well", air moves into the well when the barometric pressure rises. Oxygen in this air is consumed within the permeable zone, nitrogen and carbon dioxide are Chapter 3 – Sullivan Mine Accident  54 picked-up, and when the barometric pressure drops, the oxygen-depleted air is blown up the well and into the pit or sump at surface. A drop in barometric pressure of about 270 Pa is needed for the oxygen level in air to decline from 20.5% to 19.0%. When the air pressure rises, fresh air is forced back into the well and the cycle continues. The air flow rate is higher in summer than during colder fall and winter periods whenever a pressure change causes exhalation. However, the rate of oxygen removal by oxygen-depleted water was found to be independent of temperature. 3.4.3. Breathing Coal Mines Water wells are not the only structures in the ground that breathe. Abandoned coal mines do as well. In 1987, a 42 year old woman moved into her new home in Newcastle, U.K. The house had been tightly shut-up for about 2 to 3 days. Shortly after her arrival, she felt dizzy and nauseated and was unable to light a gas fireplace or even a cigarette due to "bad" air that had built up in the house. Gasses from a remediated coal mine beneath her home had leaked into the house. Inspectors could not detect CO or methane, but the air in the house was found to be significantly oxygen-deficient. Escape of oxygen-depleted air from the mine caused the problem in her house and several others over a distance of about 5 km (Hendrick and Sizer, 1992). The heavier carbon dioxide also displaced air (Jeana et al., 2004). Air diffused into an abandoned mine where oxygen was consumed by exposed coal seams. The seams oxidized internally at low temperature in the presence of water and the air became oxygen-depleted. Pyrite is also present in most coal mines in amounts of 1 to 3% which also consumes oxygen and can act as a catalyst in that sulfuric acid generated from pyrite oxidation increases the oxidation rate of specific organic compounds in the coal matrix (Wang et al, 2003). It was also believed that additional oxygen-depleted air derived from the decomposition of timbers in the mine. The sulfuric acid generated by sulfide oxidation can Chapter 3 – Sullivan Mine Accident  55 also react with carbonate minerals to produce carbon dioxide directly which can displace air (Hendrick and Sizer, 1992).  A drop in barometric pressure caused the leakage of oxygen-depleted air and carbon dioxide from the mine. The concentration reported for carbon dioxide was 9.4% while the oxygen level was 9.8 % in a shallow venting exploratory shaft sunk adjacent to the house. The oxygen level was as low as 8-9 % in a cupboard and in the kitchen sink. In living areas oxygen levels were measured as low as 16%, although with rapid drop in atmospheric pressure, it declined 13%. The minimum acceptable level for oxygen is 19.5% (Pettit and Linn, 1987; BC Mines Act, 2008). When oxygen goes below 16%, humans begin to feel physiological effects of oxygen-depletion which includes impaired judgment and difficulty in breathing (Pettit and Linn, 1987). If oxygen drops to any level below normal (20.6%), investigations must be conducted. The most relevant pathway for gas entry was determined to be an underlying defect in the concrete floor through which water pipes entered the home. Geological features (sand layer and faults) above the coal mine allowed bad air to exit from the mine (Hendrick and Sizer, 1992). The sand layer allowed the gas to transfer upwards, while a thin clay layer on top of the sand layer prevented the gas from dispersing into the atmosphere. The concentrated gas found its way to the house through a nearby fault. The National Coal Board sunk a new shaft into the mine to ventilate the mine mechanically. This mitigation was successful as oxygen and carbon dioxide in nearby buildings returned to normal (Hendrick and Sizer, 1992).  The risk of oxygen-depleted air leaking into houses built above coal mines was not recognized prior to this event. Air coming out of a coal mine has never caused any asphyxiation because generally within houses, air does not become confined for significant periods of time that cause the concentration of oxygen to drop. However, if the gas is trapped in places such as cellars and pits, these can become dangerous spaces (Hendrick and Sizer, 1992).  Chapter 3 – Sullivan Mine Accident  56 The carbon dioxide concentration in the house was about 9%. If this gas simply mixes with atmospheric air, the situation, although dangerous, may not be fatal for short term exposure. However, in most cases, displacement is accompanied by oxygen consumption, reducing oxygen to a hazardous level. A barometric pressure rise causes air to flow into shafts and adits providing fresh air for further oxidation, i.e., the mine "inhales". A pressure drop causes bad air to flow out through permeable geological features (sand, faults, or man-made structures), i.e., the mine "exhales". If the pressure drop is slow, bad air mixes with outside air creating an uncomfortable oxygen  level within a structure. If the pressure drops rapidly, bad air can accumulate in a basement. Air from coal mines blows out at several hundred L·s-1.km-2 (Hendrick and Sizer, 1992). There are many cases of oxygen-depleted air blowing into houses built above abandoned coal mines. In North America, deaths by asphyxiation have not yet been reported since bad air is diluted with air entering from outside, but cases of "sick" houses have been observed. In 2003, there was a case where carbon dioxide infiltrated into the basement of a house built above reclaimed surface land above an abandoned coal mine in West Virginia. The residents of the house complained about shortness of breath, lightheadedness, dizziness, poor concentration, and blurry vision when they went to the basement of their two-story building. Their symptoms went away within minutes after leaving the basement. NIOSH initiated an investigation to determine the cause. Gas monitoring showed elevated carbon dioxide levels about 8.8 % in the basement, while the oxygen level was as low as 13.5 %. There were no signs of carbon monoxide or methane. This agrees with the symptoms of the homeowners. Air and soil gas samples from the mine drainage pipe were analyzed for carbon isotopic composition which showed that carbon dioxide infiltrating into the house was from a carbonate source (Jeana et al., 2004).  To mitigate the adverse effect of the leaked gas, similar ventilation to that generally used for radon mitigation was setup. Other preventive measures included maintaining positive air Chapter 3 – Sullivan Mine Accident  57 pressure, sealing the cracks and ventilating by installing fans in the basement. Similar accidents have been reported on houses built on top of abandoned coal mines in many places in West Virginia (Jeana et al., 2004). In some houses, carbon dioxide and oxygen concentrations have reached 25% and 10% respectively in the U.K. and in Pennsylvania. The West Virginia department of environmental protection stated that indoor workers, homeowners, public utility workers, emergency response workers, and remediation workers should use a confined space policy when working in a facility built above an abandoned coal mine. In the case of soil gas leaking into basements and causing symptoms, homeowners should discuss the problem with the Office of Surface Mining to develop active measures. Workers at mine remediation sites should be trained for confined space entry and test the atmosphere regularly for safe entry according to Jeana et al, 2004.  Some additional examples of an atmospheric hazard caused by a change in barometric pressure and outside temperature are as follow: • According to Hendrick and Sizer, (1992) there was an accident caused by change in atmospheric conditions in Cameroon. In this case, a massive volume of carbon dioxide was released from a nearby lake (crater) which asphyxiated about 1700 residences in a nearby village. The CO2 came from a magma crater on top of a nearby volcano (http://www.neatorama.com/2007/05/21/the-strangest-disaster-of-the-20th-century/). Carbon dioxide accumulated at the bottom of the lake in the crater and eventually erupted under pressure, releasing to the surrounding vicinity in significant amounts. The gas flowed down into the village with the aid of wind and took many lives.  • There was a case where the pilot light of a gas furnace in a basement of a church in Echart, Maryland extinguished frequently and people in the basement for any length of time began to feel dizzy. The basement of the church was connected to a crawl space covered with soil beneath a chapel which had a 3 m diameter sink hole. The carbon Chapter 3 – Sullivan Mine Accident  58 dioxide  gas was coming out of a closed coal mine through the subsidence into the crawl space and then into the basement. (Eltschlager et al., 2001(a)).  3.4.4. Asphyxiation from Oil-Contaminated Soil In June 1952, a plumber working in a water manhole died in Minneapolis, Minnesota. The manhole had been built to provide water for a building under construction. It was 1.2 m in diameter with a depth of 2.7 m and constructed of concrete bricks and mortar (Michaelsen and Park, 1954). The employee had been working daily in the space for 2-3 days before the accident without any problem. On the morning of June 10th, another worker entered the manhole to turn on the water without incident. Later in the afternoon the plumber went to turn the water off and was overcome by "bad" air. Samples taken 20 days later showed oxygen levels of about 3%. The cause was found to be oil-contamination in the surrounding soil. The soil consumed oxygen three times faster than "black garden soil" producing high levels of carbon dioxide. The oxygen-depleted air diffused into the manhole because of a concentration difference with normal air. It is likely that the acute nature observed was also related to a significant barometric pressure drop (about 202 Pa) accelerating the flow of "bad" air into the space. Similar to a breathing water well, dangerous air can move into the confined space when the pressure drops. 3.4.5. Soil Penetration by CO Gas In 1997, a municipal sewer project involved installation of new pipes and manholes. Blasting was used to break up rock layers about 2m below surface before excavating pipeline trenches and manhole pits. A construction crew had installed a 3.7m-deep manhole. After the crew left, 120 kg of nitroglycerin in twenty 5.5 m deep boreholes were detonated about 15m away. A worker entered the manhole 45 minutes later and collapsed. Two coworkers tried to rescue him. One was able to retrieve the unconscious worker, but the other rescuer died in the manhole. All workers had elevated blood levels of carboxy-hemoglobin indicating they Chapter 3 – Sullivan Mine Accident  59 had inhaled air with high CO levels. It was determined that CO released from the explosion migrated through the soil into the manhole. CO concentrations in the manhole two days after the incident was 1,905 ppm – well above the IDLH (Immediately Dangerous to Life and Health) level of 1,200 ppm. Even after ventilating, high levels continued for 7 days. The incident shows that CO (and other gasses) from subsurface blasting can migrate through porous material to accumulate in a confined space. This was the first reported fatality from this exposure type although nonfatal poisonings have occurred in basements from nearby blasting (Deitchman et al., 1998). Other examples of blasting related accidents are as follows: 1- November 3-8, 1988, Bucks County, Pennsylvania (PA) - 12 blast holes with a depth of 8.8 to 10 m were filled with TNT and ANFO for a blasting project in sewer construction. Six days after the blast, two occupants of a house 30 m away from the site were reportedly poisoned by CO (Santis, 2001).  2- April 1-2, 1993, Perry Hall, Maryland (MD) – Carbon monoxide from a blast to construct a foundation for a house migrated 4.6 - 6 m to a nearby residence. The occupant of the house smelled what they claimed was “Burnt Concrete” in the evening and reported experiencing moderate to severe CO-poisoning symptoms. The following day, emergency responders measured 210 ppm of CO in the basement which did not return to normal levels even after 6 days (Santis, 2001). 3- December, 9, 1994, Havre de Grace, MD - two blasts (consisting of 110 to 220 kg of extra gelatin) were conducted for a new home foundation. Each blast hole was 3.4 m deep and filled with 2.3 to 2.7 kg of explosive per hole. Twenty minutes after the second blast, the CO level was measured in a model house 11 m away, by the contractor at a level of 87 ppm. After the third blast, this level increased dramatically (Martel et al., 2004). 4- March 7 to April 20, 2000, Adrain, PA – 20 overburden blasts in a small surface coal mine were conducted with 2300 kg of ANFO. Blast depths were 7.6 to 9.1 m Chapter 3 – Sullivan Mine Accident  60 and holes were stemmed to about 4.9 m. After blasting, the ground heaved about 2.3 m. On April 2nd, residents in a house 120 to 150 m away from the blast were diagnosed with CO-poisoning. 650 ppm CO was measured in the basement of their house as well as 400 ppm on the 2nd floor. On April 20th, the CO level was still high at 100 - 200 ppm in the basement (Santis, 2001). In Appendix C.9, gasses that emit from blasting are discussed. Factors that contribute to their emission and ways to minimize this emission are listed. Knowing about the accidents described in this chapter and reviewing the BC Mines Act, it is known that reclamation sites at surface or underground mines have not received sufficient attention for acute atmospheric hazards. Before the Sullivan mine tragedy, few people understood that mining activities on surface can emit gasses that when confined become an acute hazard. So while confined space accidents are a well-understood hazard in underground mining with modern ventilation practices ensuring that work areas of a mine have high-quality air to support the miners' need for air, similar dangers from surface mining blasts is not widely recognized. Danger comes from seam gas, battery-charging stations, diesel engines, oxygen-depletion from exposed sulfide minerals or from displacement by carbon dioxide in coal mines (black damp), coal seam gasses such as methane at explosive levels (fire damp), CO at elevated levels (white damp), or excess heat at depth (Mohammadi and Meech, 2008). As well, coal with low moisture content produces dust, and coals with higher sulfur content may spontaneously combust (Hargraves, 1983). The sources of atmospheric hazards at surface mine sites may be blasting and other explosions, fires, liberation of gasses from the orebody, sulfide dust explosions, decay of organic materials, use of diesel and gasoline in closed areas, or gas carried by water or transferred through the soil. Many of these hazards are unrecognized in surface mining regulations and practices. As a result, there is no source that discusses these issues directly, except for a few reports that suggest atmospheric problems may exist at a reclaimed site. If the basic chemical processes that control gas generation from soils and Chapter 3 – Sullivan Mine Accident  61 minerals are known, then one can begin to understand which conditions lead to the appearance of these gasses. Appendix C (in the form of an Atmospheric Confined Space Manual - especially for gasses from mines and soils) discusses existing or possible atmospheric hazards from all industrial work places (especially mine wastes, minerals, and soils) and highlights pathways for gas emission and diffusion at mine sites to confined areas. 3.5. Summary The tragedy at the Sullivan mine provides a valuable lesson that personnel in the mining industry must learn in order to account for unforeseen atmospheric risks at surface reclamation sites. The accident was compared with other similar confined space accidents, where temporal changes in temperature and pressure caused "bad" air to flow into a confined structure. How temperature and pressure control gas flow is clear in each accident, but the knowledge was obtained after the fact (Mohammadi and Meech, 2011). When a confined space is connected to an outside environment, atmospheric pressure and temperature changes affect gas flow into and out of the space. Knowledge about these diurnal and seasonal changes is essential in assessing a confined space. As a result, suspicious enclosed structures (especially ones near naturally- or industrially-contaminated areas) should be tested at different times of the day, month, season, and in different years to ensure the space is safe. A unique lesson from the Sullivan Mine tragedy is that the sampling shed was safe before the change in design in which the toe-drain was covered creating a direct connection between air in the shed and air in the dump. Before the glacial till cover was placed on the dump and the toe extended, the oxygen-depleted water seeping from the dump picked-up oxygen as it traveled along the open channel. With the new design, the till-covered dump essentially acted like a giant tea bag and the pipe like a straw. With the cover in place, air flowed into the dump and gas flowed out of the dump to be concentrated in the sump. So, what was a safe place became a dangerous one.   62 Chapter 4   4. Analysis of Factors that Contributed to the Sullivan Mine Accident  4.1. Overview It was shown that seasonal atmospheric temperature fluctuations have been found to significantly affect oxygen-depleted gas emissions in the Sullivan mine accident. Also, several examples of atmospheric-related confined space accidents were discussed in Section 3.4 in which temporal atmospheric pressure changes caused hazardous gas emissions into a confined space (Mohammadi and Meech, 2011). Although diverse with respect to industry type and other characteristics, atmospheric-related confined space accidents have many similarities. These common characteristics can aid in defining and modeling these hazards.  Currently there is no literature that discusses the combined effect of atmospheric temperature and pressure changes on hazardous gas emissions into a confined space, especially one associated with waste dumps. This chapter describes how temperature and pressure differences between the inside of the dump and the outside atmosphere affect gas flow from the dump into an ARD collection pipe. Understanding the problem will be the foundation to develop methods to conduct atmospheric risk assessment of confined spaces at mine waste dumps as well as other confined space sites such as breathing water wells.  4.2. Effect of Atmospheric Conditions on Confined Space Risk Atmospheric-related dangers have different sources that can be categorized according to the presence of organic materials, minerals and soils or gasses from specific operations and activities. With a confined space connected to an outside environment, atmospheric pressure and temperature variations due to climatic changes affect gas flow in and out of the space. Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  63 Examples of such accidents are breathing water wells; breathing coal mines, and manholes located in oil-contaminated soil. These effects are extremely important in establishing and defining risk from atmospheric conditions in the confined space. If temperature and pressure effects change the direction of gas movement, hazardous gas flows through the confined space and to outside, while if they cause air movement into the space from the outside, natural ventilation of the space results. The confined space accident at the Sullivan mine is a typical example of how atmospheric changes affect gas emission. Phillip et al., (2008) performed extensive data measurements of the Number One Shaft waste dump from 2006 to 2008. This work monitored internal and outside pressure and temperature as well as gas velocity in the effluent pipe. Their extensive dataset has allowed the effect of temperature and pressure variations on gas emission to be quantified to assist in later development of an atmospheric-related fuzzy risk assessment of a confined space at a mine reclamation site.  4.2.1. Development of Theory  While temperature alone is able to describe the gas velocity variations, there is a need (and desire) to incorporate the pressure effect. The data have been reexamined in this research to understand the general criteria for determining gas flow direction and magnitude in dumps by considering both pressure and temperature effects. In order to build on this knowledge and attempt to improve the prediction, a First Principles model using the first and second laws of thermodynamics was created. While it is clear that atmospheric temperature affects gas flow in and out of the dump and atmospheric pressure is known to affect gas flow in and out of breathing water wells and underground mines, the real variables of importance are the difference between the atmospheric and internal dump, mine, or well conditions (temperature and pressure). In order to understand this, the approach involves calculating the free-energy changes within the dump that cause gas to flow out or air to flow in. Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  64 The development of the theory in this section is based on recent work by Cengel and Boles, (2006). The first law of thermodynamics, also known as the conservation of energy, states that energy can neither be created nor destroyed during a process that takes place within a system; it can only change forms. Therefore, the sum of energy into the system should equal the sum of energy leaving. However, satisfying the first law alone does not show that a particular process takes place, because a process must proceed in a certain direction. While the first law poses no restriction on the direction of a process, the second law can be used to identify its direction. The second law shows that energy has a quality component (with respect to its utility) that indicates the degradation of energy during a process. According to the second law, systems spontaneously move from order to disorder, thereby increasing the disorder of the universe. A process will happen if it satisfies both laws. The property known as exergy is a valuable tool in determining the amount of useful work that can be gained from a system by bringing it into equilibrium with the environment. In order to satisfy both laws of thermodynamics, a system’s exergy must be calculated. The amount of energy available for useful work in an irreversible process is less than the reversible work. The difference between reversible work and useful work is because of irreversibilities present in the process and is called the lost (or destroyed) exergy (Cengel and Boles, 2006). The quantity of the energy does not change during the process (first law) but its quality must decrease according to the second law. Irreversibilities such as friction, chemical reactions, heat transfer through finite temperature differences, unrestrained expansion, nonquasi- equilibrium compression or expansion, all of these generate entropy. Entropy describes the tendency for systems to go from a state of higher organization to a state of lower organization on a molecular level. Figure 4-1 depicts two types of systems. In Figure 4-1(a), material flows through a control volume in which changes in the conditions of the material lead to work done on the external environment, while in Figure 4-1(b), work is done on the system to compress material into the control volume without material flow through that volume. Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  65   Figure 4-1. (a) single-inlet single-outlet steady flow system – exergy flow (g2258); (b) Flow of energy to                      move gas in and out of a control volume (Mc Graw Hill Ltd, Cengel and Boles, 2006,                      adapted by permission).  The exergy change in a gas flowing through a control volume as shown in Figure 4-1 (a) is defined by Equation 4-1:         Xwork-Xheat+(XO-XI)=∆X (4-1)  Where Xwork is the sum of all work interactions per unit mass of the working gas on the outside world other than that needed to push air and gas into and out of the system (flow work). The variable Xheat is the heat transferred to the unit-mass of gas. If the exergy of a flowing gas as it is processed from state 1 to state 2 over time does not change, then ∆X = 0. As well, for a system with no work and heat interactions, the exergy balance for a unit-mass of steady gas flow is:  g4666XO-XI g4667=0                                                  (4-2)  g4666hO-hI g4667+Td g4666sO-sI g4667+ g4684Vg4652g4652g1318O22 g4685 - g4684Vg4652g4652g1318I 2 2 g4685+gg4666zO-zI g4667=0                                                   (4-3)   Xdestroyed=Td g4666sO-sI g4667  (4-4)   where g3435hO-hI g3439 is the change in enthalpy as defined by Equation 4-5: hO-hI = Cg3365 g2926 g4666Tg2899 g3398 Tg2893 g4667 g3397 g1874g3031 g4666Pg2899 g3398 Pg2893 g4667 (4-5)  Combining the change in enthalpy and entropy, free energy is obtained as in Equation 4-6:     g4666hO-hI g4667+Td g4666sO-sI g4667= fO-fI (4-6) W flow P v Control Volume Pv before entering Pv after entering P v Control Volume W flow (a) (b) Tout, Pout, Vout, zout Tin, Pin, Vin , zin Control Volume ψin ψoutHeat Transfer Work Environment T0, P0Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  66 in which Td and vd are properties of the dead state where the system is in equilibrium with the environment in which it is located, i.e., the system is in equilibrium with temperature and pressure of the environment and has neither kinetic or potential energy. As a result the exergy of a system that is in equilibrium with its environment is zero. By definition, the term environment refers to the region beyond the immediate surroundings of the system whose properties are unaffected by the process. As such, a system’s exergy depends on environmental properties as well as its own properties. In an open system (control volumes), a form of energy exists that is called "flow work", v0 g3435Pout-Pin g3439. This type of energy does not apply with a closed system. This is the work that pushes gas into or out of a control volume – see Figure 4-1 (b). Flow work maintains gas flow in a pipe or duct and is included as part of free energy in Equation 4-5. The specific volume of gas, v0, is equivalent to the volume change of a unit mass of fluid as it is displaced during flow. In Equation 4-6, the entropy change, for an adiabatic, single stream, steady flow control volume for the unit mass of an ideal gas is: sO-sI=Cp dT T +R dP P = g3505 Cp g4666Tg4667dTT +R lnPOPIOI  (4-7)       If Cp as a function of temperature remains constant, then the entropy change is:  sO-sI=Cg3365pln TOTI +RlnPOP]  (4-8)  So the specific entropy of the gas must increase as it flows through an adiabatic control volume, i.e., sgen≥0. The difference in entropy for an outflow situation is positive while for an inflow condition, the difference in entropy is negative.    Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  67 4.2.2. Results of the Theory When the air inside the waste dump is heated due to the sulfide reactions in the dump, the air inside will expand and rise through the top surface of the dump doing work on the surroundings. The waste dump is considered an open system since the mass crosses the system boundary. As the process in the waste dump is irreversible, the change in exergy must be taken into account. The exergy change (∆X) of the gas flowing through the dump as it transforms on an hourly basis was calculated from the dump data. This change was found to be essentially zero over the measurement duration. Therefore the dump can be treated as being at steady state over a one hour period. The potential energy is also negligible in comparison to the free energy changes in this case, but may be required for higher dumps. By applying these conditions to Equation 4-3, the result for an adiabatic system (with no outside heat interaction) and no work interaction (i.e., at steady state) gives an exergy balance between the outside and inside of the dump as in Equation 4-9:   Cg3364g3364g3364p(TO-TI)+vg4666PO-PI g4667+TO g4666sO-sI g4667= g4684Vg4652g4652g1318I22 - Vg4652g4652g1318O 2 2 g4685  (4-9) Note that according to Equation 4-6, the left side of the above equation is the delta free energy or ∆f (outside-inside). Here the temperature measured in the monitoring station is considered to be the dead state condition. This condition is shown by T0 and was measured at waist height in the sampling shed. Temperature and pressure measured inside the dump 27.4 m down BoreHole (BH) 1A (TI, PI) (see Figure 4-2) and outside on the dump slope (TO, PO) were considered to characterize the inlet and outlet conditions of the exergy balance equation respectively. Note that the pressure was specifically recorded above each bore hole, while outside temperature was measured at one place (meteorology station). The pressure above borehole 1A was chosen as the outside pressure since borehole 1A was located at the centre of the dump and so, it was considered representative of the pressure above the dump. Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  68 As the dump height is 55 m, the pressure measured at 27.4 m down BH 1A does not represent the pressure at the bottom of the dump, but since this was the lowest level for which a pressure measurement was available, this was considered representative of the internal pressure in the  dump. As expected, all values for (PO - PI) were negative which shows that the outside pressure was lower than the internal pressure. This value varied seasonally between -1 Pa and -12 Pa in summer and between -12 Pa and -30 Pa in fall and spring.  In Equation 4-9, the gas velocity inside the dump (VI) and velocity of the gas while it moves out across the entire dump surface (VO) cannot be measured. From Equation 4-9, it can be seen that free energy due to temperature and pressure differences between the inside and outside of the dump causes the gas to flow into or out of the dump by convective and/or advective flow. Gas going into or out of the pipe is part of this massive convective gas flow. As a result, the velocity of the gas flow measured in the pipe should be correlated with the free energy difference between the inside and outside of the dump and so can be used as a proxy – see Figure 4-3. A clear relationship between ∆f and the gar or air velocity is evident.   Figure 4-2. Locations of the boreholes at No.1 Shaft waste dump (Phillip et al., 2008, adapted by                      permission).  PipeChapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  69    Figure 4-3. Correlation of gas velocity in the pipe with ∆g1806.                      Equation: y = 1E-07x4 + 8E-06x3 - 0.0002x2 - 0.0312x - 0.0087, R2 = 0.929.  As can be seen in Equation 4-9, rather than using measurements of a single temperature and single pressure to predict gas flow, First Principles use differences in these two conditions. (TO-TI)  = Difference in outside temperature and inside temperature of BH 1A (PO-PI)  = Difference in outside pressure and inside pressure of BH 1A By considering differences in outside and inside conditions, the graph should pass through zero velocity at zero free energy difference. The correlation of determination (R2) for this relationship is slightly better (0.929 vs. 0.927) than that for outside temperature alone in Figure 3-4.   In order to further improve the correlation shown in Figure 4-3, the data were separated into the four seasons. As can be seen in Figure 4-4, summer, fall, and winter data give a higher overall correlation of determination (0.954) which justifies the claim that the relationship between gas velocity and ∆f vary seasonally with different patterns. However, the data for spring show much more scatter and the relationship is linear rather than fitting a -60 -40 -20 0 20 40 60 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 G as  Ve lo cit y (m /s ) ∆f(J/g) R-Square 0.9296Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  70 polynomial expression suggesting that a transition to a smaller pore volume being involved occurs in the cold months of fall and winter through freezing of ice in the cover and upper surface of the dump as depicted in Figure 4-5. Due to a higher pore volume in spring, greater amounts of air flow into the dump at very negative delta free energy values. The maximum positive gas velocity of 1 m·s-1 occurs in spring compared to 0.75 m·s-1 in late fall and winter. When colder conditions involve lower pore volume, the gas velocity is restricted to about 0.75 m·s-1. During the summer, gas blows out of the pipe reaching a maximum negative velocity of -1.0 m·s-1, showing that much more of the internal pore volume is involved in the process as would be expected. Figure 4-6 schematically represents how the pore volume of air shrinks during colder periods to affect the pattern of inhaling and exhaling throughout the year.    Figure 4-4. Correlation of gas velocity in the pipe with ∆g1806 (Summer, Fall and Winter, 2007), the                      trendline is shown for Summer, Fall, and Winter                      Equation: y = 2E-07x4 + 9E-06x3 - 0.0003x2 - 0.0314x + 0.0132, R2 = 0.9537.   Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  71  Figure 4-5. Correlation of gas velocity in the pipe with ∆f (Spring 2007).                         Equation: y = -0.0283x - 0.0446, R2 = 0.837.     Figure 4-6. Representation of volume change in Number One Shaft waste dump                      (Chnages occur during the Spring melt)    ∆f (O-I) (J/g) Spring 07  Waste dump Pipe Changeable Internal Volume        TI, PI, VI Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  72   Figure 4-7. Diurnal changes in (PO- PI),  (TO- TI)  and gas velocity at the same time period in the                      year following the accident (May 9th to 16th 2007). Internal data are from 27 m down                      borehole 1A (data by permission from Technical Panel of the Sullivan Mine                       Accident, Teck Metals Ltd).  The dump inhales and exhales during the cold and warm times of the year respectively as well as cold and warm periods of each day respectively – see Figure 4-7. In some months or seasons, the inhalation and exhalation patterns may change to days of inhalation without exhaling and vice versa. For example, during the warm months of the year (July and August), the dump mainly exhales with long durations of negative velocity. In colder months (such as late November, December, and February) the dump mainly inhales with a positive velocity.  In the "cooling or heating" seasons (Spring and Autumn), daily variations in gas flow become very important due to fluctuations occurring in outside temperature and barometric -30 -25 -20 -15 -10 -5 0 5 10 15 5/9 5/10 5/11 5/12 5/13 5/14 5/15 5/16 Te m pe ra tu re  an d Pr es su re  D iff er en ce Time (Hours) dP (Pa) dT (  C)o -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 5/9 5/10 5/11 5/12 5/13 5/14 5/15 5/16 Ga s Ve lo ci ty Time (Hours) Gas Velocity (m/s)Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  73 pressure that straddle the internal dump temperature. In any specific climate, the balance between pressure and temperature outside and inside the dump throughout the year determines the direction of gas flow.  4.2.2.1 Effect of Pressure and Temperature Difference on Gas Flow  The relationship between (TO- TI) and (PO- PI) is shown in Figure 4-8. The minimum value of (PO- PI)  is about -30 Pa which is probably too low to cause any significant gas flow. The values of this variable are always negative showing that the pressure inside the dump is always higher than the pressure above the dump, which would tend to drive pore gas up to the top of the dump. The energy due to (PO- PI) varying from -30 Pa to -1 Pa is very small ranging from -0.04 to 0.00 J·g-1 which is three orders of magnitude less than that due to the maximum and minimum temperature differences with a (TO- TI) varying from +20oC to -28oC occurring respectively in August and December, producing a free energy range of +20 to -30 J·g-1 respectively at these times of the year. As such, the pressure driven energy has barely any noticeable effect on gas velocity and so, the temperature difference is the dominant factor affecting gas flow in reactive waste dumps. According to Phillip et al., (2008), the measured barometric pressure and air velocity are not correlated, but Figure 4-8 does suggest a small degree of correlation with a broad error band. So despite the fact that the outside pressure varies considerably, for example at a time interval of one hour for data from September 2006 and December 2007, where the outside pressure increased to a maximum of 887 Pa in one hour and decreased to a minimum of 543 Pa in another one hour period, at the same time, the internal pressure at 27.4 m down BH 1A changed by a similar value in both cases (i.e., 882, 542). So, despite dramatic changes, equilibrium with the outside pressure occurs rapidly within one hour possibly due to channeling (hot spots) in the cover. The length of time required for a response to barometric pressure changes is determined Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  74 largely by the dump permeability and only in very fine grained wastes (silts and clays) will the response be prolonged (personal communication, Andy Robertson).  It may be that the duration for the internal pressure to equilibrate with outside pressure will increase in thicker and/or higher waste dumps. As a result, during periods of decreasing outside pressure, (PO- PI) may become negatively large in high elevation waste dumps and will decrease the delta free energy to a further negative value. Increased negative free energy will result in positive gas velocity at the bottom of the dump as in Figure 4-3. This means a lower hazard of gas emission at the bottom and perhaps, safer conditions. Of course, the final value of gas velocity is determined by energies derived from both pressure and temperature differences.  Figure 4-8. Relationship between pressure difference and temperature difference.  The effect of pressure difference is very small in low profile waste dumps (~50 m) such as the Number One Shaft waste dump. According to Wels et al., (2003), the potential for horizontal advection at the toe is higher in dumps with greater height. Advective response would be more important for thicker unsaturated waste rock piles as shown by Massmann and Farier (1992), because with high dumps, the outside pressure at the top of the dump is -20 -15 -10 -5 0 5 10 15 20 25 -40 -30 -20 -10 0 dP (O - I) ( Pa ) dT(O-I)0C R-Square 0.2556 TO-TI  ( 0) P O - P I (P a)   R2 = 0.2556 Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  75 much lower than the outside pressure surrounding the dump near the bottom and the dump will inhale air at the bottom. Shaw et al., (2002) have mentioned the Chimney Effect: the combination of elevated temperature within rock piles due to ongoing oxidation, and high elevation gradients (490 m high) resulting in high rates of air inflow in the pile. According to Shaw et al., (2002), Sugar Shack South dump at the Questa mine in New Mexico is an example of a dump that shows a Chimney Effect. At this dump, the internal temperature is high at 40 oC and the oxygen content is higher during winter (6-12%) in comparison to summer (2-5%). The higher oxygen level in winter may be related to a negative (TO- TI) value in winter. The carbon dioxide profile in one borehole at Sugar Shack shows the opposite profile to the oxygen pattern in which the level drops to about 2% in the winter and increases to about 4.5% in the summer.  As a result of (the) chimney effect, higher waste dumps continue to heat up and are not oxygen-deficient. Although the higher internal temperature will cause more gas flow into the waste dump, as…oxidation continues, (this) will affect (a) great oxygen transfer in the pile and increases the leachable contaminant generation (Shaw et al., 2002).   According to Wels et al., (2003), the potential for advective flow is higher in more permeable rock piles with high height to depth ratios, while the gas flow due to temperature differences depends on spatial permeability (coarseness) and the reactivity of dump material which controls the internal temperature. This suggests that higher waste dumps are safer than low profile dumps as air blows in at the dump toe to cause less oxygen-depleted gas generation inside the dump and no emission from the toe which agrees with the findings of the proposed model. Wels et al., (2003) described the oxygen variation in response to atmospheric pressure. When the outside pressure at the sides of the dump increased, compression of the gas phase within the pile occurred leading to entry of fresh oxygen. According to Ritchie, (1994), a wind velocity of 10 m·s-1 at a dump with a height of 15m, an area of 25 ha, and a gas-filled porosity of 0.3 causes a pressure gradient of 1 Pa·m-1 which can significantly affect gas flow, but comparably less than that of a temperature gradient (gas Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  76 velocity of 2 ·10-5 m·s-1). Ritchie, (1994) explains that advective gas transport caused by pressure unlike temperature-driven flow, has not been quantitatively measured because temperature gradient lasts for years while pressure gradients only lasts for hours. It is thus concluded in this work that, while in high permeability dumps, barometric pressure changes quickly reaching steady-state throughout the entire dump, temperature changes take considerably longer to reach a new equilibrium.  In White’s dump at the Rum Jungle mine in Australia, Harries and Ritchie, (1985) observed a diurnal barometric pressure effect on oxygen concentration in borehole D before the dump was covered in which they noted similarities between the plots of barometric pressure and oxygen content. When barometric pressure decreased by about 500 Pa from 8 am to 6 pm, the oxygen level dropped from 5% to 1%, and when the barometric pressure increased by about 400 Pa from 6 pm to 12 am, the oxygen rose from 1% to 8% at a depth of 10 m in this borehole. According to Ritchie, (2003), this effect was not studied for periods longer than a day. The dependence of oxygen at a depth of 7.5 m on atmospheric pressure as a function of time over 28 hours was clear in borehole A at this dump, where similarity between the two plots strongly suggested semi-diurnal changes in atmospheric pressure caused rapid changes in oxygen and carbon dioxide levels in this borehole before the dump was covered. The energy caused by a pressure drop of 350 Pa led to an 8.5% decline in oxygen within 8 hours, and the energy caused by a pressure rise of 300 Pa caused the oxygen to increase by 8% from 4:30 pm to 22:30 pm (Harries and Ritchie, 1985). It is not known, however, much of the oxygen change was due to any accompanying temperature changes, or how much was due to the pressure difference. Smolensky et al., (1999), also recognized the effect of barometric pressure on advective flow in the waste dump at the Nordhalde mine in Germany. At this site, the oxygen content in borehole 37 varies with barometric pressure (Smolensky et al., 1999). A pressure increase of 1100 Pa caused the oxygen content at a depth of 7 m to increase from 0% to 18% in about 4 Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  77 days, followed by a pressure drop of 300 Pa that decreased the oxygen content from 18% back to 0% in 3 days. Over the same time period, the pressure change at a depth of 14 m in the same borehole increased the oxygen level from 0% to 0.8% followed by a decrease from 0.8% to 0%. Note that the thickness of the Nordhalde dump is 80 m with a length of 1450 m and a width of 800 m. It is obvious that temperature changes show a clear seasonal pattern in increasing or decreasing oxygen content in the dump, while the pressure effect is short term and does not show a regular pattern (Smolensky et al., 1999). The claim was made that the pressure effect is more dangerous since it can cause the oxygen content to drop dramatically from 19% to 0% in a few days, showing that a significant advective gas flow has occurred causing oxygen-depleted gas to blow out of the dump. According to Wels et al., (2003), no detailed study has yet been carried out to precisely determine the net effect of pressure-driven gas flow and so, much more work is needed. It is important to study the net effect of pressure since seasonal variation in oxygen concentrations in the same borehole at the Nordhalde dump has been described as being due to thermally-induced gas convection by Lefebvre et al., (2001(a)). Seasonal variation in internal oxygen content in borehole 36 near the edge of the dump is obvious in this dump (Smolensky et al, 1999), in which during colder periods when TO < TI (about 10 oC), the oxygen content inside the dump increases from 0 to a level of 8%, while in warmer times when TO > TI, the internal oxygen content returns to 0% (Smolensky et al, 1999). It is possible that barometric pressure can cause advective flow in dumps which do not have an effective cover. The oxygen content will increase in dumps where the pressure difference between the outside and inside of the dump is very high, e.g., 1000 Pa as with the Nordhalde dump.  Currently there is considerable controversy about the effect of dump permeability and cover permeability on advective flow and so, further investigations are needed.  There are no specific studies at any dump that separates out the effect of pressure difference from the effect of temperature difference and little work has been done to Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  78 demonstrate the specific impact of pressure difference on gas flow. Since the most important feature of all waste dumps is that they are being heated-up from the sulfide reactions, it is reasonable to conclude that temperature differences cause most of the changes in gas flow.  In water wells, with little to no source of heat to cause a significant temperature gradient, it is reasonable to believe that pressure changes cause gas flow. Pressure difference is more significant in breathing water wells and coal mines because of their underground nature. 4.2.2.2 Gas Flow Estimation  It is believed that ∆f rather than outside temperature should be used to calculate and predict gas velocity in order to account for both the temperature and pressure effects. However, despite the high correlation of determination shown in Figure 4-3, there is still a significant range of values in the independent variable around the zero velocity situation. The uncertainty in the change of free energy and gas velocity shows that fuzzy rules are perhaps, a better way to estimate risk at any particular waste dump (e.g., pathways changes, measurement errors, heterogeneities, etc.). Table 4-1 shows how it is possible to predict the gas velocity direction and magnitude given a value for delta free energy at different times of the year. These rules derive directly from the curves in Figure 4-4 and 4-5. Table 4-1. Fuzzy rules to predict gas velocity from free energy difference at bottom of the dump.  Free energy difference between outside and inside of dump (J·g-1) Direction of gas velocity at bottom of the waste dump Summer Fall Spring Winter Positive Very High ∆F >32 NVB - - - Positive High 10 < ∆F <32 NB NB (unlikely) NB - Positive Moderate 7 < ∆F <10 NS NS NS - Positive Low 0 < ∆F <  7 NVS NVS NVS NVS (unlikely) Negative Very Low    -2 <dF  < 0 PVS PVS NVS NVS Negative Low -7 < ∆F <  0 PVS PVS PVS PVS Negative Moderate -10 < ∆F < -7 PS (unlikely) PS PS PS Negative High  -32 < ∆F < -10 - PB (unlikely) PB PB Negative Very High  ∆F < -32 - - - PVB Negative Very Big - NVB  (< -1), Negative Big - NB (-0.7 to -1), Negative Small - NS (-0.3 to -0.7), Negative Very Small - NVS (0 to -0.3), Positive Very Small - PVS (0  to 0.3 ), Positive Small - PS (0.3 to 0.7), Positive Big - PB (0.7 to 1), Positive Very Big - PVB  (>1).  Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  79 The methodology to estimate the direction and level of gas flow in a waste dump as a function of free energy difference is clear in Table 4-1. In order to proceed with a risk assessment based on this method, internal temperature and pressure measurements of the dump as well as temperature and pressure measurements outside the dump are required.  Measurements of internal temperature and pressure in borehole 1A at 27.4 m depth at the Number One Shaft waste dump were chosen to calculate ∆f for this prediction. The internal temperature in shallower boreholes (BH 3A) varies seasonally and is affected by outside temperature more than by the internal reactions. However the internal temperature of other boreholes near the toe in deeper areas (BH 2B in Figure 4-2) do not show such effects and so will yield results similar to boreholes at the centre of the dump. Internal temperatures in the dump show seasonal changes to a depth of about 6 m in bore holes located at the toe which can be attributed to heat escape through the dump surface to the atmosphere. Therefore, temperatures at shallow depths (<6 m) in all boreholes are not representative of the real internal temperature of the dump and are unreliable in predicting the direction of gas velocity. While the driving force for gas convection is related to high temperatures at the dump centre (or perhaps, deeper areas at the toe), applying the internal temperature of boreholes located in such shallow areas should be avoided.  If the assessment is done at a particular time of the year with only a single measured data set, knowledge about maximum monthly atmospheric temperature and the pressure at that time throughout the year allows the system to calculate the direction and value of gas velocity at the bottom of the dump on a temporal basis over the year using the rules in Table 4-1. It is important to measure the outside pressure at the apex of the dump because when outside pressure drops (if the pressure difference is significant) gas escapes from the dump to the outside at the apex, not along the sides of the dump. Risk assessment is less sensitive to the position where the outside temperature is measured, although in calculating ∆f, the outside temperature should ideally be measured half way up the dump.  Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  80  Figure 4-9. Effect of seasonal changes on internal temperature of boreholes               1A, 3A and 2B. location of boreholes are shown in Figure 4-2.                               (data provided by the Technical Panel of the Sullivan Mine Accident by                                      permission, Teck  Metals Ltd)  Since the Number One Shaft waste dump is very heterogeneous with respect to rock size and material reactivity, the internal temperature may fall across a wide range spatially within the dump and so, assuming a single internal temperature is probably unrealistic. However, with current knowledge and sensor capabilities, a single measurement point is likely the most practical way to assess the direction of gas flow. Internal temperature will differ from one dump to another based on many factors related to mineralogy, climate, particle size, water flow, etc. This simplification is correct if the internal temperature is specific for one waste dump, but the result only represents the period for which the risk assessment is conducted. For example, the pivot temperature for change in the direction of gas flow ranges from 5 to 16 ºC and depends on the season (Phillip et al., 2008) – see Figure 3-4. Although this range is wide, an average of 11.5ºC is apparent. This value is equal to the average temperature measured at 27 m (90 ft) down a borehole at the centre of the dump (1A) from March 2007 to December 2007. At different depths in borehole 1A, average internal temperature varies as follows: 14.6 ºC at 21 m, 15.9 ºC at 14.6 m, and 14.2 ºC at 8.2 m. The peak in internal Te m pe ra tu re  °C Te m pe ra tu re  °C Te m pe ra tu re  °C 27.4 m 2.4 m 5.5 m 1.2 m 5.5 m 3.9 m 2.4 m 21.0m14.6m 1.8  mChapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  81 temperature (15.9 ºC) occurs at about 14.6 m depth. The internal temperature of Number One Shaft waste dump varies from surface to depth in other boreholes as well (Figure 4-9). However, note that the central peak temperature of the dump at depth is relatively stable on a yearly basis and continuous measurements of this value over a year are not needed.  To conduct a risk assessment, a single internal temperature can be measured at the centre (middle height) of the dump. The depth of maximum internal temperature varies at different dumps depending on age and reactivity. The temperature at the centre of the dump most likely will not equal the maximum internal temperature as this point will change over the years as the internal reaction zone descends through the dump. In Number One Shaft waste dump, the maximum internal temperature of about 16 ºC occurs at a depth above the centre (at about 0.3 x dump height). The pivot temperature that marks the change in direction of gas flow is below the peak temperature at 0.3 but close to the temperature at 0.5 x dump height (about 12ºC). Therefore an internal temperature reading at 0.5 x dump height is a better measurement point. Due to the abundance of oxygen at some points near the edges of the dump, oxidation rates can increase and therefore the internal temperature can reach levels higher than the peak temperature at the centre. This effect was seen in Number One Shaft waste dump, with boreholes 2B and 3A which are closer to the edges of the dump, the internal temperature reached a maximum of 21 ºC and 19 ºC respectively. An increase in oxidation rates at the edges was also seen at the Nordhalde dump (Smolensky et al, 1999). The maximum internal temperature at the edges of the Number One Shaft waste dump is not close to the pivot temperature for change in direction of gas flow and so, is not useful for risk assessment. Since the oxidation rate of sulfides in the dump changes over time, one can expect the internal temperature to increase or decrease as reactions either grow or fall in intensity, depending on many variables such as dump age, sulfide content, etc. For example, in the Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  82 case of the Number One Shaft waste dump, it appears that the internal temperature is currently in a rising mode at the rate of about 2 oC per year. 4.3. Summary This chapter has described how atmospheric pressure and temperature changes affect gas flow through a pipe installed to collect effluent from the toe of a waste dump. Combining these effects by taking into account the first and second laws of thermodynamics is proposed for predicting the gas velocity. Temperatures inside and outside a dump are the most important factors that affect gas velocities into and out of a sampling shed. Pressure differences are of a short term nature with three orders of magnitude lower impact than temperature differences for the Number One Shaft waste dump. Pressure plays a role, but to a lesser extent than with a "breathing water well" and other underground soils/rock as temperature gradients are generally much lower in these structures than in a waste dump. Advection at the toe is greater for high elevation waste dumps. The effect of cover permeability and dump permeability on pressure driven gas flow needs more investigation.  Taking free energy changes into account instead of outside temperature alone shows benefit in that both pressure and temperature differences are accounted for. The fact that the graph of gas velocity versus ∆f passes through the origin indicates the reliability of this relationship. The method is applicable to assess gas emission from other waste dumps. The dump free energy difference from inside to outside a dump is expected to accurately estimate velocity, i.e., gas emission. Since the degree of danger changes with climate changes, it is unreliable to simply rely on a random gas measurement in a possible confined space at a reclamation site. Atmospheric pressure and temperature changes occur daily, weekly, monthly, and seasonally. As a result, in a matter of a few hours, the risk of entering a confined space can transform from "safe" to "hazardous". Although such changes have regular patterns (in summer, gas flow is mostly out Chapter 4 – Analysis of Contributing Factors to Sullivan Mine Accident  83 of the bottom of the dump, while in winter the gas flows out of the top of the dump), during the cooling (fall) and heating (spring) seasons, conditions are unstable and can change daily and/or weekly. This shows the importance of being aware of the existence of such a hazard not only at reclamation sites, but also at other enclosed structures and spaces affected by pressure and temperature changes such as a manhole located in contaminated soil, a "breathing water well", or other naturally confined spaces such as caves. Assessing the risk based on free energy will take into account the seasonal and diurnal changes in temperature and pressure.   2  Many parts of this Chaper have been published previously. Meech, J.A., Mohammadi, L., January 2011, Confined Space Atmospheric Risk Assessment, Focus on Tomorrow, WorkSafeBC, pp. 81. 84 Chapter 5   5. Atmospheric Fuzzy Risk Assessment for Confined Spaces2  5.1. Overview The decisions required to operate a reclamation site effectively and safely with respect to atmospheric danger involve designing the reclamation activities to accomplish one or more of the following:  1. Minimize, eliminate, or control the generation of oxygen-depleted gas in the dump; 2. Minimize, eliminate, or control the emission  of toxic gas from the dump; 3. Identify structures or surface anomalies that accumulate or concentrate toxic gas; 4. Eliminate or control the ability of these structures and anomalies to accumulate toxic gas; 5. Control human access to the structures and implement confined space safety practices.  Although these goals may sound straight-forward, the first requirement is to identify (or perceive) that a potential atmospheric risk is high. AFRA has been designed as a logical process to understand and conduct a detailed site assessment to provide this perception. In this chapter, the key variables that affect gas generation, gas emission, gas confinement, and human exposure are identified. AFRA follows a step-by-step entry of relevant data that determines a Degree of Belief (DoB) that each of these issues is "High" or "A Problem". These elements are collected together to assign a linguistic risk probability to entering an associated confined space.  The system has been verified against six waste dumps from around the world for which sufficient data are available to provide a direct/indirect comparison of prediction and actual Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 85 conditions. The comparison in all cases is excellent. Three waste dumps not used to create the system were also examined to test the system against new situations. This validation process also shows excellent agreement with reality in terms of internal temperature and direction of gas velocity estimates. 5.2. Research Problem  To understand how the Sullivan mine accident happened, it is necessary to break down the reclamation activities into their specific contributions that increased the risk of an accident. As demonstrated in Chapter 3, covering the dump during the reclamation activities resulted in the conversion of the ditch into an underground drain – this change was the first element in a chain of events that led towards the accident. The drain acted as a hydraulic conduit for air and pore gas to flow between the dump and the shed and it prevented oxygen- depleted effluent waters from dissolving oxygen from contact with air prior to entering the shed. Neither of these dangers was recognized at the time. While the effluent flow was too low to cause oxygen-depletion at this site, recognition of that danger might have triggered an investigation of the extreme hazard created by the atmospheric connection. Seasonal temperature changes (mainly in the summer) caused oxygen-depleted gas to flow from the dump into the shed (Phillip et al., 2008). Knowledge about these diurnal and seasonal changes is essential in assessing the atmosphere within a confined space. As well, changes occur within the dump as sulfide minerals continue to oxidize consuming oxygen from the pore gas, generating acid, and producing carbon dioxide from reaction of the acid with carbonate-type minerals. Temperature changes within the dump over time due to sulfide reactions lead to periods of danger followed by dormant behaviour and then followed again by danger over decades and possibly centuries. At a waste dump, the danger varies hourly as outside temperature changes from high in the daytime to low at night. Reclamation design factors (covers, slope changes, etc.) and Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 86 dump properties (e.g., sulfide content, water content, particle size, etc.) also play important roles  (Mohammadi and Meech, 2011). For example when the cover effectiveness is "High", gas flow into the dump is inhibited which decreases the internal temperature. This will happen for 1 to 3 years after cover placement when the cover is still young. For example, with White’s Dump at Rum Jungle mine in Australia, about one year after rehabilitation commenced, the internal temperature at 10 m depth dropped from 49 ºC to 44 ºC, and the oxygen- level in the pore gas declined to <1% (Harries and Ritchie, 1983). Based on Figure 3-4, it was concluded that an increase in outside temperature (or decrease in internal temperature) creates an unobvious and dangerous situation with respect to atmospheric risk at the bottom of a waste dump. Placement of a cover is a cost-effective method to reduce dump oxidation and control ARD generation since it reduces both air and water flows. When a cover becomes eroded, its effectiveness is lessened resulting in higher oxygen inflow to the dump leading to greater rates of sulfide oxidation and higher internal temperatures. Climate plays a significant role, for example – wet periods with a low evaporation to precipitation ratio can cause the cover to saturate with water increasing its effectiveness (Mohammadi and Meech, 2011); with White’s Dump, at the end of the wet season the internal temperature dropped 2 to 3 oC because of an increase in cover effectiveness due to saturation (Harries and Ritchie, 1983).  With a varying environment, one may not recognize an atmospheric danger using a multi- gas meter unless one is aware that certain apparently safe structures may become dangerous at another time. Continuous and regular measurement of oxygen levels is necessary or else a "false perception" of SAFE may result. This may be impractical and as an alternative, a risk assessment tool may prove helpful. Such a system should give designers and operators the knowledge of possible danger before an accident takes place. Recognizing a hazard using a single gas meter measurement is unrealistic as the danger may occur at an unknown point over an unknown time frame, maybe tonight, or next year, or next decade or Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 87 100 years from now – AFRA (Atmospheric Fuzzy Risk Assessment) can help evaluate and predict that uncertain risk (Mohammadi and Meech, 2011).  It would be easy to conclude that all sulfide waste dumps pose atmospheric confined space risks, but that is not the case. Some dumps show no effect, such as the North Dump at the Sullivan mine, while others that indicate similar problems have never had an accident because an element in the chain of effects is missing – either no emission, no concentration, or no human exposure. For example, with no covered pathway that connects a shed to bad air in the dump, the shed will likely be safe. Regulations are encouraged to require reclamation designers and operators to conduct an atmospheric risk assessment using AFRA and evaluate if a hazard is present at the beginning of the reclamation work and on a regular basis thereafter. AFRA suggest ways to overcome a hazard and remove it from the site. 5.3. Methodology 5.3.1. Fuzzy Logic Rulebase AFRA is an Expert System based on Fuzzy Logic. Fuzzy Logic is a methodology able to capture knowledge and perform approximate reasoning similar to how a human copes with uncertain problems when making a decision. Fuzzy Logic can provide structure to an ill- defined problem by gathering vague and dispersed information to input into the decision- making process. Fuzzy Logic in environmental risk assessment has been successfully applied in previous work (Meech and Veiga, 1995; Veiga and Meech, 1997, Ghomshei and Meech, 2000). A number of recent remediation systems have been developed: a system to deal with remediation practices at oil-contaminated sites (Geng et al., 2001); a fuzzy expert system to recommend remediation methods for ARD sites (Balcita et al., 1999); a fuzzy expert system to predict levels of Hg emissions and bioaccumulation risk from artisanal mine sites (Veiga and Meech, 1994; Veiga and Meech, 1995a/b; Veiga and Baker, 2004). These systems all deal with groundwater and soil contamination. AFRA recognizes confined space Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 88 atmospheric hazards for sulfide mine reclamation sites. Characterizing confined spaces situations using Fuzzy Logic rules has not been done before. Because of the inherent uncertainties involved in such space responses to temporal and environmental changes, this domain is likely to benefit from the use of Fuzzy Logic.  5.3.2. Other Confined Space Systems It appears that no system has yet been built to evaluate or predict an acute atmospheric hazard. OSHA developed a program called Confined Space Advisor to help with recognizing a confined space (Anonymous, 1997). In this program there is one part that can evaluate a certain work space and give guidelines. The questions that the program asks are simple:  "1- Is the space in question large enough that a worker can bodily enter?"  "2- Is the space configured so that a worker can enter to perform work inside?" "3- Does the space in question have limited or restricted means of entry and exit?"  "4- Is the space designed for continuous human occupancy?"  The explanation given for "limited or restricted entry and exit" is considerably different from that provided by NIOSH (Pettit and Linn, 1987) and is criticized as being inaccurate. Details of the NIOSH explanation was given in Section 2.3.2 In this program "limited or restricted entry and exit" is defined as "any place where an occupant must crawl, twist, be constrained in a narrow opening, follow a lengthy path or otherwise exert unusual effort to enter or leave, or the entrance may become sealed or secured against opening from inside" (Anonymous, 1997). But there are some cases where a very large enclosed structure such as a house or basement can become a confined space (according to the terminology used in this thesis) or a permit-required confined spaces (according to the terminology used by OSHA) (Jeana et al, 2004). This definition does not cover all the possible structures that may be confined spaces.  The program developed by OSHA was run for the sampling shed at Sullivan mine and it was unable to distinguish the sampling shed as a confined space with "limited entry and exit". Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 89 OSHA's system was designed for use in industries that include "agriculture services, oil and gas extraction, food products, tobacco products, textile mill, wood products, chemical products, petroleum refining, rubber products, and leather products". In each of these industries the typical confined spaces are covered and discussed. There is no mention of the mining industry in this program (Anonymous, 1997).  Workplace Hazardous Materials Information System (WHMIS) Canada has developed an electronic training course for confined spaces called "Confined Space Awareness". The course can certify employees after taking such training. The course includes confined space hazards, and pre-entry requirements for confined spaces (http://www.yowcanada.com/course _outlines_csr.asp?source=google&gclid=CIWatv-95qcCFQkFbAodxkJtbA). Recognition of a confined space in this course is based on confined space regulations in Alberta, where a space is categorized as "restricted space" or "confined space". Note that differences in defining "restricted space" and "confined space" consist of crisp answers that do not allow entry of any uncertainty. A Restricted Space is defined in Alberta as a place where the only hazard is getting into and out of the space. There is no need for a permit, for atmospheric testing, or for an attendant. A Restricted Space can become a Confined Space following a new hazard assessment by a competent person prior to entry in which changes in the conditions or the environment suggest additional hazards now exist. The relation between restricted space and confined space is similar to OSHA's definition of confined space and permit-required confined space, so if the place is partially or completely enclosed and not designed for human occupancy and has a limited or restricted entry or exit, it is termed a restricted space in Alberta, but if any type of hazard is added to the space then it is called a confined space. These types of definitions are very helpful and it is recommended to apply such definitions in Bc Mines Act and WorkSafeBC as well. The US Bureau of Mines developed the SPONCOM computer program to assess the risk of spontaneous combustion in a coal mine. This program is an expert system program that Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 90 uses geological conditions, coal properties, mining conditions, and mining practices to predict the spontaneous combustion risk. Data are based on information from the literature, mine operators and experts. The geological conditions such as density, coal joints, dikes and channels as well as mining conditions such as degree of floor heave and sloughage are ranked based on known ranking criteria for each property as low, moderate, and high. Based on the rank of each property, the likelihood of spontaneous combustion is output as low, moderate, or high (Smith et al., 1996).  Another program has been built to estimate gas explosion hazards in coal mines. The risk assessment model uses fuzzy mathematics to provide the hazard index of an accident in a coal mine based on a gas (methane, coal dust, or sulfides) explosion. The technique was examined for five faces of a coal mine in Chongqing, China to validate the risk assessment technique (Cao et al., 2006). These programs were designed to predict flammable atmospheres and none of them deal with the atmospheric hazard of confined spaces.  5.3.3. Atmospheric Fuzzy Risk Assessment (AFRA) A hazard is recognized by AFRA using both a general and detailed assessment of numerous variables observed and measured at a site. The main focus of AFRA is a detailed risk assessment of a sulfide waste dump with the aim to estimate risk due to oxygen-depleted gas emission. The general risk assessment done in AFRA can extend to other gas problems, such as nitrogen monoxide/dioxide and carbon monoxide/dioxide or methane in which conditions for occurrence of such gasses may exist. As such, AFRA can warn about atmospheric hazards from coal waste dumps or from blasting agents. The following sections describe the detailed and general risk assessment processes. AFRA does not account for other types of confined space dangers such as flooding or asphyxiation by collapse of the structure walls, although an extension into that knowledge could be done in the future. Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 91 AFRA is written within the Visual Basic.Net 2008 environment allowing customized design of the fuzzy system and making the software portable into different operating systems. Fuzzy logic-based expert systems allow new features to be added with ease by creating different and/or new membership functions, applying different fuzzy mathematics techniques, or by modifying, adding, or removing rules in the future as required. AFRA is flexible, robust and, unlike pre-made fuzzy expert system tools (such as Exsys (http://www.exsys.com/) and LPA VisiRule (http://www.lpa.co.uk/)); there are few restrictions on knowledge transfer since no proprietary software product is required. Visual Basic is one of the most commonly-used programming languages. It is user-friendly and compatible with all PC-based technologies. The system can function without all the input information and the User is not forced to enter data for each and every variable. There are only four pages of questions – although a further set of questions may be asked if some factors are undetermined. If measurements are available, they can be entered directly, but when not known, ranges or linguistic terms can be chosen instead. Selection of one range involves entering a Degree of Belief (DoB) to represent the certainty (or uncertainty) that the user holds about an entry. In a detailed risk assessment, the confined space entry risk is synthesized into four major elements: gas generation, gas emission, gas confinement, and human exposure as shown in Figure 5-1.  Fuzzy values of the first three elements give the likelihood of a hazard being present in a particular enclosed structure. By combining this with the Degree of Belief in human exposure, the risk of entering the space is calculated. Both a numerical and linguistic output is generated, although the relationship between the numerical value (on a scale from 10-4 to 1.0) and the linguistic term can be changed by the user if desired. Heuristics is used to model Unsafe and Safe situations at a waste dump. The main focus is on the first two elements, generation and emission but, the latter two are critical in creating conditions for an accident. The first two elements evaluate situations in which oxygen-Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 92 depleted air or carbon dioxide develops (gas generation) within the pore gas inside a waste dump and then transfers outside the dump (gas emission).        Figure 5-1. Stages in the atmospheric risk assessment   In conducting an environmental risk assessment of a dump, the generation and emission of Acid Rock Drainage (ARD) is evaluated. An analysis is made of the balancing of a dump's AP (acid potential) and NP (neutralization potential) of the minerals in the dump together with the impact on water effluent discharge quality. In AFRA, the process of oxygen-depletion of air (the first element of risk) and the process of oxygen displacement by carbon dioxide coupled with processes that emit such gasses (the second element of risk) are studied. Rather than NP and AP, permeability and the forces of thermal convection and advection that draw air into the dump and allow toxic gas to exit are of interest. Oxygen consumption is characterized by the oxidation rate of sulfides that depends on factors such as sulfide Step 1. Step 2. Step 3. Step 4. Step 5. Pathway properties (fractures, cracks, pipes, tunnels, shafts, diggings, or other surface disturbances) Dump permeability and age, Free energy Reactivity Generation of O2-depleted air Emission of O2-depleted air Concentration O2-depleted air Exposure of workers Risk of fatality Surface disturbances (erosion channel or pits) or structures built on surface such as shed or a house Wind, Height, Other sources, Reactivity, Effectiveness of cover, Dump permeability Free energy Failure to identify the space and worker entry to the space Sulfides, Surface Area, Air, Water Li ke lih o o d o f h az ar d Ex po su reChapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 93 content, water content, pore gas oxygen level, particle size, channeling, etc. Many factors also affect air flow in and pore gas flow out of the dump. Oxygen is depleted from air by reacting with sulfides leading to a temperature increase which in turn enhances convective transport into and pore gas flow out of the dump. The degree of oxygen-depletion depends on a balance between oxygen consumption and oxygen replenishment by convection. The set of variables that characterize these processes are identified by AFRA. Rules were created from knowledge of waste dump behaviour (specifically the Number One Shaft waste dump). The rules were derived from detailed study of waste dumps reported in the literature. Details of these rules with respect to gas flow and the characterization of toxic gas generation and emission can be found in Appendix E.  As Fuzzy Logic is a case-based reasoning approach, the rules attempt to predict oxygen- depletion at six reference dumps from the literature. The software was validated by applying AFRA to three dumps withheld from system development (test dumps). As more sites are tested, more combinations of variables that control atmospheric risk will be covered. The dumps selected for inclusion cover a spectrum of variation in key variables, however not all combinations are yet covered. The multi-dimensional graph in Figure 5-2 shows the levels of variation defined by these dumps. The dumps and their sources are: 1. Nordhalde Dump at the Ronnenburg Mine (U) in Germany                (Lefebvre et al., 2001(a); Wels et al., 2003; Smolensky et al., 1999),   2. South Waste Dump at the Doyon Mine (Au) in Quebec                (Lefebvre et al., 2001(a); Wels et al., 2003),   3. Sugar Shack South Dump at Questa Mine (Mo) in New Mexico                (Wels et al., 2003 and 2001; Lefebvre et al., 2001(b) and 2002;                  Shaw et al., 2002; Robertson GeoConsultants Inc., 2001)    4. Aitik Mine dump (Cu) in Sweden                            (Stromberg and Bawart, 1999/1994; Ritchie, 2003; Takala et al., 2001)  5. White’s Dump at the Rum Jungle mine (U) in Australia                (Harries and Ritchie, 1980, 1983, 1986, 1987; Ritchie, 2003),  6. Number One Shaft Waste Dump at the Sullivan mine (Pb/Zn)  Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 94               (Lahmira et al., 2009) The test dumps are:  7. Main Waste Dump at Equity Silver Mine (Au/Cu/Ag) in British Columbia                (Aziz and Ferguson, 1997; Lin, 2010)  8. West Lyell Dump at Mt. Lyell Mine (Cu) in Tasmania                (Garvie et al., 1997)  9. North Dump at the Sullivan mine (Pb/Zn)                 (Lahmira et al., 2009; Dawson et al., 2009)Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 95    Figure 5-2. Input data for reference waste dumps                        (fuzzy values are given in parenthesis). Koppen Climate, Reclamation Particle size Internal Temperature  (0C) O2 content (%) Average Sulfide  Content Age (yrs) Permeability (m2) Coarse FineIntermediate Moderately High to High LowModerate Extremely High LowModerately Low Extremely Low >15% >7% <15% <7% in summer LowExtremely High Middle Age Very Young Young Extremely Low Very High Very Low Moderately High G oing  fro m  center  to  o utw ard s Legend: Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 96 5.3.4. Transition from an Enclosed Structure to a Confined Space  Although an enclosed structure is not dangerous by itself, all such structures can become dangerous confined spaces when gas generation and emission are high. A logarithmic scale is used with linguistic fuzzy sets ranging from 10-4 (not a problem) to 1.0 (hazardous) to describe the risk – see Figure 5-3.     Figure 5-3. Fuzzy sets defining levels of risk for a confined space.  Transition from an enclosed structure to a confined space depends on the assessed risk. Different types of people respond differently to different risks, meaning that a space called Marginal Problem by AFRA may be hazardous to the public, but not to a First Responder whose job involves taking on risk – the very nature of their job must accept a higher level of risk. It is a subjective decision to define a linguistic threshold of acceptable risk for each type of person so a threshold level is requested of each User. Figure 5-3 shows the default setting of the linguistic risk threshold for entry of the public (Marginally Safe); for a paramedic (Problem); while spaces with higher risk (Significant Problem or greater) should be entered only when wearing an appropriate respirator. Gasses generated within a dump are not hazardous unless they are 0 20 40 60 80 100 0.0001 0.001 0.01 0.1 1 De gr ee  o f B el ie f Likelihood of Danger Marginally Hazardous Marginally Safe Hazardous Problem Marginal Problem Very Safe Safe Not a Problem Equipment RequiredFirst RespondersPublic Confined SpaceEnclosed StructureNo Structure Significant ProblemChapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 97 emitted from the dump and then confined within an enclosed structure – at which point the "enclosed structure" has definitely become a "confined space". When assessing atmospheric danger, i.e., the presence or absence of gas generation and gas emission, AFRA will convert an "enclosed structure" to a "confined space" or vice versa respectively. As a result, the atmospheric risk will change when elements are eliminated or added to the situation. 5.3.5. Fuzzy Rules Used Within AFRA Atmospheric risk reaches its highest value when the possibility of human exposure in the confined space is "high". Enclosed structures at a mine site that might become confined spaces are sheds, sumps, diggings, ditches, storage room or pump house, erosion channels, etc. all of which may confine an emitted gas. A site is considered SAFE when there is no concentration of gas emitted. Although a gas emission may exist, if it is quickly dispersed or diluted with ambient air, no significant hazard is present, but the situation may be problematic. The presence of sampling holes, ditches, coarse soil or rock, segregated rock sizes, cracks or channels, a shaft or well, a pipe, fractures around pipes, drains, or other surface disturbances can alter the emission rate. A site audit and review of past events is necessary to identify pathways that develop over time.  Assessing risk using heuristic formulae (or fuzzy if-then rules) provides a smooth and logical transition from a high to a low risk (or vice-versa) for a single input (or for a combination of fuzzy inputs). AFRA links the different ranges of each variable using piece-wise linear functions to model the multi-dimensional non-linear problem space. This makes the problem easier to understand and the model can be modified quickly as required for new analyses.  A number of fuzzy if-then rules are applied to map gas generation, gas emission, gas confinement, and human presence onto the output risk. Figure 5-4 shows the rules that link the four elements of risk to the final assessed value. For example: If enclosed structure is "Likely Present" and gas generation is "Moderate" and gas emission is "Small" and presence of people is "Low" Then confined space risk is "Marginal Problem" Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 98 Expressions such as "Low" or "Small" are called fuzzy values. The rule-base was initiated based on common sense and then was modified based on information gleaned from the literature on different dumps as well as discussions with a number of available experts. The knowledge gained was used to decide on the level of risk to match each dump with unique degrees of gas generation and emission. In calculating risk, a fuzzy risk value for each element is calculated using a weight factor for each input factor. The degrees of belief (DoBs) of these elements are combined to determine the overall risk. Appendix F contains details of questions asked by the system and typical answers as well as output for each of the nine dumps.  Figure 5-4 shows the complete set of rules that control dataflow from input to final conclusion. Minor variables affect major (or intermediate) variables and their respective risk element through each rule connection. General rules were developed that govern major variables such as reactivity, gas flow, and permeability. The Degree of Belief (DoBs) of the fuzzy value of “high” is used to characterize the major elements of risk. Each variable is described by a number of fuzzy sets to characterize different ranges (i.e., low, medium, and high). Then heuristic formulas and/or weighted inferencing are applied to derive the Degrees of Belief (DoBs) of major (or intermediate) variable from the minor factors. The weighted inference method calculates the DoB of a rule conclusion by the following formula proposed by Meech and Kumar, (1992) and applied by Veiga, (1994): DoB conclusion = g3533WiDoBin i=1    (5-1) where: DoBi = DoB of each factor Wi     = Importance of each factor (a value between 0.0 and 1.0) Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 99    Figure 5-4. Fuzzy Associated Memory Map for Confined Space Risk as a function of                          gas generation, gas emission, gas confinement, and exposure of humans. Risk Confinement Not Present Likely Present Present Gas Generation Low S S P Moderate MS P SP High P SP MH Risk Confinement Not Present Likely Present Present Gas Generation Low VS S MS Moderate S MS MP High MS MP P Risk Confinement Not Present Likely Present Present Gas Generation Low NP NP VS Moderate NP NP VS High VS VS S Risk Confinement Not Present Likely Present Present Gas Generation Low MS MP SP Moderate MP SP MH High SP MH H Risk Confinement Not Present Likely Present Present Gas Generation Low VS MS MP Moderate S MP P High MP P SP Risk Confinement Not Present Likely Present Present Gas Generation Low NP NP VS Moderate VS VS S High S S S Risk Confinement Not Present Likely Present Present Gas Generation Low MS P MH Moderate P MH H High MH H H Risk Confinement Not Present Likely Present Present Gas Generation Low S S MP Moderate MP P SP High P SP MH Risk Confinement Not Present Likely Present Present Gas Generation Low NP VS S Moderate S S MS High S MS MS H i g h           L o w                            N o n e P o s s i b i l i t y  f o r   P r e s e n c e  o f  P e o p l e Low                                                               Moderate                                            Large Gas Emissions Legend: H = Hazardous , MH = Marginal Hazard, SP= Significant Problem, P = Problem, MP = Marginal Problem, MS =  Marginally Safe , S= Safe ,VS= Very Safe NP = Not a Problem D o B   i n  t h e  P r e s e n c e  o f  P e o p l e  DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas Confinement DoB in Gas  Generation DoB in Gas  Generation DoB in Gas  r tion DoB in Gas  Generation DoB in Gas  neration DoB in Gas  Generation DoB in Gas  eneration DoB in Gas  eneration DoB in Gas  Generation DoB in Gas Generation Chapter 5 - Atmospheric Fuzzy Risk Assessment for Confined Spaces 100      Figure 5-5. Overall Flowchart of AFRA.Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 101 The major parts of a fuzzy rule-based expert system include Fuzzification (input fuzzy set definition), Inferencing (interpreting outputs from the inputs applying the rule-base) and Deffuizification (output fuzzy set deffinitions) units (Meech, 1999). The first unit is called Fuzzification wherein membership functions which are local descriptions of input variables are assigned a Degree of Belief (DoB). A membership function gives the DoB to a description of a discrete variable using linguistic terms such as "high", "medium", and "low". In the Inference unit, a conclusion is made from input facts and fuzzy rules. Fuzzy rules are expressions of the form - IF A THEN B - where A and B are labels of fuzzy sets each characterized by a designated membership function. The premise part of each rule partitions the input space into fuzzy regions that overlap with other rule-defined regions, while the conclusion or output part describes how the system behaves in each region. In the Inferencing unit, depending on the input values, some of the fuzzy rules stored in the rule-base will fire. After the inputs are fuzzified, the degree of belief (DoB or µ) for each part of the premise for each rule is defined on a scale from 0 to 1 (or 0 to 100). If the rule premise has more than one part (in the example in Figure 5-6, there are two parts), a fuzzy operator is applied to obtain the net degree of truth of the rule premise. The fuzzy operator for an OR conjunction is the maximum DoB in the premise parts while for an AND conjunction, the minimum DoB is chosen. In this example, the AND (minimum) conjunction is applied to calculate the net degree of truth - e.g., for the first rule this value is 20%, while for the second rule the value is 10%. The net degree of truth is then assigned to the fuzzy set of the rule conclusion to reshape it. This process is called implication. Implication is done by shortening the output (or conclusion) fuzzy set to the value of the corresponding net degree of truth of the rule premise multiplied by the variable weight. For implication, every rule has a weight (between 0.0 and 1.0). In the example in Figure 5-6, the weights for the rules have values of 1.0. The conclusion fuzzy set in this example is a Gaussian set - although often a fuzzy singleton is chosen for simplicity without any loss in accuracy.  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 102 If two or more rules with similar fuzzy set conclusions are active, the maximum DoB of these rules will be projected onto the fuzzy set in question. The output fuzzy set is the combination of conclusion fuzzy sets which are then defuzzified to a discrete value using a weighted average equation as shown in Figure 5-6. This method is not computationally intensive and produces results only slightly different from other techniques such as "area-centroid" or "mean of maximum".  If data is unavailable for a variable (such as wind velocity), fuzzy linguistic sets are chosen to show the relative velocity (on a scale from 0 to 10). These linguistic variables include heavy winds, light-but-frequent winds, and other terms in-between.  The weights of the effect of minor/major variables and the heuristic formulae used to assess each risk element are listed in Tables 5-1 through 5-6.   Figure 5-6. Two-rule fuzzy expert system (inferencing and defuzzification)      (Grima, 2000, adapted and changed by permission). First Rule: If dump is Young and sulfide content is Moderate then reactivity is Moderate Second Rule: If dump is Very young and sulfide content is Low then reactivity is Low Young Moderate Very young Low X=3 yr Age Age Sulfide Content Y=0.02 Moderate Low Sulfide Content Min Rule 1 Rule 2 Reactivity Weighted Average Defuzzification   z = μ 1+ μ 2 Reactivity z2 μ2 μ1 z1 μ μμμ μμ μ 80% ConclusionPremise AND (Min) operator Inputs Fuzzy Sets OR (Max) operator for rules with same output fuzzy sets μ1 z1+ μ 2z2 Min 20% 20% MinChapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 103 5.3.6. Detailed Risk Assessment Architecture In AFRA, a list of possible enclosed structures at a reclamation site is given to the User to choose one or many. Then, possible pathways that may connect the waste dump to each structure are entered by the User. Some elements of risk are specific for each structure (e.g., exposure and confinement), while other properties such as dump permeability, reactivity, internal temperature, cover properties are specific for the dump and so, are common for all structures depending on the degree of homogeneity of the dump (see Figure 5-7). A final confined space risk is determined for each enclosed structure. The output is given numerically and linguistically accompanied by recommendations and suggestions about future risk.    Figure 5-7. Structural architecture of AFRA.  The calculated risk is adapted to a final risk level using an adaptation factor (α) as in Equation 5-2 (Veiga, 1994): Structure Specific Properties Enclosed structure 1 Enclosed structure 2 . . . . Enclosed structure j Dump permeability Wind, Height Other sources Reactivity Effectiveness of cover Free energy Dump (Outside) Properties (Common for Structures) Average monthly temperature and pressure of the site Pathway 1 Pathway 2 . . . . Pathway i Gas Generation Future monthly risks Risk for enclosed structure 1 Risk for enclosed structure 2 . . . . . Risk for enclosed structure j Outputs Human exposure Gas Emission     each pathway Gas emission by massive convective flow (independent of pathway configuration) Dump ageChapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 104                                               Riskfinal = 100*(Riskinitial / 100)α   5-2 The value of the adaptation factor depends on input parameters such as observations of dead wild life around the enclosed structure, people entering the space showing signs of sickness, or people possessing a confined space entry permit, and future plans for worker protection. The adaptation factor is applied as an exponent to the initial estimated risk value. Adaptation provides considerable elasticity to the risk prediction. As shown in Figure 5-8, the adaptation factor (α) can vary logarithmically from 100 to a value approaching 0, with 1.0 representing no adaptation. When α approaches 0, even a small initial Degree of Belief is amplified significantly. When α approaches 100, even a small amount of uncertainly will move the final Degree of Belief close to 0. In AFRA, the value of α is restricted to a range from 0.6 to 20. If the system is satisfied that future plans and site observations show that a confined space is recognized and mitigation step implemented, the risk estimate is reduced. However, should AFRA believe that recognition of the problem is unclear, the final degree of belief in a "high" risk level is elevated.    Figure 5-8. Diagram showing different adaptation factors.   Adaptation Factor 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.2 0.4 0.6 0.8 1 Initial Degree of Belief Fi n al  De gr ee  o f B el ie f 6 4 2 1 0.6 0.08 0.2 0.4Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 105 The user is warned when the outside temperature lies above the internal temperature. If the outside temperature for all months of the year is not provided or available, the system can use site climate type according to Köppen’s climatic classification (Peel et al. 2007) to estimate the maximum outside temperature (climate type can also be used to verify humidity which helps to estimate changes in cover effectiveness). Given the maximum yearly outside temperature the extreme condition (highest risk) at the site is estimated. If gas emission at this extreme is "Low", there is minimal or no concern about site results for the rest of the year. The assessment results and input values are stored in an Excel spreadsheet as well as an ASCII file for future analysis or modification.   Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 106 5.3.7. Gas Emission Table 5-1 shows that the DoB in "high" gas emission depends on climatic changes (temperature and pressure) which determine the free energy difference across the dump surface, the dump configuration, the dump age, and the design of each pathway to the structure. The ∆f ranges in Table 4-1 are applied to interpret the fuzzy value for gas velocity at different seasons, although additional ranges are used to choose the weight for gas velocity values to estimate the gas generation and emission DoBs (see Table 5-1).  Depending on the pathway effectiveness to transfer gas outside, a value from 0 to 100 is assigned. This term is called PathDoB(i) - where i represents each pathway. When the dump is inhaling air, danger may still arise if O2-depleted water flows into the structure. This will deplete oxygen in the structure if the flow rate to air volume ratio is high enough and influx of fresh air is low (see Section 3.3.2) (Mohammadi and Meech, 2011). Gas emission DoB increases if sufficient effluent water runs through the pathway. The dump effluent DoB is represented by PathEffluentDoB(i). This DoB derives from flow data entered by the User from knowledge of ARD effluent samples which are generally collected monthly. The extent to which a pathway is covered will isolate water and air from the atmosphere before they reach the structure. This effect is represented by PathCoverDoB(i). Equation 5-3 combines these factors to calculate the effect of pathway properties on gas emission – the variable name is PathwayEmissDoB(i). PathwayEmissDoB(i) = Min(100, (PathDoB(i) + 0.2*(PathCoverDoB(i) +                                        PathEffluentDoB(i))))  5-3  As there may be more than one pathway, the Maximum of the various PathwayEmissDoBs is considered as the final value of the pathway influence on the DoB in "High" gas emission to each enclosed structure. The variable is MaxPathwayEmissDoB(j) (as in Equation 5-4), where j represents different enclosed structures at the site:         MaxPathwayEmissDoB (j) = Max (PathwayEmissDoB (i)), where i, j =0,…, n 5-4  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 107 The effect of pathway properties on gas emission together with other physical factors determines the final gas emission value. The overall gas emission to each enclosed structure (j) is determined by Equation 5-5. In Equation 5-5, "PermWeight" is the weight applied to permeability, "GasVelocityWeightforEmiss" is the weight applied to the effect of gas velocity direction, and "AgeWeightforEmiss" is the weight applied to the effect of the dump age on gas emission. The values for these weights are specified in Table 5-1.           DoBforGasEmissionn(j) = Max( 0, Min(100, (DoBperm * PermWeight +  5-5         DoB GasVelocityWeightforEmiss * GasVelocityWeightforEmiss + DoBAge *          AgeWeightforEmiss + MaxPathwayEmissDoB (j) * Weightfor MaxPathwayEmissDoB)))   Table 5-1. Weights to calculate DoB in "High" gas emission at dump bottom.  Inputs Fuzzy Values Range Weight Free energy difference between outside and inside of dump (J/g)  >44 0.95 31 to44 0.8 25 to 38 0.7 18 to 31 0.6 12 to 25 0.5 6 to18 0.4 3 to12 0.3 1 to 6 0.2 -2 to 3 0.18 2 to -3 0.1 -1 to -6 0.05 -3 to -12 -0.1 -6 to -18 -0.2 -12 to -25 -0.3 -18 to -31 -0.5 -25 to -37 -0.8 <-37 -1 DoB for Each Range of Permeability at Edges at the Bottom (m2) Very Low 1E-13 to 1E-12 -0.2 Low 1E-13 to 1E-11 -0.1 Moderate 1E-12 to 1E-10 0 Moderately high 1E-11 to 1E-9 0.1 High 1E-10 to 1E-8 0.2 Very High 1E-9 to 1E-8 0.25 Undetermined DoB in Table 5-2  0.25                       Age  <2 yrs -0.8 >2yrs 0.01 Maxpathway (%)  >70 0.2 <70 >50 0.1 <50 >20 -0.3 <20 -0.8 DoB for Gas Emission at the Bottom of the Dump Min ((DoB Free energy difference * W Free energy difference + …+ MaxPathway * W MaxPathway), 100)  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 108 5.3.8. Permeability The gas permeability of a waste dump is a difficult parameter to measure and/or estimate, yet it is fundamental in establishing the degree to which gas and air flows through the dump surface. Waste dumps generally consist of coarser material and large boulders with a range of permeability from as high as 10-9 to as low as 10-12 m2 (Ritchie, 1994), on the other hand, tailings dams consist of very fine material and have a lower permeability (10-11 to 10-13 m2). Permeability also varies across the dump surface, so measurements must be taken at various points to ensure a good average value is recorded. When a dump is covered with soils, clays, or glacial till, the objective is to create a less permeable surface to minimize air and water infiltration. Table 5-2 lists the factors considered important in estimating the DoB of "High" dump permeability.      Table 5-2. Weights to calculate DoB in "High" permeability   Factor Fuzzy Value Range Weight Method of dumping End dumping  0.3 Truck dumping  0.2 Push dumping  0.2 Use of dragline or bucket excavator  0.1 Percentage of coarse material  at the base of the dump (> gravel size) Undetermined 0% 0 Low 0-20% -0.5 Moderate 20-50% -0.2 High 50-80% 0.10  Very High   >80% 0.25 Water saturation High >0.75 -0.3 Moderate 0.4-0.75 -0.22 Low 0.2-0.4 -0.2 None 0-0.2 -0.1 If saturation level is unavailable then use mass water content High >10% -0.3 Moderate 5-10% -0.22 Low <5% -0.2 None <2% -0.1 Channeling  Present  0.08 Horizontal layering or prevention of high permeability channels Present  -0.3 Dump materials from processing plant Present  -0.08 Dump materials from blasting Present  0.08 Opencut mining Present  0.08 Underground mining Present  -0.70 DoB for "High" permeability Max(Min((DoB Method of dumping * W Method of dumping +…+           DoB Underground mining * W Underground mining, 100),0) Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 109 Air permeability is estimated heuristically in AFRA from variables such as: method of mining and dumping; percentage of coarse material at the dump toe, water saturation (or water content), channeling, and type of dump material. The weights shown in Table 5-2 were derived during validation of the model to fit results to the six reference dumps. 5.3.9. Gas Generation (Oxygen-Depletion) Table 5-3 indicates that the DoB in "High" gas generation depends on the cover, dump permeability, and material reactivity – all of which affect reactions that deplete O2 from air.  At the Number One Shaft waste dump, the oxygen concentration varies from 0% at the centre of the dump to a maximum of 20.6% at its edges (Phillip et al. 2008). Figure 5-9 shows the oxygen ranges and their corresponding fuzzy values.    Figure 5-9. Fuzzy representation of oxygen level.  An oxygen value below 19% is unacceptable for permanent human occupancy, so it is very likely that air blowing out of an effluent pipe at dumps like this one will exhibit an oxygen level below this value. When oxygen was blowing from the pipe at Number One Shaft dump, it was below 8% all the time (Figure 3-7) – definitely "oxygen-deficient". This occurred despite the fact that many boreholes showed "Marginal Low" oxygen concentrations at the dump edges because 0 20 40 60 80 100 10 12 14 16 18 20 22 24 Oxygen Level (%) D eg re e o f B el ie f     Oxygen Deficient Problem   Marginal  Low Enough Marginal High Oxygen Surplus Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 110 of a higher permeability at the bottom due to end-dumping together with the wedge shape of the dump making it shallower at the edges leading to higher oxygen contents. The weights in Table 5-3 were derived initially by common sense judgement and then modified to adapt to the expected behaviours of each of the six reference dumps. Similar to gas emission, the free energy difference between the inside and outside of the dump determined from temperature and pressure measurements also affects gas generation. In this case the effect of a negative free energy was reduced (lower weights than for gas emission) to account for a lower influx of air in winter versus the absolute value of the outflow in summer. Other variables of importance to gas generation include permeability at the edges of the dump, reactivity of materials, cover effectiveness, dump height, the presence of wind, and other sources of gas generation such as carbonates which create CO2. The reactivity weight trend may seem counter-intuitive showing a decline in generation as the material becomes more reactive. However, there are multiple effects of reactivity through the impact on internal temperature (free energy) as well as interactions with the presence of water and the quantity of sulfides, so this trend is designed to mute the final trend of reactivity on gas generation. A 1.5 m deep "low permeability" cover was placed on Number One Shaft waste dump, but this did not stop convective gas flow especially at the toe where the cover may help intensify gas influx, because of topography and the effluent pipe pathway. "High" permeability dumps with a low-permeability cover are likely more hazardous regarding oxygen-deficient air. Although the fuzzy value of hazardous gas generation in some "High" permeability dumps may be "Low", when gas is blowing out through a specific pathway, the danger may be "High". When the sulfide content is "Low" (0.015-0.02), and the waste dump has a "High" permeability, oxidation is inhibited by lack of sulphide, despite oxygen being widely available. In this situation, gas generation is "Low". If the site is located in an area of high winds, advection may be high for high elevation dumps (Wels et al., 2003) so more oxygen can enter and the oxygen level in the dump will be "Enough" to sustain oxidation in the central core.  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 111 The Number One Shaft waste dump has passed its neutralization stage, yet the CO2 level in gas blowing out of the dump measured as high as 5% which is an additional hazard in its own right. The amount of carbonate rock in the dump varies from 0.1 to 0.7%. In some boreholes, carbon dioxide varied from 4 to 10% during August and 0 to 5% during November 2007.  Table 5-3. Factors affecting "High" gas generation (oxygen-depletion within the dump)  Inputs Fuzzy value Ranges Weight Free energy difference between outside and inside of dump (j/g)  >44 0.95 31 to44 0.8 25 to 38 0.7 18 to 31 0.6 12 to 25 0.5 6 to18 0.4 3 to12 0.3 1 to 6 0.2 -2 to 3 0.1 2 to -3 0.08 -1 to -6 0.05 -3 to -12 -0.05 -6 to -18 -0.1 -12 to -25 -0.15 -18 to -31 -0.2 -25 to -37 -0.3 <-37 -0.4 DoB for each range of permeability at edges (%) Very Low 1E-13 to 1E-12 0.1 Low 1E-13 to 1E-11 0.05 Moderate 1E-12 to 1E-10 0 Moderately high 1E-11 to 1E-9 -0.15 High 1E-10 to 1E-8 -0.2 Very High 1E-9 to 1E-8 -0.25 Undetermined DoB in Table 5-2 -0.25 Cover effectiveness (%) (DoB in Table 5-5) Very High 80-100 0.6 High 60-80 0.3 Moderate 60-40 0.2 Low 40-20 0.1 None 0-20 0 DoB for reactivity (%) (DoB in Table 5-4) High >60 1 Moderate <40 1.4 Low <20 1.6 Height (m) Short <50 m 0 Moderately High 50-100 m -0.025 High 100-400 m -0.05 Very High >400 m -0.08 DoB for each range of wind (%) Not Windy 0 Light Winds -0.03 Heavy Winds -0.05 Frequent Heavy Storms -0.08 DoB other sources DoB of other sources of hazardous gas generation in the enclosed structure DoB for "High" oxygen-depleted gas generation from dump (%) Min(DoB Free energy difference * W Free energy difference + ... + DoB Wind* W Wind), 100) Total DoB for "High" gas generation (%) Max(DoB other sources, DoB for "High" oxygen-depleted gas generation, 100)   Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 112 5.3.10. Dump Reactivity Reactivity is a term that describes how rapidly oxygen is consumed from the air in a waste dump (Wels et al, 2003). Reactivity is a function of dump age, sulfide content, particle size, water saturation, and extent of weathering as listed in Table 5-4. The difference in oxygen  concentration at the slopes of a dump and its centre is due to high reactivity that causes the pore gas to become oxygen-deficient. If reactivity is low, oxygen will be brought to even deeper levels in the dump and if reactivity is not too small, the chance of heat being retained is high leading to high temperatures inside the dump (Wels et al, 2003). As the dump heats up, the influence of bacteria to sustain oxidation reactions by controlling the ratio of ferric to ferrous ions in water will dominate yielding higher temperatures with low oxygen levels in the pore gas.  Another factor affecting exposed sulfide is water infiltration which washes the surface of the rock removing reactant products. More important than infiltration, there must be at least 2% water content to meet the reaction requirement – below this level; oxidation virtually ceases (private conversation, Ward Wilson). While water is needed for oxidation, saturation is not required. When the fuzzy value of degree of saturation is "None", this has a very high weight in lowering the estimate of internal temperature. When the degree of saturation is well above the critical level of 2%, the DoB in a "high" internal temperature is unaffected - see Table 5-4.  Many of the factors affecting dump reactivity depend on the method of dump construction. For example, underground mining or dump relocation and resloping generates finer, more uniform waste rock with higher exposed sulfides which in turn leads to higher internal temperatures. Age affects the reactivity of a waste dump. Older dumps will be more reactive mainly because bacterial activity is established and the internal temperature has become elevated.    Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 113 Table 5-4. Factors affecting waste rock dump reactivity. Inputs Fuzzy Values Range Weight Sulfide content Extremely High >0.035 1 Very High >0.03 <=0.035 0.9 High >0.025 <=0.03 0.8 Moderate >0.02 <=0.025 0.7 Low >0.015 <=0.02 0.6 Very Low >0.01 <=0.015 0.2 Extremely Low <= 0.01 -0.2 None 0 -0.8 Percentage of exposed sulfide (ave. particle size) Low <20% -0.07 Moderate 20-50% 0.02 High >50% 0.03 Water saturation  (if not measured, use Water content) None <0.2 -1 Low 0.2< 0.02 Moderate 0.4< 0.02 High 0.75< 0.03 (OR) Water content None <2% -1 Low <5% 0.02 Moderate 5-10% 0.02 High >10% 0.03 Percentage of fine grained materials High  >20% finer than sand 0.08 Moderate >7 <20% 0.02 Low >2 <7% finer than sand 0 Very Low <2% finer than sand -0.01 Weathering or Slaking Highly weathered > 20% is weathered to fine 0.04 Slightly weathered >2% is weathered to fine 0.03 Not weathered < 2% is weathered to fine -0.02 Permeability     Very Low 1E-13 to 1E-12 0.3 Low 1E-13 to 1E-11 0.25 Moderate 1E-12 to 1E-10 0.2 Moderately High 1E-11 to 1E-9 0.15 High 1E-10 to 1E-8 0.1 Very High 1E-9 to 1E-8 0.05 Undetermined DoB for “High” permeability inTable 5-2 0.05 Age  > 181 yrs -0.13 171 - 193 yrs -0.1 162 - 181 yrs -0.05 150 - 171 yrs 0.005 137 - 162 yrs 0.04 121 - 150 yrs 0.1 106 - 137 yrs 0.18 92 - 121 yrs 0.2 78 - 106 yrs 0.18 71 - 92 yrs 0.1 64 -78 yrs 0.04 57 -71 yrs 0.005 45 - 64 yrs -0.05 30 - 57 yrs -0.055 15 - 45 yrs -0.06 4 -15 yrs -0.1 8 - 30 yrs -0.12 4 - 8 yrs -0.13 0 - 4 yrs -0.15 If percentage of coarse and fine materials in undetermined -Underground mining  0.08 If percentage of coarse and fine materials in undetermined -Open Pit mining -0.02 Resloping  Was done 0.04 Relocation Was done 0.05 Mining with blasting (as it affects exposed sulfide surface) - considered if permeability is undetermined -0.02 Waste material from processing plant  - considered if permeability is undetermined 0.08 DoB for ”High” reactivity  Min((DoB Sulfide content * W Sulfide content + ...+ DoB Material from processing plant * W Material from processing plant), 100) Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 114 5.3.11. Cover Effectiveness  The factors that affect the effectiveness of the cover to reduce permeability are listed along with their weights in Table 5-5. These factors include: type of cover; thickness of the growth medium; depth of established root system; age of the cover; evaporation:precipitation ratio; cover permeability; water infiltration rate; hydraulic conductivity of the compacted layer; slope angle; various defects in the cover (crusting, hotspots, etc.); thickness at the toe; freezing in winter; water management practices; and use of erosion-resistant rock. The more effective the cover at minimizing influx of water and air, unfortunately the more dangerous may be the atmospheric hazard. Best practice reclamation activities aim to minimize gas and water flow into the dump. Hence rehabilitation is done by covering a dump surface with clay and/or soil to reduce water infiltration; reshaping the slope to reduce erosion; and planting suitable growth material (revegetation). Temperature profiles measured within a covered waste dump shows that sealing can reduce and, in some extreme cases, stop pyrite oxidation (Harries and Ritchie, 1987). Placement of a cover is a cost-effective method to reduce dump oxidation and control ARD generation since it reduces both air and water flows. Lundgren, (2000) studied the effect of cover on oxygen concentration within two centuries-old waste dumps in Sweden. One was covered with about 0.5 m of glacial clay compacted in 3 lifts, while the second was covered with a concrete-type layer consisting of cement slurry with coal fly ash grouted into a 0.25 m thick layer of crushed aggregate. Before covering, the oxygen concentration of air in the dump was close to 21% throughout. Time-variations were small and related mainly to shifting wind direction and the presence of snow and ice, despite intense oxidation of sulfides taking place in the dump due to its very high permeability.  After covering, the oxygen concentration in both dumps immediately dropped to low levels remaining consistently at 0.5%. Some zones of higher concentrations were still seen indicating leakage paths resulting from cover imperfections around monitoring installations. A seasonal Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 115 variation was observed even in this cold climate with variations from 0 to 5%. The peak level occurs in winter even though the covers are saturated with moisture, showing that even when sealed with a "perfect" cover (i.e., cemented), convective flow is still active, although it is not enough to cause elevated oxygen content or "low" gas generation within the dump.  At the Number One Shaft waste dump, the oxygen concentration was "low" (0 to 14%) in August 2007. Covering this "high" permeability waste dump with "low" permeability material (glacial till) stopped air from entering and exiting freely by convection. The type of cover is the most important factor affecting cover effectiveness. These include: - store and release (or evapotranspiration) cover (highly effective if saturated) - capillary barrier cover (must maintain near-saturation conditions) - conventional or low hydraulic conductivity cover - clay or geosynthetic membranes - simple soil cover (cheapest, but least effective) - water cover (most effective, if water level > 2 m) - concrete-like cover (very effective, but can degrade or crack)  Oxygen ingress reduces exponentially with cover thickness (Wels and O’Kane, 2003). When a proper cover is installed, air will not flow freely into the dump even if a pipe is installed at the bottom of the waste dump – while it can flow into the dump, it cannot move too far as the seal limits outflow. However such a pipe can still act as a pathway for pore gas movement out of the dump when it exhales. Cover age is important. Aging of the cover leads to deterioration and loss in effectiveness. In mine sites subject to heavy rain fall, permeability will decrease after heavy rains retarding air inflow (private conversation, Ward Wilson). Crusting may occur in older waste dump (>70 yrs) which are more reactive, e.g., Sullivan North dump. Crusting may initially impede air flow, but as the crust cracks, enhanced infiltration may result.   High permeability waste dumps with a low-permeability cover are more hazardous regarding oxygen-deficiency since gas can flow out of the dump if it finds a pathway. The cover acts as a seal leading to oxygen-deficiency ("High" gas generation in the dump) although the internal temperature may not rise as much when there is no cover. So with the internal temperature Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 116 falling below the outside temperature, convection can move the gasses out of the dump from the base. So although covers are beneficial and cost-effective for rehabilitating a waste dump by reducing the flow of air and water, a "higher" atmospheric risk arises because of "high" gas generation and emission.    Table 5-5. Factors affecting cover effectiveness to control oxygen transfer  (adapted from O’Kane and Wels, 2003)  Inputs Range Characteristic Thickness Fuzzy value Weight Cover type  Store and release cover (evapotranspiration) Saturated > 2m  0.9 2-1 m  0.8 <1m  0.7 Partially saturated > 2m  0.7 2-1 m  0.6 <1m  0.5 Not saturated > 2m  0.5 2-1 m  0.4 <1m  0.3 Capillary barrier cover Completely saturated > 2m  0.3 1-2 m  0.2 <1m  0.1 Nearly saturated >2m  1.0 1-2 m  0.8 <1m  0.6 Partially saturated >2m  0.7 1-2 m  0.5 <1 m  0.4 Not saturated > 2m  0.5 1-2 m  0.3 <1m  0.1 Conventional low hydraulic conductivity cover Active clay saturated > 1.5 m  1 <1.5m  0.9 partially saturated > 1.5 m  0.8 <1.5m  0.7 not saturated > 1.5m  0.6 <1.5  0.5 Stable clay saturated >1.5m  0.8 <1.5 m  0.7 partially saturated >1.5 m  0.6 <1.5m  0.5 not saturated >1.5m  0.4 <1.5 m  0.3 Geo-membrane- >1.5m  0.9 <1.5 m  0.8 Simple soil cover  0.7 Water cover  1 Concrete-like cover  0.9 Root depth   Shallow -0.08   Intermediate -0.05   Deep 0.05 Thickness of growth medium   Thicker -0.08   Intermediate -0.05   Thinner 0.05 Cover age < 6 months  Very young 0.1 Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 117 Inputs Range Characteristic Thickness Fuzzy value Weight 6 months to 3 yrs  Young 0.06 2 to 6 yrs  Moderately young 0.03 4 to 8 yrs  Moderately old -0.06 6 to 12 yrs  Old -0.08 >12 yr  Very old -0.1 Evaporation /precipitation > 10%  - -0.02 >  5%  - -0.01 <  5%  - 0.05 Cover infiltration rate (Percentage of rainfall) < 5%  Low 0.3 5-10%  Moderate -0.1 >10%  High -0.4 Cover permeability (m2) <1E-12  Low 0.3 1E-10 to 1E-12  Moderate -0.1 >1E-10  High -0.4 Hydraulic conductivity of the compacted layer (m/s) <1E-9  Low 0.3 1E-9 to 1E-5  Moderate -0.1 >1E-5  High -0.4 Season Wet season (or wet season has just ended) 0.08 Dry season (or dry season has just ended) -0.08  Slope  <20 Low 0.08 20-40 Moderate 0.02 >40 High 0.01  How does the user rank the effectiveness of the cover?  None - 1 Low -0.2 Moderate 0.08 High 0.3 Crusting  Present 0.1 Defects in cover  Present -0.08 Hotspots  Present -0.08 Thicker cover at the toe  Present 0.08 Does the cover freeze?  Present 0.08 Water management solution, i.e. diversion channels  Present 0.08 Erosion resistant rock Adds to the life of the cover Present 0.08 Degree of Belief in "High" performance of cover Max(Min((DoB cover type * W cover type +...+ DoB Thicker cover at the toe * W Thicker cover at the toe), 100),0)  5.3.12. Internal Temperature Since internal temperature together with atmospheric temperature is the most important factor that drives pore gas out of a dump, it is necessary that AFRA be given data on internal dump temperature. At least one internal measurement is needed, although the position (height) in the dump is important since a temperature gradient will exist. If internal temperature measurements are not currently available, AFRA can still function and provide an estimate of risk by attempting to predict the internal temperature. Factors used by AFRA to calculate the internal temperature are listed along with their weights in Table 5-6. In addition to temperature, Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 118 pressure is also useful although the position (height in the dump) of temperature measurement can be used to predict the difference in pressure between the inside and outside of the dump. The central problem is to obtain internal temperature and pressure profiles. These can only be obtained by drilling at least one borehole through the dump. If the top pressure (PatmT) and bottom pressure (PatmB) are available, then pressure in the dump can be estimated from the relationship PDump= Average( PatmT, PatmB) for periods of time in which atmospheric pressure does not change abruptly since within pressure changes reach equilibrium in about one hour.  AFRA can infer the internal temperature using other factors. These include: position in the dump where average temperature conditions are considered to exist; reactivity of dump material; permeability at the edges of the dump; dump height and height:width ratio; presence of benches; presence of fumeroles; slope of dump sidewalls; cover age and effectiveness; effluent pH; and dump age. Some of these factors increase temperature and some decrease temperature as shown in Table 5-6. The output is given as a fuzzy value varying from "extremely high" to "extremely low" with each term covering a range of about 5 oC. The weights were adjusted to match reported measurements for the six reference dumps used to create AFRA. These validation and verification results are given in Section 5.5.  Table 5-6. Factors affecting internal temperature.  Inputs Effect Fuzzy Values Ranges Weight Level in the dump at which the internal temperature is preferred to be estimated Varies  >0.10 < 0.20 Height Reactivity * 0.92 >0.20 < 0.35 Height Reactivity * 1.00 >0.35 < 0.50 Height Reactivity * 0.92 >0.50 < 0.70 Height Reactivity * 0.85 Location in the dump where the   internal temperature is estimated Varies Edges  Reactivity * 1.20 Center  Reactivity * 1.00 Reactivity  (DoB from Table 5-4)   Positive              100 2 95-100 1.25 92-95 1.15 90-92 1.1 85-90 1.05 80-85 1 75-80           0.95 70-75 0.9 65-70 0.85 Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 119 Inputs Effect Fuzzy Values Ranges Weight 60-65 0.4 50-60 0.2 40-50 0.1 30-40 0.05 20-30 -0.1 10-20 -0.2 <10 -0.3 Permeability at edges Positive Very Low 1E-13 to 1E-12 -0.20 Low 1E-13 to 1E-11 -0.10 Moderate 1E-12 to 1E-10 0.00 Moderately high 1E-11 to 1E-9 0.10 High 1E-10 to 1E-8 0.20 Very High 1E-9 to 1E-8 0.25 Undetermined DoB from Table 5-2 0.25 Height Positive  <50 m -0.05 50-100 m 0.00 100-400 m 0.15 >400m 0.20 Benches Positive Present  0.1 Not present  -0.1 Fumaroles Positive Present  0.2 Height:width ratio (h/x) Negative x>>h (>7) 0.01 h=x about 1 0.0 X<h (<7) -0.01 Dump slope Positive Steep >30 0 0.10 Moderately Steep 20-30 0 0.05 Gentle <20 0 0.0 Cover age (when the cover is just placed (<1 yr) the cover doesn’t affect internal temperature. Effect is high when cover is young 1-5 yrs, When cover is > 7 yrs it loses effectiveness due to erosion) Factor (A) is multiplied by the weight for cover effectiveness <1yr A=0.20 1-3yr A=2.00 3-5yr A=1.80 5-7 yrs A=1.70 7-10 yrs A=0.90 >10yrs A=0.80 Cover and crusting  DoB for "High" effectiveness  - Table 5-5)  Negative None 0-20 0.000*A Extremely Low 20-30 0.125*A Very Low 30-40 0.070*A Low 40-50 0.050*A Moderate 50-60 -0.025*A High 60-70 -0.170*A Very High 70-100 -0.200*A pH (acidic) effluent Positive Low > 4 -0.10 Moderate 3-4 0.05 Highly acidic 2-3 0.15 Extremely acidic <2 0.20 DoB for internal temperature Max(Min ((DoB Reactivity * W Reactivity + ... + DoB pH* W pH ), 100),0) Estimation of internal temperature based on ranges of DoBs for internal temperature  Extremely High 95-100 >40 oC  Very High 90-95 35-40 oC  High 80-90 30-35 oC  Moderately High 70-80 25-30 oC  Moderate 60-70 20-25 oC  Moderately Low 50-60 15-20 oC  Low 30-50 10-15 oC  Very Low 15-30 5-10 oC  Extremely Low <15 2-5 oC     Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 120 5.3.13. Gas Confinement An emitted gas must be trapped in one place without being vented for the hazard to be realized. Any enclosed structure existing on the dump surface or in a working place that can accumulate oxygen-depleted air can create a dangerous condition. A sump, manhole, large pipe, space with internal baffles, surface depression, tank, shed, tunnel, well house, pump house, basement, storage room, steep surface, trench, and even an erosion channel (see Figure 5-10) are all possible sites to confine and accumulate gas at a mine reclamation site.   Figure 5-10. Erosion channel at a rehabilitation site  (Matsui et al., 2004, by permission).  Table 5-7. Initial degree of belief in confinement for different enclosed structures.  Structure DoB in Confinement Structure DoB in Confinement Large Pipe 100 Pump House 80 Place with Internal baffles 100 Basement 60 Depression on the ground 70 Storage Room 80 Tank 100 ARD settling pond 60 Sump 90 Room 50 Manhole 90 Steep surface 40 Shed 60 Trench 80 Tunnel 60 Erosion Channel 80 Well House 80 Depression on 80 Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 121 The initial degree of belief in whether these structures confine a gas is set by default using the values in Table 5-7. Gas Confinement is also affected by artificial ventilation or other devices used for gas control such as a U-tube on the effluent flow pathway. If there is no confined structure or the structures are not large enough for a human to enter, the hazard of gas emission from the surface is estimated. Surface emissions can pose a hazard to a person bending down or lying on the ground. Gasses emitted from a waste pile are quickly dispersed or vented at the surface over a short distance under most circumstances. Wind velocity, surface topography and geometry as well as temperature inversions can play important roles in dissipating or intensifying the danger however. Other factors, such as snow covering a structure and the presence of a door or window will also affect the Gas Confinement DoB. A number of rules are used to calculate the effect of these factors on effective ventilation of a confined space. Some of these rules are as follows: If 'Artificial ventilation' is None  and 'Window and door ventilation' is None  and 'Snow cover' is None  and 'Other devices' are None  then 'Effect of ventilation' is Very Low  If 'Artificial ventilation' is None  and 'Window and door ventilation' is Enough  and 'Snow cover' is Large  and 'Other devices' are Quite Effective  then 'Effect of ventilation' is High  If 'Artificial ventilation' is Moderately Enough  and 'Window and door ventilation' is None  and 'Snow cover' is None  and 'Other devices ' are None  then 'Effect of ventilation' is Low If 'Artificial ventilation' is Moderately Enough  and 'Window and door ventilation' is Enough  and 'Snow cover' is Large  and 'Other devices ' are None  then 'Effect of ventilation' is Moderate  If 'Artificial ventilation' is Enough  and 'Window and door ventilation' is Enough  and 'Snow cover' is Small  and 'Other devices' are None  then 'Effect of ventilation' is Very Low If 'Artificial ventilation' is Enough  and 'Window and door ventilation' is Enough  and 'Snow cover' is Large  and 'Other devices' are Quite Effective  then 'Effect of ventilation' is Very High  These rules together with others are combined to formulate an "alpha" factor similar to that shown in Equation 5-2 to adjust the initial DoB in Confinement.     Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 122 5.3.14. Human Exposure  People can be exposed to the atmosphere inside an enclosed structure either by entering the structure, or by extending their head into the structure, or by bending down or lying or falling on the ground. If a structure ( j ) is large enough for a human to enter, then the Degree of Belief in Human Exposure is calculated using Equation 5-7. If the structure is not large enough to enter, then the DoB of Human Exposure is only due to head exposure:  DoB HumanExpose( j ) = Max(DoB exposure by entering ( j ), DoB exposure                                           by entering in the future ( j ), DoB head exposure ( j ), 0) 5-7  5.4. Simplified Representation of the Rules in AFRA The following examples show pertinent output from AFRA; however, there are many additional sub-factors and rules that affect the risk estimate. All rules are active at all times in an analysis, albeit to different degrees. In these examples, permeability is assumed high enough (>1E-10 m2) to allow air or gas to flow freely in and out of the dump. 1. High reactivity (average age, high sulfide content, “no cover” or “ineffective cover” (age (>5 yrs) or eroded cover, high fine particles)  barb2right High internal temperature and higher internal temperature than outside barb2right Air flows in at the bottom edges while pore gas rises from the centre to the top of the dump (gas emission is low)  barb2right Oxygen content inside the dump is moderate because of a balance between high reactivity and gas inflow  barb2right A buried pathway connects the atmosphere in the dump to a shed  barb2right The shed has no open window and artificial ventilation  barb2right   Shed is in use by people barb2right Risk is low (e.g. A Problem) at this time as O2-depleted gas does not find a pathway to flow from the dump to the shed.  2. High sulfide content  barb2right High reactivity  barb2right High internal temperature and lower internal temperature than outside barb2right Gas sinks from top to centre and flows out of bottom edges (gas emission is high)  barb2right Oxygen content in dump is very low because of high reactivity and gas outflow Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 123 barb2right A buried pathway connects the atmosphere in the dump to a shed  barb2right The shed has no open window or artificial ventilation  barb2right   Shed is in use by people barb2right  Risk is very high (e.g. Hazardous) as highly O2-depleted gas blows into an unventilated confined space.   3. Young age (>2  >50 years) or Old age (>160 years), low sulfide content, effective cover (young age: 1-3 yrs) and few fine particles  barb2right Low reactivity  barb2right Low internal temperature and lower internal temperature than outside barb2right Gas sinks from top to centre and flows out bottom edges, negative gas velocity (gas emission is high)  barb2right Oxygen content in dump is moderately low because of balance between low reactivity that consumes less O2 and low gas inflow because of effective cover  barb2right A buried pathway connects the atmosphere in the dump to a shed  barb2right The shed has no open window or artificial ventilation  barb2right   Shed is in use by people barb2right Risk is high (e.g. Marginally Hazardous) since O2-depleted gas is blowing into an unventilated confined space.  4. Low sulfide content barb2right Low reactivity  barb2right  Low internal temperature, but higher internal temperature than outside barb2right  Gas rises to top from centre and air flows into dump bottom (gas emission is low)  barb2right  Oxygen content in dump is moderately high because of low reactivity of dump material and higher air inflow  barb2right  A buried pathway connects the atmosphere in the dump to a shed  barb2right  The shed has no open window or artificial ventilation  barb2right   Shed is in use by people barb2right Risk is very low (e.g. Marginal Problem) at this time since O2-depleted gas is not blowing into an unventilated confined space and oxygen-depleted gas generation is not high. 5.5. Validation and Verification of the AFRA Model  Comparisons of the estimated internal temperature range with actual measurements for each of the reference and test dumps are shown in Table 5-8. In each case the measured value falls within the range predicted by AFRA which shows that the system is able to estimate internal temperature of the nine waste dumps with considerable accuracy.  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 124 To compare the effects of gas generation and gas emission on risk at each of the dumps, the two other elements (gas confinement and human exposure) were each considered high at 100%. So it is assumed that an ARD collection sump exists at the bottom of each dump connected to the pore gas in the dump through a buried pipe and an underground drain.      Table F-1 in Appendix F shows the answers to question related to the enclosed structure (sump in this case), while Table F-2 demonstrates the answer to the questions related to properties of the pathway (pipe in this case). Table 5-8. Estimated and measured internal temperatures in reference and test dumps.  Dump Internal Temperature ( ºC)1 Estimated by AFRA Measured Nordhalde                      10-15                         14 Doyon                    >40                         45 Sugar Shack South                    >40                         40 Aitik Mine                       2-6                         0-3 White’s Dump                    >40                         44 (one year after cover) Sullivan No. 1 Shaft                      10-15                         12 (two years after cover) Equity Silver Main                    >40                         52 (four years after cover) West Lyell                      35-40                         38 Sullivan North                      30-35                         33 (eight years after cover)  1   For details see Table F-10 in Appendix F.  Table 5-9 is a summary of the key findings predicted for each of the reference and test dumps used to build AFRA. Four of the dumps (Doyon, Sugar Shack South, Aitik, and West Lyell) either have no cover in place or the cover is old and judged completely ineffective. Dump reactivity is judged to be 100% high for Doyon, White's Dump, and Sullivan North with the Number One Shaft dump having 78% belief in high. Gas generation is fully certain for the Nordhalde dump and for Number One Shaft dump. Both of these dumps show predicted gas velocities that are negative as does the Aitik dump. Although White's dump shows 100% gas generation belief, its emission belief is only 27%. The combination of these factors leads to risk values that are at least "A Problem" in all of the dumps. Number One Shaft dump and Nordhalde dump are judged to be "Hazardous" while Aitik, White's and North dumps are each a "Significant Problem". Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 125  Table 5-9. Summary of reference and test dump assessment assuming                                            100% confinement and 100% human exposure.  Dump Cover  effectiveness1 Dump reactivity2 Gas velocity3 Gas generation4 Gas emission5 Risk Value Linguistic Risk Term Nordhalde 74 48 Negative Small 100 63 0.53 Marginally  Hazardous Nordhalde (winter) 74 48 Positive Big 49 18 0.16  Problem  Doyon 0 100 Positive Big 31 15 0.15  Problem   Sugar Shack South 0 78 Positive Big 23 18 0.15  Problem   Aitik Mine 0 19 Negative Big 75 86 0.47 Significant Problem White’s Dump 100 100 Positive Very Small 100 27 0.33 Significant Problem Sullivan No. 1 Shaft 89 78 Negative Very Big 100 100 0.90  Hazardous  Sullivan No1.  Shaft (May)  89 76 Negative Big 100 73 0.60 Marginally Hazardous Equity Silver Main 100 70 Positive Big 64 18 0.16   Problem  West Lyell 0 83 Positive Big 36 18 0.16  Problem  Sullivan North  100 100 Positive Very Small 100 24 0.31 Significant Problem  1 For details see Table F-4 in Appendix F 2 For details see Table F-5 in Appendix F 3  Assuming a pipe and sampling shed exist – for details see Table F-6  4 For details see Table F-7 in Appendix F 5 For details see Table F-8 in Appendix F   The estimates of gas velocity direction during the warmest time of the summer for Number One Shaft dump and North dump at Sullivan mine show results comparable to the measurements. At Number One Shaft dump, when the outside temperature is 32ºC, the gas velocity is about -1 m/s). This agrees with AFRA’s estimation of gas velocity of "Negative Very Big" for an outside air temperature of 32ºC. According to Dawson et al., (2009), monitoring of the North dump seepage collection system has not shown any significant oxygen-depletion or Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 126 carbon dioxide-elevated gas emission in comparison to the Number One Shaft dump. This agrees with AFRA’s estimation of a gas velocity at the North dump of "Positive Very Small" for an outside temperature of 32oC.  There are no gas velocity measurements for any of the waste dumps in the literature and therefore it was not possible to compare gas velocity direction with real data for the remaining dumps. However by comparing AFRA’s risk assessment results with oxygen content measurements at these dumps, a useful verification of the magnitude of the estimated gas generation can be made. A comparison of seasonal changes in oxygen content can be used to infer if gas or air is flowing into or out of the dump respectively. Such a comparison reported by Smolensky et al., (1999) allows conclusions about the direction of convective gas/air flow in the Nordhalde dump. According to Smolensky et al., (1999), during the late autumn and early winter the oxygen profile in Nordhalde dump shows an increase in oxygen content, for a borehole near the edges of the dump. At this dump, AFRA estimated a gas velocity value at the bottom of the dump of "Positive Big" during the winter resulting in a gas emission belief of only 18%. During the summer, the gas velocity was estimated as "Negative Small", resulting in a gas emission belief of 63%. These results are in accord with oxygen levels in summer (0%) and winter (8%) in this dump (Smolensky et al., 1999). The results presented in Table 5-10 represent three possible scenarios for each of the reference and test dumps. The worst case scenario for all dumps involves the presence of a sampling shed connected through a buried pipe to an underground drain (similar to the situation in May 2006 at the Sullivan Mine Number One Shaft waste dump). Note that all the dumps are at least problematic should a sampling shed and pipe pathway be installed. The Nordhalde dump is "Marginally Hazardous" with the Aitik Mine dump, White's dump at Rum Jungle, and Sullivan North each classified as a "Significant Problem". As would be expected, the Sullivan Number One Shaft dump is classified as "Hazardous" with almost full certainty.  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 127 Table 5-10. Atmospheric Risk of reference and test dumps for different scenarios.   Dump Risk with 100% confinement  and  100% human exposure Risk with 100% confinement  and  0% human exposure  Risk with 0% confinement     And 0% exposure and no pathway Worst Case Moderate Case Best Case Nordhalde   0.53 (Marginal Hazard) 0.21 (Significant Problem)   0.02     (Marginal Problem) Doyon   0.15 (Problem) 0.06 (Marginal Problem)   0.0002   (No Problem) Sugar Shack South   0.15 (Problem) 0.07 (Marginal Problem)   0.00019 (No Problem) Aitik Mine   0.47 (Significant Problem) 0.21 (Significant Problem)   0.019     (Marginally Safe) White’s Dump   0.33 (Significant Problem) 0.11 (Problem)   0.009     (Marginally Safe) Sullivan No. 1 Shaft   0.90 (Hazardous) 0.50 (Marginally Hazardous)   0.05     (Marginal Problem) Equity Silver Main   0.16 (Problem) 0.06 (Marginal Problem)   0.008   (Safe) West Lyell   0.16 (Problem) 0.08 (Marginal Problem)   0.00018 (No Problem) Sullivan North   0.32 (Significant Problem) 0.13 (Problem)   0.012     (Marginally Safe)  The Moderate Case reflects continued use of the confined space as a sampling shed although proper confined space procedures are now in place and the shed can only be entered by someone wearing proper SCBA equipment with a person watching from outside the shed. The Number One Shaft waste dump is still considered hazardous, although the Nordhalde dump is now reduced to a Significant Problem. The Best Case scenario is where there is no enclosed structure connected to the dump, humans are not exposed to a possible hazard, and the pathway has been eliminated. In this situation, four of the dumps are now rated as No Problem, and three are Marginally Safe, while the Number One Shaft waste dump has been reduced to a Marginal Problem.  In fact, this is what has taken place at the Sullivan Number One Shaft waste dump. The shed has been removed and a U-tube has been built to prevent emission of gas through the pipeline. The reason the site is still considered a Marginal Problem by AFRA relates to the fact that an accident did indeed occur at this site. The results show that complete elimination of Emission and/or Confinement elements can switch a potentially dangerous problem into a potentially safe one.  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 128 5.6. Risk Assessment under Different Climate Conditions   It is important to examine the sensitivity of the risk assessment process to changes in climate conditions. The question asked is: how would the Number One Shaft waste dump behave if it was located in a different part of the world? Four scenarios have been examined using Köppen’s climatic classification system (Peel et al., 2007):   1. AF - tropical wet with no dry season 2. Bwh - Subtropical desert, low latitude 3. Cfc -Marine west coast, mild with no dry season, cool humid summer 4. Dfd - Subarctic - severe winter, no dry season, cool summer Table 5-11 shows the estimated risk under different climate conditions during the summer for entering a sump at the toe of the Sullivan Mine Number One Shaft waste dump in 2006, which is connected to the dump by a buried pipe and is in use by workers. In this assessment it is assumed that no internal temperature measurements are available and the evaporation to precipitation ratio and water saturation values of the cover are Undetermined. As can be seen, for an arid, hot climate (Arizona) and a very wet, hot climate (Brazil) the risk is assessed as "Hazardous". Without measurements, the temperate climate conditions generate a "Marginally Hazardous" risk. For the very cold, sub-arctic-like climate (Yukon), the risk of this situation is only a "Significant Problem". Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 129       Table 5-11. Estimated risk for different climates during the summer with no internal temperature measurement for                      entering an enclosed structure similar to that at the toe of Sullivan Mine Number One Shaft waste                      dump in 2006, which is connected to a dump by a buried pipe and in use by Workers.   Climate Type Atmospheric Temperature    in Summer (ºC) Gas Generation DoB Gas Emission DoB High Reactivity DoB Estimated Internal Temperature (ºC) Cover Effectiveness Direction of Gas Velocity Risk Very wet and hot: AF – Tropical wet climate with                      no dry season (i.e., Brazil) 31 100% 100% 76% 10-15 100% Negative  Very Big 0.896  – Hazardous Very dry and very hot:      Bwh – Subtropical desert,                  low latitude desert                 (i.e., Mexico, Arizona) 40 100% 100% 76%     15-20 a      90 % b Negative  Very Big 0.896  – Hazardous Temperate, moderately wet: Cfc – Marine west coast –mild with no dry season, cool humid      summer (i.e. Norway) 22 100% 79% 76% 10-15 100% Negative  Big 0.641 – Marginally      Hazardous Very cold and wet:   Dfd – Subarctic – severe winter,             no dry season, cool summer 8 100% 25% 76% 10-15 100% Positive    Very Small c 0.316  – Significant    Problem  a  Increase in internal temperature is due to reduced cover effectiveness b  Lower cover effectiveness due to lower saturation in a dry climate c  Despite the fact that the gas velocity is Positive Very Small, the hazard is still considered a Significant Problem to ensure a conservative approach.   Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 130 5.7. Cyclic Behaviour of a Reactive Waste Dump No atmospheric accident similar to that which occurred at the Sullivan Mine reclamation site has ever occurred anywhere else in the world. As such, the results of this project cannot be compared directly with another study since atmospheric investigations of waste dump behaviour is a relatively young field of research. Further studies of other waste dumps need to be conducted to provide more evidence about atmospheric risk. The industry appears to have been fortunate not to have seen a similar incident at another dump site. The following analysis describes our belief as to why other waste dumps have not shown an atmospheric hazard in the past. The reason may lie in the fact that there are several cyclic behaviours exhibited by a waste dump with respect to oxygen-depleted air being blown out the bottom of the dump into an associated confined space. These include: 1. Diurnal: Safe at night / Dangerous in day time Each day as the temperature cycles from hot in the daytime to cool at night, the dump may transition from blowing to sucking - this will occur if the maximum internal dump temperature lies between the daytime maximum and night time minimum temperature; 2. Seasonal: Safe in winter / Dangerous in summer In the summer, the minimum night time temperature may lie above that of the maximum internal dump temperature - in this case the dump will blow toxic gas throughout the entire day. On the other hand, during the winter, the maximum day time temperature may lie below the maximum internal dump temperature - in this case the dump will suck in air at the bottom throughout the entire day. 3. Decadal: Safe(r) when the maximum internal temperature has reached its maximum value / Dangerous when it is transitioning either up to or down from this value.  Initially, there is a low reaction rate of sulfides with oxygen so the pore gas is not depleted of oxygen and high convective flow is not yet established. But as the dump temperature rises due to internal heat generation from the oxidation reactions, more air is sucked in through convection Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 131 and the rate of reaction intensifies (especially as sulfiferous and ferriferous bacteria begin to accelerate the surface reactions) – the pore gas becomes depleted of oxygen and dangerous. As years pass, the maximum internal dump temperature continues to rise, perhaps climbing to a level above the maximum diurnal temperature in the summer. When this occurs, the dump will suck in air at the bottom all the time and no danger will exist in a confined space connected to the bottom of the dump. As the mineral surfaces continue to oxidize, eventually the sulfides approach the point of depletion and the maximum internal dump temperature will begin to fall. As it passes below the maximum diurnal temperature in summer, the dump will begin to exhale toxic gas at the bottom of the dump once again thus recreating the hazard. Eventually the reactions will stop altogether and the pore gas will no longer be depleted of oxygen, hence the danger is now gone forever. Exactly when each of these transitions occur is difficult to predict with any degree of accuracy, but it will depend on the sulfide content, the reactivity of the sulfides, the dump permeability, the flow of water through the dump, and atmospheric conditions that include temperature and pressure changes, among many other variables. Depending on the outside and internal temperatures, the danger can be conceptualized as follows (estimated for the Number One Shaft Waste Dump based on historical information): 0 - 10 years Initial period with rising danger 10 - 60 years Maximum danger - extremely hazardous 60 - 80 years Declining danger - transitioning from hazardous to a problem 80 - 150 years Constant reduced danger - internal temp > max. outside temp 150 - 170 years Rapid increase in risk - internal temp falls below max. outside temp 170 - 180 years Maximum danger returns - extremely hazardous 180 - 190 years Declining danger - transition from hazardous to safe (pore gas O2 levels rise) 190 - onward Site is safe - no oxygen-depleted gasses are generated or emitted Recognize that the temporal boundaries between these projected risk levels are fuzzy concepts which vary significantly by changes in the site setting and waste dump properties. Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 132 Currently many of the dumps that have been reviewed in this thesis appear to be at the stage of reduced danger due to their extremely high internal temperatures and so; most of these dumps do not show an atmospheric risk. Of course, this stage is followed by rapid increase in atmospheric risk if confinement and exposure exists. Dumps with high sulfide content may reach this stage as early as 20 years old. Figure 5-11 shows the idealized risk and internal temperature over time for low, typical, and highly reactive waste dumps. In this assessment, the internal temperature changes as the dump ages. The maximum outside temperature is assumed to be 32 oC each year to ensure the evaluation detects the maximum likelihood of risk in that year (although warmer conditions could prove more hazardous). Figure 5-11 (c), presents a conceptual graph of the decadal variations in maximum internal temperature and the corresponding risk for Number One Shaft dump. In this assessment, the internal temperature was estimated by varying the age of the dump and considering all the real dump properties as given in Appendix F and has been presented idealasticly. The Number One Shaft waste dump was about 66 years old when the accident took place in 2006, although this is really an estimate since the dump is very heterogeneous and was in use off and on over its life to closure; reaction rates may not have varied uniformly over the years as AFRA assumes. The maximum internal temperature was about 16 ºC in 2010 and from measurements taken between 2006 and 2010; this appears to be increasing at the rate of about 1.5 to 2 ºC per year. If this rate of increase continues, by 2030 the maximum internal dump temperature will probably reach about 36 oC, and remain at this steady state value for about 60 years (perhaps longer). In 2090, the reaction rate will begin to decline as the sulfides become depleted. The internal temperature is estimated to then begin dropping by about 3 ºC per year until a final equilibrium temperature of 10 ºC is reached at which point all atmospheric danger at the site may be gone. This might occur around 2140 or so.  In the early years (<5 yrs) the risk is a "Marginal Problem". At this stage, although gas generation is "None", the DoB in a "Low" gas emission is 81%. Gas emission is not "None" Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 133 since it is assumed the dump is connected through a buried ARD collection pipe to an enclosed structure at the bottom of the dump in which water flow is significant. If gas generation and emission were both "None" instead of one being "Low", the risk level would be "Very Safe". Here, a "Low" gas emission in combination with "High" values for confinement and human exposure gives a risk value much higher than "Very Safe". If the confined structure did not exist (which would yield "No" concentration and exposure) the risk would be "Not a Problem".  Between 5 to 50 years of age, oxidation increases and the pore gas oxygen level declines to a very low value. The internal temperature increases and convective air flow is established with internal temperatures below the outside temperature. As such, risk increases to "Hazardous". Between 50 to 80 years, the internal temperature continues rising until it exceeds the maximum reported outside temperature of 32 ºC causing the risk to decline to a "Significant Problem". From 80 to 150 years the internal temperature reaches its maximum (about 32 ºC) and flow reversal occurs year round – at this point, the risk is judged to be "A Problem". At 150 to 170 years, the internal temperature begins to drop due to a decrease in dump reactivity as the sulfides become depleted causing the risk to increase to "Hazardous" once again. From 170 to 190 yrs the danger starts to decline as the sulfides become depleted. At a very old age (>200 yrs) the sulfides in the dump are completely depleted and so, the pore gas is no longer oxygen- depleted, i.e., "No" toxic gas is generated. Although the internal temperature is lower than the maximum atmospheric temperature which may continue to induce air emission from the toe, the risk is a "Marginal Problem" since the pore gas oxygen level will have increased to that of normal air. It must also be understood that spatial differences in these transitions may occur at different times due to dump heterogeneities. Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 134    Figure 5-11. Idealized risk and internal dump temperature over time for various reactivities. It is assumed that heat is retained to an extent                         sufficient to eventually reach an internal temperature above the maximum outside temperature except for (a) which represents                                     the case of a slow rate of temperature increase to a level below the maximum outside temperature. 0 5 10 15 20 25 30 35 40 1 10 100 1000 T e m p e r a t u r e  ( 0 C ) Age of Dump (Years) 0.0001 0.0010 0.0100 0.1000 1.0000 1 10 100 1000 R i s k  V a l u e High reactivity Hazardous Marginally Hazardous Problem Marginal Problem Marginally Safe Safe Very Safe Not a Problem Significant Problem 0 5 10 15 20 25 30 35 40 1 10 100 1000 T e m p e r a t u r e  ( 0 C ) Age of Dump (Years) 0.0001 0.0010 0.0100 0.1000 1.0000 1 10 100 1000 R i s k  V a l u e Typical Hazardous Marginally Hazardous Problem Marginal Problem Marginally Safe Safe Very Safe Not a Problem Significant Problem (a) (b) (c) (d) 0 5 10 15 20 25 30 35 40 1 10 100 1000 T e m p e r a t u r e  ( 0 C ) Age of Dump (Years) 0.0001 0.0010 0.0100 0.1000 1.0000 1 10 100 1000 R i s k  V a l u e Low reactivity Hazardous Marginally Hazardous Problem Marginal Problem Marginally Safe Safe Very Safe Not a Problem Significant Problem 0 5 10 15 20 25 30 35 40 1 10 100 1000 T e m p e r a t u r e  ( 0 C ) Age of Dump (Years) 0.0001 0.0010 0.0100 0.1000 1.0000 1 10 100 1000 R i s k  V a l u e Low reactivity With Limited Temperature Increase Hazardous Marginally Hazardous Problem Marginal Problem Marginally Safe Safe Very Safe Not a Problem Significant Problem Atmospheric Risk of Confined Space Predicted Internal Temperature Maximum Outside Temperature r wi  limiting temperature increase  Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 135 Figure 5-11 (a) represents the risk for a dump with a low reactivity along with a limited temperature increase due to other factors such as a more effective cover and geometrical properties of the dump (e.g., gentle slopes and shorter height). In this dump, it may take longer (about 200 years) for the internal temperature to reach its maximum value of 25 ºC. If the reactivity of the dump is low but has a faster internal temperature increase, due to a lack of cover, then in the same amount of time, the internal temperature may increase to around 35 ºC  see Figure 5-11 (b). In a dump with a much higher reactivity, the internal temperature may reach its highest value speedily (about 10 years or less) - Figure 5-11 (d).  The yearly trend of the internal temperature can help in future risk analysis. A cooling dump initially becomes more hazardous while a heating one will eventually become safe for some considerable time. For the reference dumps this information is in Table F-6 in Appendix F.  A sensitivity analysis was done for Number One Shaft waste dump to show how the confined space risk value varies when the values for gas emission, gas generation, gas confinement and human exposure change from "Low" to "High" - the results are presented in Figure 5-12. Gas emission changes were studied by considering three different sets of internal temperatures. Gas generation, gas confinement, and human exposure have been tested at three levels [0% (Low) 50% (Moderate) and 98% (High)]. The outside temperature is assumed to be at the highest measurement reported for Kimberley B.C (about 32 ºC). In this table, gas emission varies for different internal temperatures: for T = 12 ºC, Emission belief = 100%, for T = 30 ºC, Emission belief = 51%, while for T = 38 ºC, Emission belief = 18 %. The results show that as the values of the four elements of risk vary from "Low" to "High", the level of risk changes smoothly from "Not a Problem" to "Hazardous". After discovering the existence of a risk, management measures should be implemented. These measures are learned from regulations, through other effective methods, or because of past accidents. The best practices may be passive or active ventilation or a confined space Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 136 entry procedure suggested for the site. The lowest level of risk management is simply placing warning information at the site.  The first step in managing risk is to ensure that any person who enters a dangerous confined space is aware of the possible problem and has the proper authority and equipment to survive should that problem actually exist - i.e., HUMAN EXPOSURE The second step in eliminating RISK is to eliminate the danger. In the case of a confined space this can be done in a number of ways. The space itself can be removed - i.e., CONCENTRATION; its connection to the source of the toxicity can be removed - i.e., PATHWAY. The ability of the toxic gas to enter a pathway can be eliminated - i.e., EMISSION. And finally, the source of the toxic gas can be eliminated - i.e., GENERATION. Any of these actions will eliminate the RISK of death. However, step 1 which ensures awareness and preparedness does not preclude a misunderstanding or a failure on the part of a person to follow proper procedures, so even if properly done, the possibility of death still exists to a certain degree. With step 2, each of the individual actions may not be completely possible. For example: - removing the confined space may not be practical if it serves a proper function - i.e., sampling ARD; the connection demands a pipeline for the ARD to flow, but it can be designed to avoid the collection of a toxic gas in the structure (use of a U-tube or venting pipe); controlling emission and generation involve complex issues related to the properties of the dump, its cover, and the associated environment. Achieving success with actions aimed at these elements is unlikely.  Once the RISK of an identified confined space at a reclamation site has been eliminated or minimized the analysis must move to possible emission and concentration of the toxic gas in an unrecognized confined space. This might include a surface depression, an erosion channel, a bore hole, a small excavation or trench, or one of many other types of pathways. The summer is more dangerous at the toe of the dump while in winter, emissions can occur on the upper walls or top surface of the dump. A further unrecognized state might take place at some point in the future, decades from the present time in which a house or structure is erected on the dump with Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 137 a basement or cellar or well-head because the memory or knowledge about the hazard has been forgotten or mislaid. It is for this latter reason that a regular atmospheric risk assessment must be done whenever a change in the environment or design of the site has taken place. Surface erosion or fallen trees or uprooted shrubs may create a new confined space situation that should be evaluated. The residual risk after implementing these changes should be assessed. After the design process, performance evaluations should be done to understand if the redesigned system meets the requirements to reduce or eliminate the risk. Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 138    Figure 5-12. Sensitivity analysis for various internal temperatures at No.1 Shaft Waste Dump assuming a maximum atmospheric                       temperature of 32 °C. Output in each rule box represents the risk value estimated by AFRA. Values are interpolated       when moving from one box to the next. The values shown are the risk at the centre position of each box. 0.047 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.078 0.243 0.464 30 0.023 0.103 0.157 38 0.012 0.041 0.066 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.040 0.172 0.240 30 0.015 0.091 0.103 38 0.008 0.055 0.059 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.001 0.027 0.072 30 0.0005 0 0.015 0.023 38 0.0003 0.008 0.012 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.236 0.422 0.706 30 0.103 0.199 0.2646 38 0.059 0.128 0.159 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.118 0.238 0.358 30 0.09 0.173 0.199 38 0.050 0.115 0.130 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.005 0.104 0.151 30 0.002 0.077 0.089 38 0.001 0.039 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. w 0.440 0.697 0.800 30 0.160 0.289 0.4386 38 0.084 0.163 0.228 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.164 0.384 0.697 30 0.09 0.180 0.257 38 0.054 0.115 0.159 Risk Confined Structure L (0%) M (50%) H (98%) Internal Temp. 12 0.007 0.163 0.449 30 0.002 0.077 0.130 38 0.001 0.041 0.066 H i g h  ( 9 8 % )                 M o d e r a t e  ( 5 0 % )                  L o w  ( 0 % )  P r e s e n c e  o f  P e o p l e Low (0%)                                                     Moderate (50%)                                           High (98%) Gas Generation Legend: H = Hazardous , MH = Marginal Hazard, SP= Significant Problem, P = Problem, MP = Marginal Problem, MS =  Marginally Safe , S= Safe ,VS= Very Safe NP = Not a Problem 12 MH HMSS MPNP VS SPP DoB in Gas Generation D o B  i n  t h e  P r e s e n c e  o f  P e o p l e  DoB in Gas Confin ment DoB in Gas Confinement DoB in Gas Confin ment DoB in Gas Confin ment DoB in Gas Confin ment DoB in Gas Confin ment DoB in Gas Confin ment DoB in Gas Confinement DoB in Gas Confin ment Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 139 5.8. General Atmospheric Fuzzy Risk Assessment For a general assessment regarding the nature of the waste materials, the presence of confined structures or human activities at the site, AFRA simply decides if there is an atmospheric risk to human health. The risk at this stage is assessed based on the degree of certainty that a confined structure or a source of hazardous gas exists (user defined); and the magnitude of danger based on possibility of death due to exposure (predefined for each hazard). A number of waste materials may be present that can create an atmospheric hazard at reclamation sitr. These are listed so a user can choose those that are known or considered to be present. For example sump or well head or drill holes, erosion channels, ditches, and/or drains, or pit or hole caused by trees falling on the waste dump surface can create confined structures in which hazardous gas can accumulate, water in  tailing pond/waste dump water is O2 deficient and can cause O2 deficiency, soil around maintenance facilities after closure is contaminated with oil and gas and can create atmospheric hazards in confined space nearby and any pit or manhole created in soils in a dangerous confined space also blasting activities agents or residues can create atmospheric hazards.  This part of the system is referred to as Source Identification and Recognition and is at the heart of the general risk assessment technique. Source investigation needs a good understanding about the kinds of reactions that may occur within different materials and the types of gasses that might be generated. To comprehend the sources, the literature for gas generation from soils and mined materials as well as previous confined space accidents and their regulations were reviewed. The Atmospheric Confined Space Manual based on Appendix C is embedded within AFRA. This involved classification of hazardous gas or enclosed structure presence into four groups: minerals and soils, organic materials, operations and activities, and other suspect places. Each group contains a list of materials in the form of hyperlinks. Gasses Chapter 5 – Atmospheric Fuzzy Risk Assessment for Confined Spaces 140 related to each are recognized from user input and further investigation and/or use of proper respirators is recommended.  5.9. Summary The major outcome from this research has been the creation of a fuzzy methodology to perform an atmospheric risk assessment at a waste dump reclamation site. The design of this “expert system” can help prevent an accident similar to the Sullivan tragedy if atmospheric risk assessment is made a prerequisite in conducting reclamation work. An effectiveness evaluation was done to verify that the system meets the requirements of risk assessment at six mine sites used as input to the development process. The system was validated against three sites that were not used as input. Prediction of internal temperature using a thermodynamic model has demonstrated success in matching measured temperature levels. Four risk elements (generation, emission, confinement, and human exposure) are considered the steps to generating a dangerous and potentially unrecognized confined space. The failure to recognize danger from a sampling shed connected directly to pore gasses at the base of a waste dump may have been due to the cyclic nature of the emission and confinement steps. These cycles occur on a daily, seasonal, and long-term (yearly) basis. The cyclic nature can mask the danger for many years after the dump begins to oxidize. There is currently no convenient analytical way to calculate sulfide oxidation rate in a dump (Lefebvre et al., 2001a) and so, field measurements are needed to provide valid data. However, AFRA can use inference equations to interpret an approximate value of reactivity to estimate the internal temperature range and apply this to the thermodynamic model to estimate gas flow and eventually evaluate the risk. Internal temperature depends on the degree of saturation (or water content), reactivity, permeability, cover effectiveness and age, dump age and geometry and pH. The system allows a User to enter values for these intermediate variables, or if not available, it will infer these values from other parameters.  141  Chapter 6  6. Compact Version of Confined Space Risk Assessment     During the Sullivan mine incident, two of the victims were paramedics. Their deaths were ultimately a failure to recognize they were responding to a confined space accident. An additional goal of this research was to develop an application for a hand-held Pocket PC to provide first-responders with immediate answers pertaining to a particular confined space situation. The application, called AFRA CE, was built in .Net Compact Framework V.2 and can run under Windows mobile 5.0. It can be converted to .Net Compact Framework V.3 or 3.5 in Visual Basic .Net 2008, so it can be run in Windows CE, Pocket PC 2003, Windows Mobile Professional, or Standard 6 OS. AFRA CE takes user input to assist in decision- making prior to entering an accident scene for the purpose of rescue. It is also a training tool for paramedics, fire-fighters, and police officers to become familiar with confined space situations so when they respond to an accident site, they understand the need for multi-gas meters and appropriate respirators. The application is only useful in assessing the presence of an atmospheric hazard. Rescuers are encouraged to learn about the cause of the accident to determine the equipment needed to conduct the rescue safely. Other dangers with confined space are listed in AFRA CE as part of a training manual but the system does not provide advice on other types of hazards. The risk assessment for the hand-held device is done by answering questions – each question being part of a hierarchy of questions to quickly affect the Degree of Belief about "entering" or "not entering" the space without proper safety equipment. A concise manual on confined spaces is embedded within the system to aid users in accessing information rapidly on confined spaces and the types of materials responsible for hazardous gas generation. A Chapter 6 – Compact Version of Confined Space Risk Assessment 142  definition of a confined space is provided along with a complete list of confined space examples. The system guides the user in understanding which type of hazardous gasses might be generated from different types of organic materials, minerals (including warning about waste dumps), and from activities such as welding, etc. However, no claim is made that AFRA CE covers all possible atmospheric hazards in all industrial workplaces. AFRA CE is only designed to help with atmospheric hazards and as such, it is assumed that the rescuer is responding to a situation involves an enclosed structure. Since the enclosed structure is present (Concentration) and the rescuer is about to become exposed to it (Human Exposure), only the Degree of Belief in the presence of an atmospheric hazard in the enclosed structure is important in deciding to  "enter" or "not enter" the space.  Figure 6-1 shows the flowchart of AFRA CE, in which inputs are combined to calculate the DoB to not enter the space without SCBA. Appropriate advice is provided for each situation to generate an awareness of the risk to the rescuer. Table 6-1 shows how the initial User Belief (DoBi) in the presence of an atmospheric hazard can change according to factors that indicate the possibility of a problem. After calculating the final DoB in the presence of an atmospheric hazard from the user’s answers to questions such as: "Has the accident happened recently?" and "Is the victim obviously still alive?", the DoB in "not entering" the space is calculated – see Table 6-2. In all cases, if a Self-Contained Breathing Apparatus (SCBA) or air–supplied respirator is available, the rescuer should use it. If such equipment is not available, then depending on the DoB certainty level to not enter the space, the advice given is: Advice 1:   0-20 - Enter,   Advice 2:  20-40 - Maybe okay to enter,  Advice 3:  40-70 - Maybe better not to enter,  Advice 4:  70-90 - Better not to enter,   Advice 5: 90-100 - Do not enter.  Chapter 6 – Compact Version of Confined Space Risk Assessment 143   Note that the victim’s condition (dead or alive) has a double effect on the decision because not only does it affect the DoB in the presence of a hazard, it also affects the DoB to not enter the space. This is because in many confined space accidents, the victim dies quickly after exposure to the hazard. So if the victim is dead, there is no point to immediately enter the space without the correct equipment. As well, if the accident happened more than 20 minutes before arrival, then rushing into the enclosed structure for rescue without SCBA is not wise since the victim is likely already dead. Note that in all rescue attempts, the safety and life of a rescuer is considered more important than that of the victim. Rescuers should not do something that places themselves in jeopardy especially if the action is unnecessary. As a result, when uncertainty exists in the decision to enter or not enter the space, the answer is leaning more towards recommending that the space not be entered, e.g., Advice 3 and 4. Appendix G shows the results of testing AFRA CE for certain situations. In all examples, it is assumed that the user has already answered the initial questions as follows: 1. Is it obvious that a non-atmospheric effect is the cause of this accident? (Ans. No or Uncertain) 2. Is the accident definitely due to an atmospheric hazard? (Ans. Uncertain - DoB varies for each instance as shown in the examples in Appendix G) 3. Has another rescuer lost consciousness? (Ans. No) 4. Do you have SCBA? (Ans. No) With Advice 1 to 4, if the rescuer has decided to enter, he/she should wear a full body harness attached to a life line. At least one additional person (attendant) should stand outside and ensure the rescuer is responding frequently during the rescue. If the attendant does not hear a response from the rescuer, the rescuer should be pulled out of the space immediately using the life line. In decision 5, it is recommended that the rescuers wait until they have access to the proper equipment before entering the space. Firefighters are usually equipped Chapter 6 – Compact Version of Confined Space Risk Assessment 144  with oxygen tanks and other necessary tools, but might arrive on the accident scene after police or paramedics. Therefore, it is recommended for police officers and paramedics to wait until firefighters arrive. Although this decision may not save the victim (if the victim was alive on first arrival), the rescuer will not lose her or his own life by rushing (perhaps needlessly) into the space without proper equipment.  Figure 6-1. Overall Flowchart of AFRA CE.  The examples shown in Appendix G indicate that the initial DoB in an atmospheric hazard expressed by a User may be strengthened or weakened depending upon the answers to the simple questions presented to the user. Each question and answer has its own unique weight to affect the final DoB in either a positive or negative way. Chapter 6 – Compact Version of Confined Space Risk Assessment 145  In all cases, the rescuer should bear in mind that they do not have to follow the Advice given by AFRA CE as it is their own responsibility to make the final decision. The software only provides logical suggestions based on user input. The developers of this system do not accept responsibility for any consequences resulting from the use/misuse of AFRA CE.    Table 6-1. Factors affecting the DoB in the presence of an atmospheric hazard.  Factor  Answer Weight  Victim condition  Alive (100)  -0.2  Dead (0)  0.3 * (100 – Ans.)  Proper ventilation  Yes (100)  -0.3  No (0)  0.1 * (100 – Ans.) Water present  Yes (100)  0.2  No (0)  -0.05 * (100 – Ans.) Organic material present  Yes (100)  0.2  No (0)  -0.05 * (100 – Ans.) DoBatm hazard Min(Max((DoBinitial + ∑Weighti * Answeri ),0), 100)       DoBintial is the intial DoB in an atmospheric hazard as expressed by the user.     Table 6-2. Factors affecting the DoB in not entering the space without SCBA.  Factor  Answer Weight  Has the accident occurred recently (within last 20 minutes)?  Yes, or do not know (100)  0  No (0)  0.1* (100 – Ans.)  Victim condition  Alive (100)  -0.1  Dead (0)  0.2* (100 – Ans.)  DoB in not entering the space without SCBA  Min(Max((DoBatm hazard + ∑Weighti * Answeri ),0), 100)     146  Chapter 7  7. Concluding Chapter   7.1. Conclusion A detailed analysis of the Sullivan Mine confined space accident was done in Chapter 3 to recognize the direct and indirect factors contributing to the accident. The accident was compared with other similar confined space accidents, where the temporal changes of temperature and pressure caused the "bad" air to flow in to the confined structure. All of the analysis was conducted with a view to creating a tool to assist in the design of mine reclamation programs that can recognize the factors that contributed to this hazard in order to prevent future accidents. Climatic pressure and temperature changes can affect the atmospheric condition of a confined space. As such, a confined space can switch from being safe in one instance to being unsafe in another. This effect has been only mentioned in few confined space regulations. In Chapter 4, a thermodynamic model was built based on laws of thermodynamic that calculates the free energy change based on temperature and pressure differences inside/outside of the dump. Gas emission into an enclosed area at different times of the day and year can be predicted given the change in delta free energy. Such a model has never been built for confined spaces and perhaps can be applied at other confined space situations such as breathing coal mines. This methodology is helpful at such places where it is not safe to simply rely on a random gas measurement due to the fact that in a few hours, days or months the confined space risk can transform from "safe" to "hazardous". In Chapter 4, the effect of pressure differences on gas flow has been studied in Number One Shaft Waste Dump. This effect at different waste dumps has been reviewed in the literature. Pressure difference in waste dumps is dependent on height and permeability and often times is of a short term nature. Pressure plays a role in dumps but to Chapter 7 – Concluding Chapter and Future Work 147  a lesser extent than in "breathing water wells" and other underground soils/rock. Cover plays a significant role in pressure effect on gas flow in waste dumps. In Chapter 5 it was shown that confined space problems at reclamation sites involve exposure of humans to oxygen-depleted air usually accompanied by high levels of carbon dioxide. This toxic gas danger results from a chain of mechanisms that occur in sequence leading to the death of virtually any and all exposed humans. These stages are Gas Generation within the dump, Gas Emission from the dump, Gas Confinement (or Concentration) in a confined space, and Human Exposure in the confined space. The risk of a confined space accident depends on the magnitude of these four elements which interact and affect each other. Gas Generation (oxygen-depleted air) occurs in virtually all sulfide-bearing waste dumps through reaction of sulfide minerals (pyrite and others) with oxygen dissolved in water and other ion species derived from early reactions. The water only needs to percolate through the dump at relatively low rates for these reactions to be sustained. The source of oxygen derives from influx of air into the pore volume which is controlled by diffusion at the centre of the dump and by advection/convection near its edges. As the reaction proceeds, air and water become depleted of oxygen and the internal temperature rises drawing in air at higher flow rates. Convection may actually reach the central regions. Gas Emission depends on many factors. Air may be drawn in at the top of the dump and toxic gas emitted through the toe (usually quickly diluted and dispersed) when the dump’s internal temperature is below that of the atmosphere. This direction is reversed when the atmosphere is cooler than the dump (i.e., at night perhaps, and in the winter months). On these occasions, air is drawn in at the toe and emitted from the top of the dump. Difference in free energy between inside and outside of the dump is recognized to be correlated with direction of air movement. Of course a direct connection is needed to cause the oxygen-depleted air or carbon dioxide to an enclosed structure. For example in Sullivan Mine tragedy the sampling Chapter 7 – Concluding Chapter and Future Work 148  shed was safe before covering the toe-drain which created a direct connection between air in the shed and air in the dump.  Gas Confinement can take place in an enclosed structure near to the point of emission which could be a pipe or a surface depression. If the accumulation remains undiluted, a confined space hazard exists that must be permit-required for entry. As well, oxygen-depleted water can be emitted through a buried conduit which can also act to deplete oxygen from the air in an attached confined space that it enters. Human Exposure to the toxic gas can occur in a number of ways. In the case of the Sullivan Mine tragedy, gas accumulation took place inside a shed covering a sump installed to collect samples of the Acid Rock Drainage every month. Sampling had been done in this shed for about five years prior to the accident and entry into the shed occurred without incident two weeks prior to the accident. Failure to identify the danger is central to all types of confined space accidents – in this case, this failure occurred by the site operators as well as the individuals who entered the shed and died. In Chapter 5 it was shown that certain reclamation activities that are implemented to protect the aquatic environment from Acid Rock Drainage (ARD) and Metal Pollution actually enhance the effects of one or more of these mechanisms. These activities include: installation of a cover to seal the dump to reduce infiltration of water and air; conversion of an open ARD collection ditch to an underground drain; installation of a sampling shed to monitor the ARD effluent; and creation of a hydraulic connection between the sampling shed and pore gas in the dump. In Chapter 5 an atmospheric fuzzy risk assessment tool (AFRA) has been devised to assist in recognizing atmospheric hazards of confined spaces at a mine reclamation site before the danger can cause death. The hazard can then be mitigated either through redesign of the reclamation practices or by adhering to proper confined space entry procedures (permit- required, proper signage, oxygen-meter, and use of proper respirators, etc.). AFRA is a fuzzy- logic rule-based expert system developed from three sources: from knowledge gained from the Chapter 7 – Concluding Chapter and Future Work 149  Sullivan Mine accident; from atmospheric emission studies on other waste dumps reported in the literature; and from discussions with recognized experts in the field. Fuzzy logic provides a way to conduct a dialog between a user and the system using linguistic terminology and to report the findings in a common sense way that is easy to understand and adjust. It is considered that a non-numeric method is more likely to be accepted for use by mine personnel. The model was verified against six waste dumps used to derive the rules of thumb within the system. Another three dump reports not seen before creation of the system were used to validate the system’s ability to predict behaviour of previously unseen dumps. Verification and validation of AFRA shows excellent agreement with actual measurements at the chosen dumps. In Appendix C, possible gasses that can be emitted from mine wastes and soils especially at mine reclamation sites as well as previous atmospheric related confined space accidents were reviewed to develop an Atmospheric Confined Space Manual (Especially for Gasses from Mines and Soils). This manual which contains important parts of existing confined space regulations is very complete in helping people with recognizing the danger at workplaces especially at mine sites. The electronic version of the manual is embedded within AFRA as under the category called "A list of possible gasses in confined spaces" and is a complete reference which relates possible gasses to different sources, designs or activities at different sites. Although society and industry seems to have a good understanding of the reasons for different types of confined space accidents, these tragedies continue to occur. A key factor in preventing such accidents is sharing knowledge about how elements of each situation combine to result in a hazard. Investigating and focusing on the causes of the accident although useful, is insufficient. The goal must apply knowledge in a way to prevent future accidents at the same site as well as elsewhere. Otherwise the knowledge is simply stored in books and reports as plain statistics. AFRA can transfer lessons from the Sullivan Mine accident in regard to confined space atmospheric hazards and apply this knowledge to assess risk at other dump sites. AFRA recognizes confined space atmospheric hazards for sulfide mine reclamation sites. This is while Chapter 7 – Concluding Chapter and Future Work 150  it appears that no system has been done yet to assess the acute atmospheric problems especially at mine sites. Regarding these issues, changes have been suggested by the Technical Panel of the Sullivan Mine Incident to eliminate and manage atmospheric hazards at the Number One Shaft Waste Dump site, e.g., placement of a cover to seal the oxygen-depleted gas, and decoupling enclosed structures from dump pore gas. A U-tube was installed on the buried pipe and the sampling shed has been removed from the site.  AFRA is a tool that can help people understand why the atmosphere within a confined space can change from safe to hazardous in a matter of minutes, but it is important to take multiple gas measurements at different times of the day and year. Since accidents such as this one have never happened before in an ARD sampling shed, designers and operators of reclamation sites are unlikely to have this problem or its causes "front of mind". When a structure has been entered safely without any problem on numerous occasions, it is unlikely that someone will consider assessing the oxygen level in the air within the space. AFRA is able to predict an oxygen-depletion hazard even when the space is currently safe. It can offer advice about design changes or climatic changes that may convert an enclosed structure into a dangerous confined space. This will force users to think about possible problems before it is too late - things that no one considers important unless warned beforehand.  7.1.1. Contributions The major contributions of the thesis are: • Applying thermodynamic laws to understand the combined effects of climatic pressure and temperature changes on the atmospheric risk of an enclosed structure near a waste dump. • Recognizing oxygen-deficient water as a source that can cause oxygen-depletion in an enclosed structure near a waste dump. Chapter 7 – Concluding Chapter and Future Work 151  • Developing and testing an atmospheric fuzzy risk assessment tool (AFRA) based on fuzzy logic to recognize atmospheric risk at an enclosed structure around a sulfide waste dump. • Providing a possible explanation as to why an accident similar to that which occurred at the Sullivan Mine has not been seen elsewhere (see Section 5-7). • Recognizing that certain reclamation activities potentially increase the atmospheric risk. 7.2. Recommendations  The terminology used in the field of confined spaces is extremely confusing, varying by jurisdiction and agency. In Chapter 5, this thesis suggests a simplified approach to improve understanding. Any structure that is enclosed should be deemed an enclosed structure. Any enclosed structure that has the potential to become hazardous should be designated a confined space that requires a permit to control entry. Two terms can encompass all situations: "enclosed structure"; and "confined space". The former is safe, the latter is hazardous. The transition from one term to the other is the central focus of risk assessment procedures such as that performed by AFRA.  During the conduct of this research, the Mines Act (2008) regulations were revised by the Ministry of Energy, Mines and Petroleum Resources – two years after the accident. Yet there is still no mention of atmospheric hazards at mine reclamation sites in the confined space section of this Act. There is insufficient definition and regulations about confined spaces in the BC Mines Act. In fact there is no specification of the possibility of oxygen-deficiency at sulfide mine reclamation sites in the BC Mines Act and any mining-related regulation throughout the world (Mohammadi and Meech, 2011). This oversight must be addressed so mine operators and mine inspectors are informed about this potential hazard that may exist during the operating phase of a mine and following closure. Few confined space regulations acknowledge that a change in atmospheric temperature or pressure can transform a safe enclosed structure into a deadly confined space. This issue Chapter 7 – Concluding Chapter and Future Work 152  should be stressed in all confined space regulations, notifications, and brochures. As well  gasses from remediated mines can create a hazard either at the site or by travelling through the ground into people’s homes or any other enclosed structure. The thesis draws the attention of society to the existence of these types of hazards. The possible pathways for such gas migrations were reviewed in Appendix C. Many houses are built near remediated coal mines, others may be built near remediated mines in the future. Assessment should be done to examine soil surfaces, water, and enclosed areas (nearby houses and yards) for the presence of methane, carbon dioxide, and oxygen-depleted gas. If any hazardous gas is present then its extent and sources should be assessed and if necessary, it should be managed (Mohammadi and Meech, 2011). A brief analysis of previous confined space accidents that included their detailed cause and location has been done by a few organisations such as NIOSH. Such a resource can help to explain which types of atmospheric problems are possible in certain workplace based on past experience. This is important in preventing future confined space accidents. Finally having one single regulation for confined spaces for all industries can result in omission of site-specific atmospheric hazards. A customized confined space regulation is needed for each industry that should be accompanied by an specific risk assessment methodology. AFRA is a  risk assessment tool specific for mine reclamation sites. AFRA is available free-of-charge through a UBC web site (www.mining.ubc.ca/AFRA). It can be downloaded for use by mining companies, government agencies, and/or safety professionals. It should be a requirement to use this tool to evaluate reclamation practices at all mine sites and a reevaluation of risk should be mandated whenever a change occurs in the environment, design, or operation of the site. First Responders (paramedics, fire-fighters, and police officers) should be encouraged to download the associated educational tool on confined space situations. The tool can be placed on a hand-held device for reference purposes to assist a First-Responder in deciding to enter or not a suspect site. Chapter 7 – Concluding Chapter and Future Work 153  The Sullivan mine tragedy should focus people’s attention on the fact that insufficient time is spent on studying hazards at reclamation sites, especially when these dangers are associated with new reclamation techniques designed to protect the environment. Continued risk assessment is needed to investigate atmospheric hazards in industry. It is crucial to investigate possible ground faults and to amend mine regulations regarding atmospheric problems instead of waiting until the next "first accident" occurs. 7.3. Future Work The proposed methodology is applicable to other workplaces by changing consideration of the types of hazardous gasses generated and emitted. The four major elements of atmospheric risk are the same in all confined space accidents across all industries. The differences involve the degree to which different variables affect each risk element and the type of toxic gas. These depend on the type of operation, design and modifications, climatic conditions, technologies, and types of materials. For example, with breathing water wells, gas generation is due to low-oxygen levels from displacement by nitrogen and carbon dioxide. Gas emission is influenced by barometric pressure changes much more than by temperature changes. Air moves into the well when the pressure rises. Problems of this kind can be modelled by a structure similar to AFRA. A thermodynamic model similar to that used in AFRA can model the effect of barometric pressure on gas flow in breathing water wells. There are other examples in the literature that could benefit from using a similar expert system structure to solve their problems, such as reclaimed coal waste tips and sites above underground coal mine operations. Tools such as AFRA can take available knowledge and distribute it to workplaces to help prevent similar accidents from occurring. The most important lesson learned from the Sullivan Mine accident is that each new technology, material, or operation should be retested from Chapter 7 – Concluding Chapter and Future Work 154  different contexts of health and safety before application. Assessment should continue throughout the life of the structure should any change occur at the site.  AFRA is based on knowledge of current waste dumps in the literature that have been instrumented and studied. As more sites are tested, better understanding of the risk will be achieved and other combinations of variables that control atmospheric risk can be included. The weights and rules can be updated as understanding of these new waste dumps is achieved. this analysis has focused on structures located at the toe of  a dump. Danger also exists in a reverse manner with structures built on top of a dump. Further attention must be given to regulation and policies about housing construction on top of dumps that might take place decades after closure.  The last issue that affects human health and can make a work place more convenient for the workers is "communication requirements". This is a key factor in managing risk and should be considered in safety guidelines at reclamation sites. In evaluating the risk for confined spaces the presence of good emergency response and communication between closed mines and emergency services is important and will significantly reduce risk. Controlling worker access to the workplace as well as proper and efficient emergency response may stop additional fatalities or prevent the first from occurring.  155  Bibliography Acer, K. S., Bayer,  S. R.,1990. OSHA revamps rules for monitoring confined space, Pollution Engineering, v 22, n 10, p 40-44. Alonso-Azcarate, J., Bottrell, S. H., Tritlla, J., 2001. Sulfur redox reactions and formation of native sulfur veins during low grade metamorphism of gypsum evaporites, Cameros Basin (NE Spain), Chemical Geology, v 174, n  4, 15, 389-402. Anonymous, 1997. OSHA Confined Space Advisor 1.1, OSHA Software Expert System (http://www.osha.gov/dts/osta/oshasoft/csa.html). Amanti, M., Pecci, M., Scarascia, M.G., Vittori E., 1996. Environmental reclamation and safety Conditions for Recreation Use of Dismissed Rock Quarry: Case Studies in Central Italy. Proceedings 5th International Symposium on Mine Planning and Equipment Selection, Sao Paulo, Brazil, October 22-25, 647-652. ANU (Australian National University), Human Resources, 2008. Confined Space Risk Assessment and Entry Permit, Available at: http://info.anu.edu.au/hr/OHS/_Procedure_Attachments/Confined_Spaces_Risk_Assessm ent.asp  Aziz, M.L., Ferguson K.D., 1997. Equity Silver Mine - Integrated Case Study. Proc. 4th Intern. Conf. on ARD, Vancouver, B.C., May 31- Jun. 6, v 1, 181-196. Baker, G., 1986. Anatomy of a confined spaces accident, Water Bulletin, v 31, 8-10. Ball, T.K., Crow, M.J., Laffoley, N., Piper, D., Ridgway, J., 1990. Application of soil-gas geochemistry to mineral exploration in Africa, Journal of Geochemical Exploration, v 38, 1- 2, p 103-115.  Ball, T.K., Nicholson. R. A.,  Peachey, D., 1985. Gas Geochemistry as an Aid to Detection of Buried Mineral Deposits, Transactions of the Institution of Mining & Metallurgy, Section B, v 94, p 181-188. Balcita, J.V., Meech, J.A., Ghomshei, M.M., 1999. ARDx - a Fuzzy Expert System for ARD Site Remediation. Proc. 2nd Inter. Conf. on IPMM, Honolulu, HI, pp.12. Banerjee, S.C., 1995. Scope of Application of Bio-Technology for Fire-Control and Ameliorating Environmental Pollution of Coal Mines, Journal Mines, Metals & Fuels, v 43, n1-2, pp 17-21.  Barbarick, K.A., Colorado State University, Nitrogen Sources and Transformations, 2006. Soil and crop sciences. 2/96. Available at: http://www.ext.colostate.edu/pubs/crops/00550.html. Barnhart, C.R., 2003. Understanding the “Orange Smoke” Problem in Cast Blasting. February 2-5 Proceedings of the 29th Annual Conference on Explosives and Blasting Technique, Cleveland, OH: International Society of Explosives Engineers, pp. 1-11. Bibliography 156  Barrett, M.A., 1969. Oxygen Consumption and Hydrogen Production BY Shrink-Resistant Grouts in Confined Places, U S Bur Mines-Report Investigations 7267, 8p.  Bayah, J., Meech, J. A., Stewart, G.,1984. Oxygen-depletion Of Static Air by Backfill Material at The Thompson Mine, Mining Science & Technology, v 1, n 2, p 93-106. BC Mines Act, 2008. Queens Printer, Victoria, British Columbia, Canada, (RSBC. 1996).  Bennett, J.W., Pantelis, G., Ritchie, A.I.M., 1989. Limitations on pyrite oxidation rates in dumps set by air transport mechanisms. In Proceeding Internet symp. on biohydrometallurgy (J. Sally, R.G.L. McCready & P.L. Wichlacz, eds.). CANMET Report SP89-10 , 551-561.  Bishop, P.K., Lloyd, J.W., 1990. Chemical and isotopic evidence for the hydrogeological processes occurring in the Lincolnshire Limestone, J. Hydrol. V 121, pp. 293–320. Bronswijk, J.J.B., Nugroho, K., Aribawa, I. B., Groenenberg J. E., and Ritsema, C. J., 1993. Modeling of Oxygen Transport and Pyrite Oxidation in Acid Sulfate Soils, American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Published in J Environ Qual 22:544-554 (1993). Brooks, R.H., Corey, A.T., 1964, Hydraulic Properties of Porous media, Hydrol. Papers, Colo. State Univ., n 3, March 1964, pp 36. Bush, R.T., 2000. Iron sulfide micromorphology and mineralogy in acid sulfate soils: their formation and behavior. PhD thesis, University of NSW, Sydney.  Bush, R.T., Sullivan, L.A., 1997. Morphology and behavior of greigite from a Holocene sediment in eastern Australia, Australian Journal of Soil Research, v 35, 853-861.  Camacho, A., Vicente, E., Miracle, M. R., 2000. Spatio-temporal distribution and growth dynamics of phototrophic sulfur bacteria populations in the sulfide-rich Lake Arcas, Aquat. sci., v 62, 334–349. Canfield, D.E., 1989. Reactive iron in marine sediments, Geochimica et Cosmochimica Acta, v 53, n 3,p 619-632. Cao, S.G., Wang, Y.P., Liu, Y.B., Xu, A.M., 2006. Risk assessment model of gas explosion in coal mine based on the hazard theory, Meitan Xuebao/Journal of the China Coal Society, v 31, n 4, p 470-474.  Cathles, L.M., Schlitt, W.J., 1980. A Model of the Dump Leaching Process that Incorporates Oxygen Balance, Heat Balance and Two-Dimensional Air Convection. In Leaching and Recovering Copper From As-Mined Materials, Chapter 2, SME Publications, p.18. CCOHS,2002. Available at:   http://www.ccohs.ca/oshanswers/hsprograms/confinedspace_program.html  Cengel, Y.A., and M.A., Boles, 2006. Thermodynamics an Engineering Approach, Fifith Edition. Bibliography 157  CGWA, Factsheet#5, 2004, The Hidden Danger in Water Well Pits. Canadian Ground Water Association.  Chaulya, S.K., Singh, R.S., Chakraborty, M.K., Srivastava, B.K., 2000. Quantification of stability improvement of a dump through biological reclamation, Geotechnical and Geological Engineering, 18(3), 193-207.  Chaulya, S.K., Jain, S.C., Sharma, D.K., 1993. Reclamation of mine dumps and environment: a case study. Proc. ISRM International Symposium, 1, 275. Clean Air, 2007. Oxygen-depletion Project, Available at:  http://www.osh.govt.nz/publications/series/hb-24-oxygendepletion.html Collin, M., 1987. Mathematical modeling of water and oxygen transport in layered soil covers for deposits of Pyrite Mine Tailings. Royal Institute of Technology, Department of Chemical Engineering, Stockholm, Licentiate Treatise, 16 April.  Confined Space Entry Program - A Reference Manual, 2005. ISSN 1715-4189, Available at: www.worksafebc.com Confined Spaces Hazards, Part III, 1979. NIOSH Criteria Documents Criteria for a Recommended Standard: Working in Confined Spaces, December 1979. DHHS (NIOSH) Publication No. 80-106, Confined. Available at: http://www.cdc.gov/niosh/80-106.html Confined Space Policy, 2006. Environmental Health and Radiation Safety, Available at: http://www.ehrs.columbia.edu/ConfinedSpacePolicy.html Daniel, J.A., Harries, J.R., Ritchie, A.I.M., 1983.The quality and quantity of Runoff and Groundwater in two Overburden Dumps Undergoing Pyritic Oxidation, Australian Water & Wastewater Assoc, v 8, 1-24.  Dawson, B., Kuit, W.J.; Phillip, M.; Thomson, D, 2009. Sullivan Mine waste dump characterization, pt.1, British Columbia Mine Reclamation Symposium, Vancouver, B.C. pp.11. Deitchman, S., Decker, J., Santis, L. September 1998. A Novel Source of Carbon Monoxide Poisoning: Explosives Used in Construction, Annals of Emergency Medicine, v 32, n 3, 381-384. Dick, R. A., Fletcher, L. R., D’Andrea, D. V., 1983. Explosives and Blasting Procedures manual, U.S. Bureau of Mines Information Circular 8925, pp 4. Elberling, B., Nicholson, R.V., 1996. Field determination of sulfide-oxidation rates in mine tailings, Water Resour. Res., v 32, 1773–1784.  Eltschlager K.K. , Harris M.L., Baldassare F., 2004. Fugitive Carbon-Based Gasses: Blasting Related or Not, in: Proceedings of the 30th Annual Conference on Explosives and Blasting Technique (New Orleans, LA, Feb. 1-4, 2004). Vol. 1. Cleveland, OH: International Society of Explosives Engineers; 409-418.  Eltschlager, K.K., Hawkins, J.W., Ehler, W.C., Baldassare, F., 2001(a). Technical measures for the Investigation and mitigation of Fugitive methane Hazards in Areas of Coal Mining, Bibliography 158  Office of Surface Mining Reclamation and Enforcement Appalachian Regional Coordinating Center Pittsburgh, PA. Eltschlager, K.K., Shuss, W., Kovcalchuk, T. E., 2001 (b). "Carbon Monoxide Poisoning at a Surface Coal Mine A Case Study", In: Proceedings of the 27th Annual Conference on Explosives and Blasting Technique. Orlando, FL: International Society of Explosives Engineers. Engels J.M., Dixon-Hardy, D.W., Kreft-Burman, K., McDonald, C.P., 2007. Parameter framework – a risk management tool to improve the safety of tailings facilities, Land Contamination and Reclamation,15, 223-233.  Fala, O., Aubertin, M., Molson, J., Bussière, B., Wilson, G. W., Chapuis, R., Martin, V., 2003. Numerical Modeling of Unsaturated Flow in Uniform and Heterogeneous Waste Rock Piles, 6th ICARD, 895-902. Fenchel, T., King, G., Blackburn, T.H., 1998. Bacterial biogeochemistry: the ecophysiology of mineral cycling. San Diego: Academic Press. Franseen, H. W.,1995, Hydropower application of confined space regulations, Proceedings of the International Conference on Hydropower - Waterpower, v 2, p 1248-1255. Franseen, H.W.,1995. Hydropower application of confined space regulations, Proceedings of the International Conference on Hydropower - Waterpower, v 2, p 1248-1255. Garvie, A.M., Bennett, J.W. and Ritchie, A.I.M. 1997. Quantifying the spatial dependence of the sulfide oxidation rate in a waste rock dump at Mt Lyell, Tasmania. In: Proceedings of the Fourth International Conference on Acid Rock Drainage, Volume I (Vancouver, Canada, May 31 to June 6, 1997). The Abatement of Acidic Drainage, v 1, 148-156.  Geng, L., Zhi, C., Chan, C.W, Huang, G.H., 2001. A decision support system for management of oil-contaminated sites. Expert Systems Applications, v 20, n 3, 251-260. Ghomshei, M.M., Meech, J.A., 2000, Application of Fuzzy Logic in Environmental Risk Assessment: some thoughts on Fuzzy Sets. Cybernetics and Systems; 31(3), 317-332. Gooddy, D.C., Darling, W.G., 2005. The potential for methane emissions from groundwaters of the UK, Science of The Total Environment, v 339, Issues 1-3, pp 117-126 Grima, M.A., 2000. Neuro-fuzzy modeling in engineering geology: applications to mechanical rock excavation, rock strength estimation, and geological mapping. Rotterdam, A.A. Balkema, Netherlands, 244pp. Harder, A.H., Whitman, H.M., Rogers, M.S., 1963. Methane in the Fresh-water Aquifers of Southwestern Louisiana and Theoretical Explosion Hazard, US Geological Survey, Water Resources Pamphlet No. 19, 22pp.  Hargraves, A.J., 1983. Sources and Extent of Mine Atmosphere Contaminants, Ventilation of coal Mines, The Aus. I.M.M Illawarra Branch Symposium.  Harris, M. L., Mainiero, R. J., 2004. CO Migration from Trench Blasting in Amherst, February 1-4. New York. Proceedings of the 30th Annual Conference on Explosives and Blasting Technique, Cleveland, OH: International Society of Explosives Engineers, v II, pp 25-39 Bibliography 159  Harris, M. L., Rowland III, J. H., Mainiero, R. J., 2004. Blasting-Related Carbon Monoxide Incident in Bristow, Virginia. Proceedings of the 30th Annual Conference on Explosives and Blasting Technique, Vol. II, February 1-4. (pp 319-327). Cleveland, OH: International Society of Explosives Engineers.  Harries, J.R., Ritchie, A.I.M., 1987. The effect of rehabilitation on the rate of oxidation of pyrite in a mine waste rock dump, Env. Geochem. and Health, v 9, 27-36.  Harries, J.R., Ritchie, A.I.M., 1986. The impact of rehabilitation measures on the physiochemical conditions within mines wastes undergoing pyritic oxidation, Process Metal., v 4, 341-351, 1986.  Harries, J.R., Ritchie, A.I.M., 1985. Pore Gas Composition in Waste Rock Dumps Undergoing Pyritic Oxidation, Soil Science, v 140, n 2, 143- 152.   Harries, J.R., Ritchie, A.I.M., 1983. Runoff fraction and pollution levels in runoff from a waste rock dump undergoing pyritic oxidation, Water, Air and Soil Pollution, v 19, 155- 170. Harries, J.R., Ritchie, A.I.M., 1980. The use of temperature profiles to estimate the pyritic oxidation rate in a waste rock dump from an open-cut mine, Water Air and Soil Pollution, v 15, 405-423.  Hazardous Substances Data Bank, 1994. MEDLARS Online Information Retrieval System, National Library of Medicine. Hazards of Confined Spaces, 2004. http://www2.worksafebc.com Hendrick, D., Sizer, K., 1992. "Breathing" coal mines and surface asphyxiation from stythe (black damp). British Medical Journal, v 305, 509-510. Hendry, M.J., Wassenaar, L.I., and Birkham, T.K., 2002. Microbial respiration and diffusive transport of O2, 16O2, and 18O, in unsaturated soils: A mesocosm experiment.  Geochim. et Cosmochim. Acta, v 66, n 19, 3367-3374. Hendry, M.J., Mendoza, C.A., Kirkland, R., Lawrence, J.R. 2001. An assessment of a mesocosm approach to the study of microbial respiration in a sandy unsaturated zone. Groundwater, v 39, n 3, 391-400. Hendry, M.J., Lawrence, J.R., Zanyk, B.N., Kirkland, R. 1993. Mircobial production of carbon dioxide in unsaturated geologic media in a mesoscale model. Water Resources Research, v 29, 973-984. Hess, J.W., White, W.B., 1989. Chemical Hydrology in Karst Hydrology; Concepts from Mammoth Cave Area, Ed. White, W. B., White, E.L., Newyork: Van Nosrtrand Reinhold, 145- 150.  Hill, S.R., 2002. Physical and Geochemical Characterization of oxygen-depleted Breathing Wells in Central Alberta. M.Sc. Thesis, Earth & Atm. Sci., University of Alberta, pp.117. Hinkle, M. E. Dilbert, C. A., 1984. Gasses and Trace Elements in Soils at the North Silver Bell Deposit, Pima County, Arizona, , Journal of Geochemical Exploration, v 20, n 3, 323- 336. Bibliography 160  Hinkle, M.E. Ryder, J.L. Sutley, S.J. Botinelly, T., 1990. Production of sulfur gasses and carbon dioxide by synthetic weathering of crushed drill cores from the Santa Cruz porphyry copper deposit near Casa Grande, Pinal County, Arizona, Journal of Geochemical Exploration, 38, 1-2, 43-67. Hockley, D., Smolensky, J., Jahn S. and Paul M. 2000. Geochemical investigations and gas monitoring of an acid generating waste rock pile in Proceedings Fifth International Conference on Acid Rock Drainage, 181-189 (Society for Mining, Metallurgy and Exploration, Inc: Littleton). Hollings, P., Hendry, M. J. and Kirkland, R. A., 2000. Quantification of oxygen consumption rates for gneissic waste rock piles, Key Lake Uranium Mine, Northern Saskatchewan, Canada: application of the kinetic cell technique, in Proceedings Fifth International Conference on Acid Rock Drainage(Society for Mining, Metallurgy and Exploration, Inc: Littleton), 1s63-171. Hooker, P.J., Bannon, M.P., 1993. Methane: its occurrence and hazards in construction. (R130), CIRIA, pp 140.  Hutchison, I.P.G, Ellison, R.D., 1992. "Mine Waste Management, " Chapter 6: Climatic Considerations. Lewis Publishers Inc. Chelsea, Michigan. Imhoff, J.F., 1992. Taxonomy, phylogeny, and general ecology of anoxygenic phototrophic bacteria, (eds) Mann, N.H. and Carr, N.G., photosynthetic prokaryotes, plenum press, New York, 53-92.  International Society of Explosives Engineers (ISEE),1998. Blasters’ Handbook, 48-89. Jeana, V., Harrison, M., Rao, C. Y., Benaise, L. G., 2004. NIOSH health hazard evaluation report, HETA #2004-0075-2944, WV Department of Environmental Protection Fairmont, Department of health and human services, Centers for Disease Control and Prevention, NIOSH. Kaczynski R., 1986. Seismic Stability on a Waste-Dump. Geologica Applicata Idrogeologia, v 22, n 2, 343-352.  Katsabanis, P.D., Liu, Q., 1993. Publication: Calorimetric determination of. the heat of detonation of commercial explosives, In Proceedings of the 9th Annual Symposium on Explosives and Blasting Research, San Diego, California, International Society of Explosives Engineers, Cleveland, Ohio, 31-39. Kelly, D.P. 1974. Growth and metabolism of the obligate photolithotroph chlorobium thiosulfatophilum in presence of the added organic nutrients. Archive of Microbiology, v 100, 163-178.  Konhauser K., Bertola G., 2007. Introduction to Geomicrobiology, Published by Blackwell. 440.  Kreft-Burman, K., Saarela, J., Engels, J.M., Dixon-Hardy, D.W., 2007. Efficient implementation of mining waste regulations and future tailings related qualifications; Land Contamination and Reclamation, v 15, 219-223.  Bibliography 161  Kuo, E.Y, Ritchie, A.I.M, 1999. The impact of convection on the overall oxidation rate in sulfidic waste rock dumps, in Proceedings Mining and the Environment II 1999 (Eds: D Goldsack, N Belzile, P Yearwood and G Hall), Sudbury, Ontario, Canada, 211-220. Lahmira, B., Lefebvre R., Hockley, D., Phillip, M., 2009.Sullivan Mine Fatalities Incident: Numerical Modeling of Gas Transport and Reversal in Gas Flow Directions, Securing the Future and 8th ICARD, June 23-26, 2009, Skellefteå, Sweden.  Lawrence, D.L.,1995. A Study of Post Blast Generation of Nitrogen Dioxide, Proceedings of the 22nd Annual Conference on Explosives and Blasting Technique, Cleveland, OH: International Society of Explosives Engineers, 1-12. Lazarov, S.B., Brinkley, R.F., Tole, D.M., 1975. Explosives and Formation of Toxic Gasses in Large- Scale Blasting, 16th International Conference on Coal Mine Safety Research, September 22 (p IV5.1- IV5.9). Bureau of Mines, Pittsburgh, PA. Lee, E.S., Hollings, P., Hendry, M.J., 2003. Use of oxygen consumption and carbon dioxide production in kinetic cells to delineate pyrite oxidation-carbonate buffering and microbial respiration in unsaturated media; J. Cont. Hydro., 65, 203-217. Lefebvre, R., Lamontagne, A., Wels, C., Robertson, A., 2002. ARD Production and Water Vapor Transport at the Questa Mine, in Proceedings Ninth International Conference on Tailings and Mine Waste, Tailings and Mine Waste ’02 (AA Balkema Publishers: Lisse), 479-488.  Lefebvre, R., Hockley, D., Smolensky, J., Gelinas, P., 2001(a). Multiphase transfer Processes in Waste Rock Piles Producing Acid Mine Drainage 1. Conceptual Model and System Characterization, Journal of Contaminated Hydrology, v 52, 137- 164.  Lefebvre, R., Lamontagne, A., Wels, C., 2001(b). Numerical simulations of acid rock drainage in the Sugar Shack South rock pile, Questa Mine, New Mexico, U.S.A. Proceedings, 2nd Joint IAH-CNC and CGS, Groundwater Specialty Conference, 54th Canadian Geotechnical Conference, Calgary, Canada, pp 8.  Lefebvre, R., Gelinas, P., Isabel, D., 1992. Heat transfer analysis applied to acid mine drainage production in a waste rock dump, La mine Doyon (Quebec). 1992 Conference of the Canadian National Chapter, International Association of Hydro geologists, Modern Trends in /hydrogeology, Hamilton , May 11-13, 294-307. Lentini, J.J, Armstrong, A.T., 1997. Comparison of the Eluting Efficiency of Carbon Disulfide with Deithyl Ether: The Case for Laboratory Safety, Journal of FORENSIC SCIENCES, 307 - 311 Lin, T.Y.D., 2010. Modeling the 3D Net Infiltration Distribution at the Equity Silver Mine Waste Dump, Masters thesis, The University of British Columbia, pp.181. LIoyd, D., 2006. Hydrogen sulfide: clandestine microbial messenger, Trends in Microbiology Vol.14 No.10. Lovell, J.S., Hale, M., 1983. Application of Soil- Air Carbon Dioxide and Oxygen Measurement to Mineral Exploration, Transactions of the Institution of Mining & Metallurgy, Section B, v 92, 28-32. Bibliography 162  Lovell, J.S., Hale, M., Webb, J.S., 1982. Soil Air carbon Dioxide and Oxygen Measurements as a Guide to Concealed Mineralization in Semi-Arid and Arid Regions. Journal of Geochemical Exploration, v 19, n 1-3, 305-317. Lunarzewski, L., Battino, S., 1984. APPLICATION OF GAS PRE-DRAINAGE FOR PREVENTION OF CO//2 OUTBURST, Inst of Mining & Metallurgy, 159-165. Lundgren, T., 2000. The Dynamics of oxygen Transport into Soil covered Mining Waste Deposits in Sweden, Journal of Geochemical Exploration, v 74, 163-173,  Luther, G.W., 1991. Pyrite synthesis via polysulfide compounds, Geochimica et Cosmochimica Acta, v 55, n 10, 2839-2849. Mainiero, R.J., Harris, M.L. Rowland III , J.H., 2007. Dangers of Toxic Fumes from Blasting, International Society of Explosives Engineers, v 1, 1-6.  Mainiero, R.J., Rowland III, J.H., Harris, M.L., Sapko, M.J., 2006. Behavior of Nitrogen Oxides in the Product Gasses from Explosive Detonations, International Society of Explosive Engineers, 1-10.  Martel, R., Trépanier, L., Lavigne, M. A., Lévesque, B., Sanfac on, G., Brousseau, P., Auger, P., 2004. Carbon monoxide poisoning associated with blasting operations close to underground enclosed spaces. Part 2. Special working procedures to minimize CO migration, Canadian Geotechnical Journal, v 41, n 3, 383-391.  Mantel, R., Trepanier, L., Boutin, L., Lavigne, M., Levesque, B., Sanfacon, G., Auger, P., Galarneau, L., Brousseau, P., 2001. Carbon Monoxide poisoning from Blasting Operations in Construction Works, An Earth Odyssey.  Massmann, J .W., Farier, D.F., 1992.Effects of atmospheric pressures on gas transport in the vadose zone, Water Resources Research, v 28, n 3, 777-791. Material safety Datasheets, 2006; Available at: http://www2.worksafebc.com/i/posters/pdfs/2009/ws_09_02.pdf http://www2.worksafebc.com/i/posters/2009/WS%2009_03.html Matsui, K., Shimada, H., Itoi, R., Ueda, T., Nakagawa, H., Shiraishi, T., Furukawa, H., Kramadibrata, S., Sulistianto, B., 2004. Rehabilitation at a Surface Coal Mine in Tropical Countries, Coal and Safety, n 25, 7-15. May, E.R., Schmidt, P.G., 1982. The discovery, geology and mineralogy of the Crandon Precambrian massive sulfide deposit, Wisconsin, Precambrian Sulfide Deposits, In: R.W. Hutchinson et al., (Editors), Precambrian Sulfide Deposits, Geol. Assoc. Can., Spec. Pap., 25: 447-480. McAleenan, C., McAleenan, P., 1999. Confined Spaces Working - Towards Zero Fatalities. Expert Ease, pp 25. McCarthy Jr, J.H., Lambe, R.N., Dietrich, J.A., 1986. A case study of soil gasses as an exploration guide in glaciated terrain - Crandon massive sulfide deposit, Wisconsin. Economic Geology, v 81, n 2, 408-420. Bibliography 163  Meech J.A., Kumar, S.,1993. A HyperManual on Expert Systems; Proc. 25th Canadian Mineral Processors Conference, Ottawa, Ontario, Paper 22, pp 10. Meech, J.A., Veiga, M., 1995. Fuzzy Assessment of Environmental Risk. IEEE-SMC, Vancouver, 1640-43. Meech, J.A., 1999. Fuzzy Logic Control,  Encyclopedia of Electrical and Electronics Engineering, Editor - J.G. Webster, J. Wiley and Sons, New York, v 8, 138-154. Michaelsen, G.S., Park, M.D., 1954. Asphyxiation in street manholes. Pub. Health Rep., v 69, n 1, 29-36. Millero, F.J., 1986. The thermodynamics and kinetics of the hydrogen sulfide system in natural waters, Mar. Chem, v 18 , 121–147.  Mohammadi, L. Meech, J.A., 2008, Implementing Atmospheric Risk Assessment in Mine Reclamation, 23rd International Conference on Solid Waste Technology and Management, Philadelphia, PA, USA, March, pp.12. Mohammadi L., Meech, J.A., 2011. Atmospheric Occupational Health and Safety Issues at Mine Reclamation Sites, Proceedings of the International Conference on Environmental Pollution and Remediation, Ottawa, Ontario, Canada, 17-19 August 2011. Morin, K. A., Gerencher, E., Jones, C. E. and Konasewich, D. E., 1991. Critical literature review of acid drainage from waste rock, MEND Report 1.11.1. Morin,  K.A., and N.M. Hutt. 1994. Observed preferential depletion of neutralization potential over sulfide minerals in kinetic tests: Site-specific criteria for safe NP/AP ratios. IN: Proceedings of the Third International Conference on the Abatement of Acidic Drainage, Pittsburgh, Pennsylvania, USA, April 24-29, v 1, 148-156. Newman, L.L, Herasymiuk, G.M., Barbour, S.L., Fredlund, D.G. and Smith, T., 1997. The hydrogeology of waste rock dumps and a mechanism for unsaturated preferential flow, in Proceedings Fourth International Conference on Acid Rock Drainage, 551-564 (Natural Resources Canada: Ottawa).  Nordstrom, D.K., Southam, G., 1997. Geomicrobiology of sulfide mineral oxidation. In: J.F. Banfield and K.H. Nealson (Eds.), Geomicrobiology: interactions between microbes and minerals. Rev. Mineral., v 35, 361-390. NIOSH Pocket Guide to Chemical Hazards, 2005. Available at:        http://www.cdc.gov/niosh/npg/. Oakes, B.W., Hale, M., 1985. Dispersion Patterns of Carbonyl Sulfide Above mineral Deposites, Journal of Geochemical Exploration, v 28, n 1-3, 235-249. OHS Regulation, 2008. Part 9 confined Spaces, Available at: http://www2.worksafebc.com  Operation of Mines Regulation in MB, 1994. Available at: http://web2.gov.mb.ca/laws/regs/pdf/w210-228.94.pdf Bibliography 164  Ormerod J., 1992. Physiology of the Photosynthetic prokaryotes. In: Mann, N.H. and Carr, N.G. (eds), Photosynthetic prokaryotes. Plenum Press, New York, 93-120. Orr, W.L., 1975. Geologic and geochemical controls on the distribution of hydrogen sulfide in natural gas. Ed. Campos, R., Goni, J.,Advances in Organic Geochemistry, , Madrid, Spain, Pergamon Press (1977), 571–597. OSHA, 2004. Permit-Required Confined Spaces, Available at: www.osha.gov  Pantelis, G., Ritchie, A.I.M., 1991. Macroscopic transport mechanisms as a rate-limiting factor in dump leaching of pyritic ores, Applied Mathematical Modeling, v 15, n 3, 136-143.  Pfennig, N.,  1978. Rhodocyclus purpureus gen. nov. and sp. nov., a Ring-Shaped, Vitamin B12-Requiring Member of the Family Rhodospirillaceae, Int J Syst Bacteriol, v 28, 283- 288. Pearson, F., Edwards, J.S., 1989. Methane entry into the Carsington Aqueduct system. In: Methane - Facing the Problems Symposium, Paper 4.3, Nottingham, U.K. Pearson, F.H. and McDonnell, A.J., 1975. Use of crushed limestone to neutralize acid wastes. J. Environ. Engin. Div., ASCE, v 101, n 11131, 139-158.  Peel, M. C. and Finlayson, B. L. and McMahon, T. A., 2007. Updated world map of the Köppen–Geiger climate classification. Hydrol. Earth Syst. Sci. v 11, 1633-1644. Pettit, T.A.,1994. Confined-Space-Related Fatalities, Part I, Prevention: Elements of a Confined Space Program, DHHS (NIOSH) Publication, 27-31. Pettit, T., Linn, H., 1987. A Guide to Safety in Confined Spaces, U.S department of Health and Human Services, Public Health Service, center for Diseases Control, NIOSH, DHHS (NIOSH) publication, NO. 87-113, 1-20. Pettit, T.A., Sanderson, L. M., Linn, H. I.,1987. Workers/ Rescuers Continue to Die in Confined Spaces, Professional Safety, v 32, n 2, 15-22. Phillip M., Hockley D., and Dawson B., 2008. Sullivan mine fatalities incident: preliminary technical investigations and findings, 6th Australian Workshop on Acid and Metalliferous Drainage. Pierson, B.K., Parenteau, M.N., 2000. Phototrophs in high iron microbial mats:  Microstructure of mats in iron-depositing hot springs.  FEMS Microbiology Ecology, v 32:181-196. Raiswell, R., Berner R.A., 1986. Pyrite and organic matter in Phanerozoic normal marine shales, Geochimica et Cosmochimica Acta, v 50, n 9, 1967-1976.  Rawls, W.J., Brakensiek, D. L., 1985. prediction of Soil Water Properties for Hydrologic Modeling, Watershed management in the eighties : proceedings of a symposium, Denver, Colorado, Jones, E.B., Ward, TJ. (eds.),- New York, ASCE, 293-299.  Ritchie, A.I.M., 2003. Oxidation and gas transport in piles of sulfidic material Chapter 4. in: J.L. Jambor, D.W. Blowes and A.I.M. Ritchie, Editors, Environmental Aspects of Mine Wastes, Short Course, Mineralogical Association of Canada (2003), v 31, 73–94.  Bibliography 165  Ritchie, A.I.M., 1994. Sulfide oxidation mechanisms: controls and rates of oxygen transport, in Short Course Handbook on Environmental Geochemistry of Sulfide Mine Wastes, Mineralogical Association of Canada, 22 (Ed. J.L. Jambor and D.W. Blowes), 201-245.  Robertson GeoConsultants Inc., 2001. Progress Report on Water Balance Study for Mine Rock Piles, Questa Mine, NM. Report 052008/17 prepared for Molycorp Inc., February, pp.28.  Rowland III, J.H., Mainiero, R.J., 2000. Factors Affecting ANFO Fumes Production. Proceedings of the 26th Annual Conference on Explosives and Blasting Technique, Vol. I, February 13-16, Cleveland, OH: International Society of Explosives Engineers, 163-174. Santis, L.D., 2001. A Summary of Subsurface Carbon Monoxide Migration Incidents, Proceedings of the 27th Annual conference on Explosives and Blasting Technique, January 28-31. Cleveland, OH: International Society of Explosives Engineers, 143-150.  Savci, G., Williamson, A., 2002. Hydrologic Assessment of Tailing Impoundments: A Case Study from an Existing Facility in Copper Mine Queen Mine, Bisbee, Arizona, 2002 SME Annual Meeting (Feb. 25-27, 2002, Phoenix, Arizona),  Scharer, J.M., Nicholson, R.V., Halbert, B., Snodgrass, W.J., 1994. A computer program to assess acid generation in pyretic tailings. In: C.N. Alpers and D.W. Blowes, Editors, Environmental Geochemistry of Sulfide Oxidation, Am. Chem. Soc. Symp. Series, v 550, 132–152,  Schatzel, S.J., Hyman, D. M., Sainato, A., LaScola, J.C., 1987. Methane Contents of Oil Shale from the Piceance Basin, CO, Report of Investigation 9063, U.S. Department of the Interior, Bureau of Mines. Report of Investigations 9063, NTIS No. PB87-216453, 1-32. Shaw, S., Wels, C., Robertson, A., Lorinczi, G., 2002. Physical and geochemical characterization of mine rock piles at the Quest Mine, New Mexico: An overview, in Proceedings Ninth International Conference on Tailings and Mine Waste, Tailings and Mine Waste ‘02, 447-458 (AA Balkema Publishers: Lisse). Sherwood, B., Fritz, P., Frape, S.K., Macko, S.A., Weise, S.M., Welhan, J.A., 1988. Methane occurrences in the Canadian Shield, Chemical Geology, v 71, Issues 1-3, 223-236. Smith, C., Rumancik, W.P., Lazzara, C.P., 1996. SPONCOM - A Computer Program for the Prediction of the Spontaneous Combustion Potential of an Underground Coal Mine, Proc of the Fifth Conf on the Use of Computers in the Coal Industry, Thompson SD, Grayson RL, Wang YJ, eds. Morgantown, WV: West Virginia University, 134-143. Smolensky, J., Hockley, D., Lefebvre, R., Paul, M., 1999. Oxygen Transport Processes in the Nordhalde of the Ronnenburg Mining District, Germany, in Proc. Mining and the Environment II 1999, Sudbury, On, (Eds.: D Goldsack, N Belzile, P Yearwood, and G Hall), 271- 280.  Solomon, D.K., and T.E. Cerling. 1987. The annual carbon dioxide cycle in a montane soil: observations, modeling, and implications for weathering. Water Resour. Res. v 23, n 12, 2257-2265. Bibliography 166  Sracek, O., Choquette, M., Gélinas, P., Lefebvre, R. and Nicholson, R.V., 2004. Geochemical characterization of acid mine drainage from a waste rock pile, Mine Doyon, Québec, Canada, Journal of Contaminant Hydrology, v 69, n 1-2, 45-7.  Sracek, O., Gelinas, P., Lefebvere, R., Nicholson , R.V., 2006. Comparison of methods for the Estimation of Pyrite Oxidation Rate in a Waste Rock Pile at Mine Doyon Site Quebec, Canada, Journal of geochemical Exploration, v 91, 99-109. Canadian Vital Statistics, Death Database. Available at: http://cansim2.statcan.ca/cgi- win/cnsmcgi.exe?Lang=E&RootDir=CII/&ResultTemplate=CII/CII___&Array_Pick=1&Arra yId=1020540 Stedman, D.H. Creech, M.Z. Cloke, P.L. Kesler, S.E. Gardner, M., 1984. Formation of CS2 AND COS from Decomposition of Metal Sulfides, Geophysical Research Letters, v 11, n 9, p 858-860. Stromberg, B. , Banwart, S. , 1999. Experimental Study of Acidity- consuming processes in mining waste rock : some influences of mineralogy and particle size, Applied geo chemistry, 14, 1-16. Stromberg, B., Banwart, S., 1994. Kinetic modeling of geochemical Processes at the Aitik mining waste Rock Site in Northern Sweden, Applied Geochemistry, 9, 583-595, 1994. Sullivan Mine Incident Technical Panel, 2010. Sullivan mine fatalities - Technical Investigations, Summary Report, BC MEMPR, March, pp.20. Suruda, A.J. S., Castillo, D.N., Helmkamp, J.C.H., Pettit, T.A., 1994. Epidemiology of Confined-Space- Related Fatalities, Confined-Space-Related Fatalities, Part I, Worker Death in Confined Spaces, A Summary of NIOSH Surveillance and Investigative Findingfs, DHHS (NIOSH) Publication, 11-27 Suruda, A.J., Pettit, T.A., Noonan, G.P., Ronk, R.M., 1993. Deadly rescue: the confined space hazard. Journal of Hazardous Materials, v 36, n 1, 45-53. Suruda, A.J., Agnew, J., 1989. Deaths from Asphyxiation and Poisoning at Work in the United States 1984-86, British Journal of Industrial Medicine, v 46, 541-546.  Swanson, D.A, O'Kane, M., August 1999. Application of Unsaturated Zone Hydrology for Mine Waste Management: Design of Soil Covers and Prediction of Seepage. 16th Annual Conference of American Society for Surface Mining and Reclamation, Scottsdale Arizona.  Takala, I., Kontturi, M., Rutkvist, J., Lindvall, M., 2001. The Aitik Waste Rock Decommissioning Project, Proceedings - Securing the Future – international conference on mining and the environment, Skelleftea, June 25-July 1, pp.8. Tauziède, C., Pokryszka, Z., Barrière, J.P., 2002. Risk assessment of surface emission of gas from abandoned coal mines in France and techniques of prevention. Mining Tech.: IMM Transactions A, v 111, n 3, 192-196 Taylor, C.H. Kesler, S.E. Cloke, P.L., 1982. Sulfur Gasses Produced by the Decomposition of Sulfide Minerals- Application to Geochemical Exploration. Journal of Geochemical Exploration, v 17, n 3, 165-185. Bibliography 167  Timms, G.P., Bennett, J.W., 2000. The Effectiveness of Covers at Rum Jungle After Fifteen Years, ICARD 2000: Proceedings from the Fifth International Conference on Acid Rock Drainage Volume II, Chapter 7 - Prevention and Remediation of Problematic Mine-Waste Drainage. Tritlla, J.; Alonso-Azcarate, J.; Bottrell, S.H., 2000. Molten sulfur-dominated fluids in the origin of a native sulfur mineralization in lacustrine evaporites from Cavera del Rio Alhama (Cameros Basin, NE Spain) , Journal of Geochemical Exploration, v 69-70, 183-187.  U.S Bureau of Labor Statistics (U.S. BLS), 2010. Occupational Injuries/Illnesses and Fatal Injuries Profiles. Available at: http://data.bls.gov/GQT/servlet/InitialPage U.S. Department of Health and Human Services, 2005. Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 2005-113. ( www.cdc.gov/niosh/mining/pubs/pdfs/2005- 113.pdf). U.S. Department of Labor, 2011, available at: http://www.laborcommission.utah.gov/UOSH/Outreach/ConstructionCD/www.osha.gov/SLTC /smallbusiness/sec12.html Veiga, M., Baker, R., 2004. Protocols for environmental and health assessment of Hg released by artisanal gold miners. Global Mercury Project, UNIDO. Veiga, M., Meech, J.A., 1997. Fuzzy-Neural Systems for Adaptive Reasoning on Environmental Risk Analysis. MineIT '97, IT in the Mineral Industry, Athens, 10p. Veiga, M., Meech, J.A., 1995a. An Adaptive Fuzzy Model for Risk Assessment of Hg Pollution in the Amazon. IFSA-95, 6th Inter. Fuzzy Systems Association World Congress, Sao Paulo, Brazil, 489-492. Veiga, M., Meech, J.A., 1995b. HgEx: Heuristic System on Hg Pollution in Amazon. Water, Air, Soil Pol. 80, 123-132. Veiga, M.M., 1994. A Heuristic System for Environmental Risk Assessment of Mercury from Gold Mining Operations. Ph.D. thesis. Dept. Mining and Mineral Process Engineering, University of British Columbia, Vancouver, BC, Canada. 196p. Veiga, M., Meech, J.A., 1994. Application of Fuzzy Logic to Environmental Risk Assessment. Proc. Intern. Congress on Min. Proc., Santiago, Chile, 355-370. Wang, H., Dlugogorski, B., Kennedy, E., 2003. Coal oxidation at low temperatures: oxygen consumption products, reaction mechanism, kinetics. Energy & Comb. Sci, v 29, 487–513. Ward, D.M., Weller, R., Shiea, J., Castenholz, R.W., Cohen, Y., 1989. Hot Spring Microbial mats: anoxygenic and oxygenic mats of possible evolutionary significance. In: Cohen, Y., and Rosenberg, E., (eds), microbial Mats: physiological ecology of benthic microbial communities. American Society for Microbiology. Washington, D.C., pp 3-13. Ward, N.J., Sullivan, L.A., Bush, R.T., 2004. Soil pH, oxygen availability, and the rate of sulfide oxidation in acid sulfate soil materials: implications for environmental hazard assessment, Australian Journal of Soil Research, v 42, 509-514. Bibliography 168  Watmuff, I.G., Morris, B.J., 1993. Soil air carbon dioxide and oxygen measurements as an aid to exploration for concealed sulfide mineralization in South Australia - a practical assessment, Conference Series - Australasian Institute of Mining & Metallurgy, 63-72. Weiss, E. S., Cashdollar, K.L. , Sapko, M.J., Bazala, E.M., 1995. Secondary Explosion Hazards During Blasting in Oil Shale and Sulfide Ore Mines Reference:Pittsburgh, PA: U.S. Department of Energy, RI 9632. NTIS stock number: PB97-127658, pp 1-50 Wels, C., Lefebvre, R., Robertson, A., 2003. An Overview of Prediction and Control of Air Flow in Acid-generating Waste Rock Dumps. Robertson GeoConsultants Inc. (RGC) / EnviroMine – Environmental Tech. for Mining.  Wels, C., O’Kane, M., 2003. Design of mine Waste Cover Systems- Linking Predicted performance to groundwater and surface Water Impact, in proceeding of the 6th International Conference on Acid Rock Drainage , Carins, QLD, Australia, 341-349. Wels, C., O’Kane, M., Fortin, S., Christensen, D., 2001. Infiltration Test Plot Study for Mine Rock Piles at Questa Mine, New Mexico, Proc. 9th Conf. of the Am. Soc. For Surface Mining Reclamation, Albuquerque, NM, June, 1995-209. White, W.B., 1988. Geomorphology and Hydrology of Karst terrains. Oxford University Press, New York. 464 pp.  Worden, R.H., Smalley, P.C., Barclay, S.A., 2003. H2S and diagenetic pyrite in North Sea sandstones: due to TSR or organic sulfur compound cracking?, Journal of Geochemical Exploration, v 78-79, 487-491. Workers Compensation Board of Manitoba WCB, 2006. Fatalities in Manitoba Workplaces for the Period 1995 – 2005, Available at: http://www.safemanitoba.com/uploads/Workplace_Fatalities_2000_2005.pdf Work safe alerts, 1998, Victoria, Available at: http://www.analox.net/product- docs/HOSP_incident_cellar_service.pdf WorkSafeBC, 2004, Hazards of Confined Spaces for Food and Beverage Industries, Available at: http://www.worksafebc.com/publications/high_resolution_publications/assets/pdf/bk82.pdf WorkSafeBC, 2010(a). Publication Index, Annual reports, WorkSafeBC's past results. Available at: http://www.worksafebc.com/publications/reports/statistics_reports/default.asp  WorkSafeBC, 2010(b). Table 6A Claim Counts by Group of the 1991 Standard Occupational Classification (SOC), Group of Accident Type and Year; Injury Years 2000-2009, Available at: http://www.worksafebc.com/publications/reports/statistics_reports/occupational_injuries/19 97-2006/assets/pdf/Table%206A.pdf, and http://www.worksafebc.com/publications/reports/statistics_reports/occupational_injuries/20 00-2009/assets/pdf/Table6A.pdf Wu, A., Wang, H., Xi, Y., Yang, B., Li, J., Yin, S., Zha, K., 2005. Heat transfer law in leaching dump. Central South University of Technology, PRC, v 12, n 6. Bibliography 169  Zadeh, L. A., 1965. Fuzzy Sets. Information and Control, v 8, 338-353. Zadeh, L.A., 2002. From Computing With Numbers to Computing with Words — from Manipulation of Measurements to Manipulation of Perceptions, International J. Applied Mathematics & Computer Science,  v 12, n 3, 307–324.  170  Appendices  Appendix A Statistics on Confined Space Fatalities  Table A-1. Confined-space accident deaths from 1980 to 2009 in the United States                     characterized by event and exposure and presented in numerically. Data from 1989 and                     before are from NIOSH (Suruda et al., 1994) while data after 1992 are from Dep't. of                     Labour statistics. Death from drowning is not considered from 1992 (U.S. BLS, 2010).                     (Data in both NIOSH (by CDC) and BLS are public domain and do not need specific                      permission)  Event or Exposure Inhalation, oxygen- deficiency, engulfment and drowning in enclosed, restricted, or confined spaces Depletion of oxygen in enclosed, restricted, or confined spaces(code 384XXX) Inhalation in enclosed, restricted, or confined space (code 3411XXX) Depletion of oxygen from cave-in or collapsed materials (383X) Total 1980 83 Not classified Not classified Not classified 83 1981 86 Not classified Not classified Not classified 86 1982 68 Not classified Not classified Not classified 68 1983 52 Not classified Not classified Not classified 52 1984 87 Not classified Not classified Not classified 87 1985 58 Not classified Not classified Not classified 58 1986 62 Not classified Not classified Not classified 62 1987 45 Not classified Not classified Not classified 45 1988 61 Not classified Not classified Not classified 61 1989 68 Not classified Not classified Not classified 68 1990 Not classified Not classified Not classified Not classified - 1991 Not classified Not classified Not classified Not classified - 1992 Not classified 13 51 0 64 1993 Not classified 3 43 4 50 1994 Not classified 3 44 3 50 1995 Not classified 10 35 0 45 1996 Not classified 12 56 0 68 1997 Not classified 5 25 0 30 1998 Not classified 7 27 0 34 1999 Not classified 0 23 0 23 2000 Not classified 5 22 0 27 2001 Not classified 13 25 0 38 2002 Not classified 10 20 0 30 2003 Not classified 7 27 0 34 2004 Not classified 9 13 Not classified 22 2005 Not classified 0 21 Not classified 21 2006 Not classified 4 15 Not classified 19 2007 Not classified Not classified 32 Not classified 32 2008 Not classified Not classified 17 Not classified 17 2009 Not classified Not classified 11 Not classified 11    Appendices 171     Table A-2. Asphyxiation in different industries in 1984 in 47 states of US between 1984 and 1986                         (data from Bureau of Labor Statistics Survey of OII) "Reproduced by permission from                       BMJ Publishing Group Limited. [from Asphyxiation and Poisoning at Work in the                        United States 1984-86, Suruda, A.J., Agnew, J., v 46, 541-546, 2011] ".  Industry Mechanical Simple Asphyxiation Toxic gasse s Oxygen deficiency Solvents Other Report incomplete Total No. of Employees in 1984 (million) Rate per million workers Agriculture 3 0 0 0 0 1 0 4 933.8 1.68 Construction 3 6 4 8 6 0 0 27 4345 2.44 Oil/gas 1 8 11 4 2 0 0 26 612.7 16.64 Manufacturing 20 26 13 7 11 6 5 88 19412 1.78 Services 1 2 7 2 5 1 1 19 20761 0.36 Trade, wholesale and retail 10 1 12 4 5 0 1 33 22134 0.58 Transport 3 1 3 1 5 1 0 14 2673.1 2.05 Utilities 1 4 7 0 0 0 0 12 898.9 5.24 Other 0 0 8 1 1 0 0 10 7213.3 0.54 Total 42 48 65 27 35 9 7 233 78983.8 1.16   Table A-3. Death from inhalation and oxygen-deficiency in a confined spaces in mining       industry in U.S (U.S. BLS, 2010). (Data in BLS is public domain and do not need                     specific permission)  Year Depletion of oxygen in enclosed, restricted, or confined spaces(code 384XXX)  Inhalation in enclosed, restricted, or confined space (code 3411XXX)  Depletion of oxygen from cave-in or collapsed materials (383X) Total  1992 0 0 0 0 1993 0 0 4 0 1994 0 6 private mining 3 6 1995 0 4 private mining 0 4 1996 0 5 private mining 0 5 1997 0 0 0 0 1998 0 0 0 0 1999 0 3 mine and quarry 0 3 2000 0 0 0 0 2001 0 0 0 0 2002 3 mining private industry 0 0 3 2003 0 4 Natural resources and mining 0 4 2004 0 4 Natural resources and mining 0 4 2005 0 0 0 0 2006 0 0 0 0 2007 0 7 0 7 2008 0 4 0 4 2009 0 0 0 0   Appendices 172      Table A-4. Work-related fatalities in Manitoba, 1995-2005 (WCB, 2006, adapted by permission).  Cause of Work-Related Fatalities in Manitoba Number of Fatalities Percentage of Fatalities Occupational Disease Fatalities Heart disease 15 4% Other Cancer 16 4% Asbestosis and Mesothelioma 96 25% Other diseases 19 5% Acute Hazard Fatalities Mobile Vehicle 113 30% Machinery Contact 17 5% Excavation and Structural Failure 28 7% Explosion/Fire/Electric Current 21 6% Confined Entry 6 2% Other Acute- Hazards 45 12% Total - 377 100% Appendices 173   Table A-5. Number of confined space related fatalities in British Columbia based on number of claims accepted for fatal benefits by subsector                     and the year in which it was accepted (WorkSafeBC, 2010 (a), data are extracted by permission).  Sector/ Subsector 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Agriculture                         3w  Amusement Facilities                           Canadian Pacific ltd. (Rail and Mining)                       1s    Construction trades    1d                      Fishing               1n            Food products               1o            Heavy manufacturing 1a   1e                      Hotel and restaurants     1f                      General retail         1j                  Logging                           Misc. Manufacturing                           Mineral products                          Mining (not coal)                           Municipalities         1k                  Oil and gas well drilling     1g                    1x  Other services   1b               1q    1y  Petroleum, Coal, Rubber, Plastic and chemical products             1l           1v   Provincial government                      2t    Pulp and paper    2h                      Road and Construction             1m              Sawmill   1c                        Technical Services                       1u    Transportation and related services                 4p        1z  Trucking     1i                      Wood and paper                    1r     1zz Total 1 2 7 0 2 0 2 2 4 0 2 3 1 3 1  a A welder was exposed to paints, solvents and fumes and the cause of death was heart attack b A 27 year old tireman died by CO poisoning when he parked his car in lot and left the motor running c Vapours from lumber irritated lungs of one mill worker. Cause of death: complications of severe bronchial asthma. d Sealer died applying concrete sealant within a cement structure when an explosion and then a fire occurred. e A welder was welding an aluminum fuel tank into a boat when an explosion occurred Appendices 174  f Guide buried in an avalanche while heli-skiing g a labourer was exposed to CO in well- building site h 2 worker was found dead in a confined space possible asphyxiation i A guide and his client were caught in an avalanche j A guard asphyxiated by propane heater in an enclosed room k A contractor collapsed in bottom of manhole, asphyxiated from methane gas L A 55 year old labourer was buried in a collapsed excavation m A 45 year old welder entered barge- lack of oxygen in hold caused worker collapse n A 34 year old fisherman died by a exhaust fumes that were blown by wind into the boat and caused carbon monoxide exposure o A 47 year old owner fell into wine fermentation tank and overcame by carbon dioxide fumes and drowned p A 47 year old welder, 36 year old yard helper, 46 year old lead hand and a 38 year old shop foreman entered watertight void space of barge and was overcame by oxygen-depleted air ( Anoxia) q A 52 year old production worker died in chemical explosion r A worker died by exposure to cedar dust which resulted in cedar dust asthma, he died of respiratory failure s A 50 year old project manager died by oxygen-deficiency In a sampling shed  t A 44 years old and a 21years old paramedics died in sampling shed because of oxygen-deficient air u A 48 year old technical engineer died in a sampling shed by oxygen-deficiency v A 25 year old operator died when equipment failed and released uncontrolled flow of H2S w A 55 years old farm worker was working in pumphouse of brown water containment pond. Pipe came apart and sludge released, creating an oxygen- depleted , hazardous atmosphere X A 60 years old equipment operator tumbled into a water-filled pit. Worker was trapped in cab and drowned Y A 60 years old carpet cleaner died while two persons cleaning carpets using a gas engine powered cleaning unit (compressor). Cause of death: carbon monoxide poisoning z A 32 years old vacuum truck operator died while a storage tank exploded. Worker was thrown against an adjacent building. zz 75 years old Pipefitter died of exposure to chlorine gas which caused asthma           Appendices 175  Table A-6. Oxygen deficiency fatalities in BC 1997-2009 in all sectors (WorkSafeBC, 2010 (b), data are exztracted by permission)  The sector 1997 1998 1999 2000 2001 2002 2003 2004 2005  2006  2007  2008  2009 Number of fatalities Agriculture Jobs, not Labourers            6  6 Art, culture technical jobs 1       1      2 Childcare; home support       1       1 Construction Trades          1    1 Forestry, mine ,fish not labourer    1   1 1 1     4 Health technical occupations          2    2 Other managers          1    1 Other sales and service  1   1   1      3 Other Trades  3  4     1 2 3   13 Primary production labourers      1        1 Private Service Jobs           1   1 Manufacturing supervisors       1       1 Manufacturing machine operator 1    1 1 1 1      5 Registered nurses  1            1 Science professionals 1     1        2 Science technical jobs      1    2    3 Trade, transport, contractor, supervisor 2      1       3 Trade, transport labourers 1 1 1    1       4 Transport Equipment Operators          1    1 Total              55 Appendices 176  Appendix B Calculating Oxygen-depletion in the Sump by Oxygen-Deficient Water (No Leakage)  B.1. No Leakage of Air in to the Shed and Sump.  )/1200C 33.33+33.33)C-(1200 (= 'C : )/()()/(*)/( ))(/(/ ** 369.210.2095*102 * /451493305.093)EXP(1799/2*0.0001*9.73 min/00538.0/3.32)293( 1/0.5 0.5)*1.2*1.2)/(2*(2 A/V 4800m 4.81.2*2*2=sump in the Air volume 12002.12.1*2*5.0 5.2 2.1*2 5.0 33.33min min/33.33/2000 1- tat time 1)C(with   volumeflow water new with mixingafter  )/( C 1- tat time 1)C(with   volumeflow water new with mixing before )/('C )/(* )/( )/( % /429.1)( 29320 102 01-t1-t ' ' 3 3 2 021-t 021-t 2 2 20 2 2 0 1 01 flowwaternewbyDilution tVAKCPKLnVAKdtCPKdC CPKVAKdtdC PKC kPaP XPP kPaLmgK mmhrcmKK L LmsumptheinwaterofVolume mdepthSump mareaSump msumptheinlevelWater LuteoneinflowWater LhrLrateflowWater LmgsumptheinwaterinOdissolved LmgsumptheinwaterinOdissolved LmgttimeationconcentratOdissolvedSaturatedC LmgttimeatsumptheinwaterinODissolvedC LmgsumpthetoflowingwaterinOdissolvedInitialC ttimeatairtheinOX LgO KCT kPaP t t t t C CgtH t L C C gtH gtHL gtHt gt gtgt H L t t gt =−==− −= = == = −== == == == === = = = = == = = = == = − − ∫∫ ρ  Appendices 177  4800(L))*(g/L)100/(1.429* t(g)in time sump in theO Remaining=(%)X min)each in  water by the taken (O - t)in timeair  sumpin  (O  t(g)in time sump in the O Remaining 43.14401.429*0.21 *4800one in timeair  in the O 1.429*X*4800=(g) t in timeair  sumpin  O 1200*)/1000)C-((C=(g)min each in  water by the taken O /')1(* )*('* )/()*( )'*( )/()( )'( )/()()'( 2gt 222 2 gt2 1-tt2 )/( 1 )/( )/( 1 1 1 1 = == +− = −=−= = − − = = − − = =−−−= − − − − − g Lmg e CeCC CCeCC tVAK CC CC Ln tVAK CPK CPK Ln tVAKCPKLnCPKLn tVAK t tVAK t t tt tVAK tt L tt tt L tgtH tgtH LtgtHtgtH L L L  B.2. Considering Gas Flow from the Pipe into the Sump  1.-B similar to isn calculatio  theofrest  The X Given the VX = V X  V X (%). pipe,  thefrom sump  theinto blowing gas in thecontent  O =X (%).  t,at time gas  with themixesair after  sumpin content  O =X (%).  t,at time gas  with themixesair  before sumpin content  O =X  volumeSump = V )(m t at time  gas of  volume(s), 60 * Q = V /s)(m t at time rate flow Gas ,Av = Q (m/s) t at time velocity Gas = v )(m 0.125 =A gt sumpgtptptsumpg0t 2pt 2gt 2g0t sump 3 tpt 3 tt t 2 +   Appendices 178   Appendix C Atmospheric Confined Space Manual (Gasses from Mines and Soils) Note: Format of this manual is borrowed from BHP-Billiton safety manual.  C.1. Atmospheric Confined Space Manual While all of details given in all regulations are useful none of the regulations are a complete source for all of the information. In this section the important parts of the existing regulations are reviewed and combined together to form a document which is as complete as possible to help with identification of confined spaces. Also, the possible gasses that were recognized in this section and can create a hazardous atmosphere at mine reclamation sites are included in this document. Some parts of this manual that have been described previously, are repeated in the manual mainly due to the importance that manual has to stand on its own. Also, the format of the manual is different from the rest of thesis so that one can take this part of the thesis and use it in their work place.  C.1.1. Intent To eliminate the risk of fatalities and injuries arising from confined spaces in surface and underground mining as well as transport, storage, production, disposal and reclamation activities.  C.1.2. Application This manual is written to prevent fatalities in confined spaces that involve exposure of humans to hazardous atmosphere that can cause immediate death. Typically, these accidents involve asphyxiation due to the emission of a hazardous gas into an unventilated enclosed structure.   This protocol applies to all controlled activities at a mine site and to all its employees, contractors, and visitors when involved in controlled activities. C.1.3. Definitions A bump test Uses a known concentration to verify if an instrument is responding correctly. The equipment shall be calibrated and bump-tested according to the frequency specified by the manufacturer (calibration may be needed every 30 days while bump-testing is typically done daily at the start of the shift) (Confined Space Entry Program- A Reference Manual, 2005, by permission).  Hazards in a confined space Confined spaces may contain a hazardous atmosphere (toxic gasses, asphyxiants, poisoning gasses and agents (inhalation/skin) and oxygen-deficiency) and mechanical hazards (entrapment and engulfment in unstable materials and/or moving parts of equipment). Confined spaces may contain other hazards such as non-ionizing and ionizing radiation, electric shock/electrocution, noisy environment, fire/explosion, head impact, heat stress, constrained movement, eye impact, poor visibility, cold stress, falling objects, drowning and hazards such as slips, trips or falls, materials release through the lines, electrical or mechanical machinery Appendices 179  activation can cause injury to the workers, objects falling (Confined Space Entry Program- A Reference Manual, 2005; Confined Space Hazards, Part III, 1979; Hazards of Confined Spaces, 2004).   Confined space A confined space is one that is not designed for human occupancy for permanent or extended periods of time. Most of the time entering and exiting such a space is restricted or difficult and is typically entered temporarily on regular or irregular basis to complete a particular work task (OSHA, 2004). Confined space refers to any enclosed area or structure that prohibits the natural ventilation in which the possibility of a dangerous atmosphere exists or can occur after the work starts. When a hazard is added to an enclosed structure like this, it will become extremely dangerous permit-required confined space. Enclosed structure is any space that is open in maximum three side (out of the 6 sides). A room with a closed door and windows is closed in all sides while a trench which is open in two sides and the top has only three closed sides.                                  Any structure built on top of a confined space or located adjacent to a confined space may contain the same hazards as confined space. Therefore, it is necessary that all confined structure and structures connected to them be examined for hazard (especially atmospheric hazard) evaluation. This type of structure is simply called an enclosed structure in this thesis and manual.   A place having any of these characteristics is a confine space: Limited entry and exit due to size or locations The size of the openings is as small as 18 inches in diameter, and is difficult to move through easily and it may contradict with getting the needed equipment such as respirators, life-saving equipment in or out of the spaces when rescue is needed. In some open-topped confined spaces such as pits, degreasers, excavations, and ships’ holds openings may be very large. Access to these spaces may require the use of ladders, hoists, or other devices, that make the escape from such places very difficult in the case of emergency (Pettit and Linn, 1987). This term is defined by WorkSafeBC as a place which does not need to be small, tight or fully enclosed space to be considered a confined space, it can be large or small and may not be enclosed on all sides and be partially enclosed, workers may be able to move freely inside the space (Confined Space Entry Program- A Reference Manual, 2005).   Not designed for continuous worker occupancy Confined spaces are designed mostly to store the product, enclose materials and processes or transport products or substances rather than for the worker to enter and work in them in the routine basis. Therefore they contain danger due to chemical or physical hazards to workers who occasionally enter them for inspection, maintenance, repair, cleanup, construction or sampling (Pettit and Linn, 1987). This often means the space is not normally ventilated (Confined Space Entry Program- A Reference Manual, 2005). Some confined spaces like the sampling shed in Sullivan mine are entered "in regular basis". Permit-required confined space A permit-required confined space is a place that has any one of the characteristics of a confined space (plus) within which the contained atmosphere can become dangerous to human health mostly Appendices 180  because the air may not move freely due to the design and does not have natural ventilation: lack of oxygen leading to asphyxiation; high oxygen- level leading to fire; emission of a toxic, flammable, or explosive gas; emission of air contaminants (fumes, dusts, or mists) that can pose an immediate threat to life or interfere with a person's ability to escape unaided from the space; threat of engulfment from surrounding bulk- material and/or water flooding into the space or collapse of the side walls or any other hazard exists that can cause serious burning, poisoning, engulfment, fracture, damage of the occupant. Accumulation of water can occur or in such a space can drown a human occupant, or water can flow in to cause consumption of oxygen by dissolution in the water or emit other dissolved hazardous gasses. Change in climatic conditions can also be added to those that can transform a non-permit required confined space to a permit required one. When such hazards are identified, entry of a human into the space requires a permit and specific protocols must be followed to eliminate or minimize the threat (changed from OSHA, 2004 (some of the hazards have been added to this definition)). The drop in barometric pressure may cause the gasses to come into the confined spaces from the surrounding walls. These kinds of manholes may have been entered previously without any problem (Pettit, 1994). Permit required confined spaces might experience temporal changes due to pressure and temperature fluctuations throughout the year (different seasons, months, weeks or even days and nights). Temperature and pressure changes (caused by climatic conditions) control gas emission direction into/out of the enclosed area in different times. Therefore, the confined spaces can change between safe and unsafe conditions all t