UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An accidental release of flowback water from storage system : a risk assessment study Gandhi, Himani Rajiv 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_february_Gandhi_Himani.pdf [ 2.76MB ]
Metadata
JSON: 24-1.0357977.json
JSON-LD: 24-1.0357977-ld.json
RDF/XML (Pretty): 24-1.0357977-rdf.xml
RDF/JSON: 24-1.0357977-rdf.json
Turtle: 24-1.0357977-turtle.txt
N-Triples: 24-1.0357977-rdf-ntriples.txt
Original Record: 24-1.0357977-source.json
Full Text
24-1.0357977-fulltext.txt
Citation
24-1.0357977.ris

Full Text

 AN ACCIDENTAL RELEASE OF FLOWBACK WATER FROM STORAGE SYSTEM: A RISK ASSESSMENT STUDY by  Himani Rajiv Gandhi  B.E. Civil Engineering, Birla Vishwakarma Mahavidhyalaya, 2012  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF                                                        THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF APPLIED SCIENCE in THE COLLEGE OF GRADUATE STUDIES (Civil Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Okanagan)  November, 2017 ©Himani Rajiv Gandhi, 2017  ii  EXAMINATION COMMITTEE The undersigned certify that they have read, and recommend to the College of Graduate Studies for acceptance, a thesis entitled:  AN ACCIDENTAL RELEASE OF FLOWBACK WATER FROM STORAGE SYSTEM: A RISK ASSESSMENT STUDY  Submitted by                       Himani Rajiv Gandhi                     in partial fulfillment of the requirements for the degree of                   Master of Applied Science                        .   Kasun Hewage Supervisor, Professor   Rehan Sadiq Co-Supervisor, Associate Dean and Professor   Dwayne Tannant Supervisory Committee Member, Professor   Kevin Hanna University Examiner, Professor        iii  ABSTRACT In the last decade, extraction of unconventional oil and gas (UOG) has rapidly increased in Canada. One of the major concerns related to UOG extraction is the risk of accidental releases during wastewater management. UOG extraction uses technologies of horizontal drilling and hydraulic fracturing. Flowback and produced water are generated as wastewater and may contain high concentrations of salts, metals, oil, grease, and organic compounds. Flowback water is stored on site before being transported for treatment, reuse or disposal. It is stored in containment pond, above ground walled storage systems (AWSS) or storage tanks. A comprehensive risk assessment has been carried out for the accidental release of flowback water during the storage. Two components of risk namely probability of failure and consequence assessment on the ecology have been examined using the frameworks of Backward Integrated Analysis (BIA) and Ecological Risk Assessment (ERA), respectively. In BIA, failure modes were identified for an uncontrolled release of flowback water due to AWSS failure by developing a fault tree. The probability of failure of the system was calculated and its failure modes were ranked by assigning risk priority number (RPN). To assess the consequence of the accidental release, the toxicity and exposure of the flowback water components to the aquatic ecology were examined through ERA. Toxicity of each constituent of flowback water was assessed by developing species sensitivity distribution curves. An exposure model using dilution factor and adsorption coefficient of the flowback water constituents is proposed and risk quotient was used to characterize ecological risk. To demonstrate the methodology, a case study in Montney unconventional play in Northern BC was carried out. The risk to the aquatic ecology was found to be very low, however, scenario analysis and uncertainty analysis prove that the risk cannot be completely overlooked. A review of the regulations for storage systems was carried out and they were assessed in light of the results of the study. Enforcing regulations pertaining to the quality of water stored, citing of the storage system with respect to the water body and making secondary system mandatory were realized to be the most beneficial.    iv  PREFACE All the frameworks development was conducted by Mrs. Gandhi under the supervision of Drs Rehan Sadiq and Kasun Hewage. Portions of this research have been submitted to a scientific journal for publication. The papers are written by Himani Gandhi under the supervision of Drs. Rehan Sadiq and Kasun Hewage. Dr. Guangji Hu is also an author on these paper due to his contribution in the review of the papers. Parts of Chapter 2, Chapter 3 and Chapter 4 of this thesis are:  • Accepted for publication on September 24, 2017: Gandhi, H., Sadiq, R., Guangji, H., & Hewage, K. (2017). Ecological Risk Assessment of Accidental Release of Flowback Water: A Conceptual Framework. Human and Ecological Risk Assessment: An International Journal, Taylor and Francis.  • Working Paper: Gandhi, H., Sadiq, R., Guangji, H., & Hewage, K. (2017). Assessing failure of the above ground walled storage systems for UOG production: An effectiveness of the provincial regulations in Canada.     v  TABLE OF CONTENTS ABSTRACT .........................................................................................................................................iii PREFACE ............................................................................................................................................ iv TABLE OF CONTENTS .................................................................................................................... v LIST OF TABLES ............................................................................................................................viii LIST OF FIGURES ............................................................................................................................ ix LIST OF SYMBOLS AND ABBREVIATIONS ............................................................................... x ACKNOWLEDGEMENTS .............................................................................................................xiii DEDICATION ................................................................................................................................... xiv CHAPTER 1. INTRODUCTION ....................................................................................................... 1 1.1 Background................................................................................................................................ 1 1.2 Research Objectives .................................................................................................................. 2 1.3 Thesis Outline ............................................................................................................................ 3 CHAPTER 2. LITERATURE REVIEW ........................................................................................... 4 2.1 Unconventional Gas Industry in Canada ................................................................................... 4 2.1.1 Global resource overview............................................................................................ 5 2.1.2 Canada’s resource overview ........................................................................................ 5 2.1.3 British Columbia’s resource overview ........................................................................ 7 2.2 Hydraulic Fracturing ................................................................................................................. 8 2.2.1 Impacts on air ............................................................................................................ 10 2.2.2 Impacts on land ......................................................................................................... 11 2.2.3 Human health and social impacts .............................................................................. 12 2.2.4 Impacts on water ....................................................................................................... 13 2.3 Flowback Water....................................................................................................................... 15 2.3.1 Flowback water quantity ........................................................................................... 15 2.3.2 Flowback water quality ............................................................................................. 16 2.3.3 Flowback water management and potential spill routes ........................................... 18 2.4 Risk Assessment ...................................................................................................................... 23 2.4.1 Risk assessment studies for unconventional oil and gas development ..................... 24 2.4.2 Assessment of probability of failure ......................................................................... 27 2.4.3 Assessing adverse environmental effects .................................................................. 30 vi  2.5 Regulatory Perspective ............................................................................................................ 32 2.5.1 National overview ..................................................................................................... 32 2.5.2 British Columbia overview ....................................................................................... 35 2.5.3 Voluntary best practices in the US ............................................................................ 35 2.6 Summary ................................................................................................................................. 43 CHAPTER 3. RESEARCH METHODOLOGY ............................................................................. 44 3.1 Overview of Methodology ...................................................................................................... 44 3.2 Backward Integrated Analysis (BIA) Framework ................................................................... 44 3.2.1 Fault tree analysis (FTA)........................................................................................... 46 3.2.2 Failure mode and effect analysis ............................................................................... 50 3.3 Ecological Risk Assessment (ERA) Framework ..................................................................... 53 3.3.1 Problem formulation ................................................................................................. 54 3.3.1.1      Assessment endpoints ................................................................................. 54 3.3.1.2      Conceptual model ....................................................................................... 55 3.3.1.3      Analysis plan............................................................................................... 57 3.3.2 Analysis phase ........................................................................................................... 58 3.3.2.1      Analysis of effects ...................................................................................... 58 3.3.2.2      Analysis of exposure ................................................................................... 59 3.3.3 Risk characterization ................................................................................................. 65 3.3.4 Uncertainty, sensitivity and scenario analyses .......................................................... 67 3.3.5 Risk description ......................................................................................................... 69 CHAPTER 4. CASE STUDY ............................................................................................................ 70 4.1 Study Area ............................................................................................................................... 70 4.2 Data Collection ........................................................................................................................ 72 4.2.1 Data for BIA framework ........................................................................................... 72 4.2.2 Data for ERA framework .......................................................................................... 72 4.3 Implementation of BIA Framework ........................................................................................ 75 4.4 Implementation of ERA Framework ....................................................................................... 86 4.5 Risk Description and Discussion ............................................................................................. 95 4.6 Regulations .............................................................................................................................. 96 CHAPTER 5. CONCLUSIONS AND RECOMMENDATIONS .................................................. 98 5.1 Conclusions ............................................................................................................................. 98 vii  5.2 Limitations and assumptions ................................................................................................. 100 5.3 Research Contributions ......................................................................................................... 100 5.4 Recommended future work ................................................................................................... 101 REFERENCES ................................................................................................................................. 102 APPENDIX A: Dataset for deriving SSDs of the contaminants .................................................. 117    viii  LIST OF TABLES Table 2-1 An estimate of top 5 countries with the highest recoverable unconventional natural gas    resources (Wang et al., 2016) ....................................................................................................... 5 Table 2-2 Total marketable gas from unconventional sources (Heffernan & Dawson, 2010) ..................... 6 Table 2-3 Contaminants produced from unconventional activities and effects on human health  (Health      Canada, 2017; US EPA, 2016c) ................................................................................................. 13 Table 2-4 Water use for hydraulic fracturing per well and percentage total water recovered .................... 16 Table 2-5 Comparision of range of concentration of constituents of flowback and produced water for Conventional Natural Gas, Coalbed Methane, Shale Gas and Tight Gas (Alley et al., 2011) ... 18 Table 2-6 Basic terms used in FTA and their definitions (CCPS, 2000b; Ostrom & Wilhelmsen, 2012; Rausand, 2011e; US NRC, 1981) ............................................................................................... 28 Table 2-7 Basic terms used in FMEA and their definitions (Carlson, 2012; Ericson II, 2005a) ................ 30 Table 2-8 Acts and regulations pertaining to flowback water storage systems in BC (BCOGC, 2015b; Ernst and Young, 2015) .............................................................................................................. 37 Table 2-9 Guidelines for flowback water storage systems in BC (BCOGC, 2009, 2015b) ........................ 39 Table 2-10 Voluntary best storage practices followed in US  (US EPA, 2014) ......................................... 41 Table 3-1 Commonly used fault tree symbols (CCPS, 2000b; Ericson II, 2005b; US NRC, 1981; Vesely, 2002) ........................................................................................................................................... 48 Table 3-2 Failure rate and correspoonding probability values and assigned ranking (Chang et al., 2001;Pillay & Wang, 2003) ........................................................................................................ 51 Table 3-3 Severity, Occurrence and Detectability scale assigned for FMEA ............................................. 53 Table 4-1 Contaminants and their concentrations found in flowback water samples collected in Montney .................................................................................................................................................... 73 Table 4-2 Background concentrations and monthly discharge rate of the Flatbed Creek (BCOGC, 2016)74 Table 4-3 Physicochemical properties of the contaminants (Bright & Addison, 2002; US EPA, 1988, 1999, 2005, 2016f) ..................................................................................................................... 74 Table 4-4 Description and probability of fault tree events ......................................................................... 78 Table 4-5 FMEA table for basic failure events determined through FTA .................................................. 81 Table 4-6 Probability of failure calculation ................................................................................................ 85 Table 4-7 Effect concentrations of contaminants........................................................................................ 87 Table 4-8 Effect of Kd and dilution coefficient on exposure concentration ................................................ 91 Table 4-9 Estimated exposure concentration and risk quotient of contaminants ........................................ 91 Table A-1 Data points for deriving toxicity profiles of the contaminants ................................................ 117 ix  LIST OF FIGURES Figure 1-1 Objectives, sub-objectives and methodology of the assessment ................................................. 3 Figure 2-1 Ultimate marketable recoverable shale and tight gas in Canada (ARI, 2013; NEB, 2016) ........ 6 Figure 2-2 Unconventional (UOG) and conventional (COG) well production stages .................................. 9 Figure 2-3 Water acquisition, use and wastewater management for UOG extraction process ................... 19 Figure 2-4 Containment pond and its components  (US EPA, 1988a) ....................................................... 22 Figure 2-5 Above ground walled storage system and its components ........................................................ 22 Figure 2-6 Typical components of risk management .................................................................................. 24 Figure 3-1 Proposed methodology for assessing risk to the aquatic ecology due to accidental release of flowback water .......................................................................................................................... 44 Figure 3-2 Backward Intergrated Analysis (BIA) framework .................................................................... 46 Figure 3-3 Steps required to perform fault tree analysis  (CCPS, 2000b; Ericson II, 2005b)..................... 47 Figure 3-4 Example use of AND and OR gate ........................................................................................... 51 Figure 3-5 Ecological Risk Assessment framework for flowback water,  adapted from (US EPA, 1998) 56 Figure 3-6 Conceptual model for flowback water release from storage and contamination pathways  ..... 57 Figure 3-7 Minimum secondary data requirement for short-term exposure in freshwater and applicable guidelines (Adapted from (CCME, 2012)) ............................................................................... 60 Figure 3-8 Exposure assessment model proposed for inorganic constituents of flowback water ............... 66 Figure 4-1 UOG play trends and Flatbed Creek in BC ............................................................................... 71 Figure 4-2 Fault tree for uncontrolled release of flowback water from AWSS failure............................... 77 Figure 4-3 a-f Cumulative distribution curves for flowback water constituents ......................................... 90 Figure 4-4 Results of uncertainty analysis for risk quotient ....................................................................... 93 Figure 4-5 Sensitivity analysis of risk quotient for effect concentration and initial exposure concentration .................................................................................................................................................. 94 Figure 4-6 Variation in cumulative risk quotient as a function of the ratio of spill volume to creek discharge ................................................................................................................................... 94     x  LIST OF SYMBOLS AND ABBREVIATIONS 𝜆   Component failure rate per time t 𝜌  Soil bulk density AER   Alberta Energy Regulator 𝐴𝑖  Amount of adsorbate  ANZECC  Australian and New Zealand Environment and Conservation Council ARMCANZ  Agriculture and Resources Management Council of Australia and New Zealand ARI   Advanced Resources International 𝐴𝑠𝑜𝑖𝑙  Area of contaminated soil AWSS  Above Ground Walled Storage System 𝐵𝐴𝐹   Bioavailable fraction BC  British Columbia BC MOE  British Columbia Ministry of Environment Bcf/day  Billion cubic feet per day BCOGC  British Columbia Oil and Gas Commission BIA   Backward Integrated Analysis 𝐶∞  Final contaminant concentration in receiving water body CBM   Coalbed Methane CCA   Council of Canadian Academies CCME  Canadian Council of Ministers of the Environment CCREM  Canadian Council of Resource and Environmental Ministers CEAA  Canadian Environmental Assessment Agency CEQG  Canadian Environmental Quality Guidelines 𝐶𝑖  Concentration remaining in solution after adsorption 𝐶𝑜  Concentration of contaminant in spilled water COG  Conventional Oil and Gas COGDPR  Canadian Drilling and Production Regulation COGOA  Canadian Oil and Gas Operations CSUG  Canadian Society of Unconventional Gas 𝐶𝑤  Background contaminant concentration in receiving water body xi  CWQG-PAL  Canadian Water Quality Guidelines for Protection of Aquatic Life 𝑑  Distance between AWSS and water stream 𝐷   Detection 𝐸𝐶50  Effect Concentration 50 𝐸𝐶𝑛  Effect concentration of the 𝑛𝑡ℎcontaminant EDF   Empirical Distribution Function 𝐸𝐸𝐶𝑛  Estimated exposure concentration of the 𝑛𝑡ℎ contaminant EMA   Environmental Management Act ERA   Ecological Risk Assessment FIA   Forward Integrated Analysis FMEA  Failure Mode and Effects Analysis FTA   Fault tree Analysis GHG   Green House Gas GWPC  Groundwater Protection Council IOGCC  Interstate Oil and Gas Compact Commission 𝐾𝑑  Adsorption Coefficient km   Kilometer L   Litre 𝐿𝐶50  Lethal Concentration 50 log 𝐾𝑜𝑤 Log octanol water partition coefficient m3   cubic meter mD   Millidarcy mg/L   milligram per litre Mm3/day  Million cubic meter per day 𝑀𝑠𝑜𝑖𝑙  Mass of soil contaminated  NEB   National Energy Board NORM  Naturally Occurring Radioactive Material NRCAN  Natural Resource Canada 𝑂   Occurrence OGAA  Oil and Gas Activity Act 𝑝   Probability of exposure xii  𝑃   Probability of failure PI   Prediction Interval 𝑄𝑤   Creek discharge  𝑄𝑠  Flowback water discharge       𝑅  Risk quotient 𝑅𝑃𝑁   Risk Probability Number 𝑆   Severity  SSD   Species Sensitivity Distribution 𝑇   Component exposure time tcf   Trillion cubic feet tcm   Trillion cubic meter TDS   Total Dissolved Solids UOG   Unconventional Oil and Gas US EPA  United States Environmental Protection Agency USGS   United States Geological Survey VOC   Volatile Organic Compound 𝑉𝑠  Volume of flowback water spilled       xiii  ACKNOWLEDGEMENTS This research would not have been accomplished without the support and guidance of several people and organizations. First and foremost, I would like to thank my loving husband, Aakash, for his unrelenting support, patience and for providing me the strength to complete this research.  I feel extremely privileged to have accomplished my research under the supervision of Dr. Rehan Sadiq. I would like to thank him for his constant encouragement, guidance and unfailing faith in me. Without his support, this research would not have been possible. I am grateful to Dr. Kasun Hewage for giving me the opportunity to carry out this research, his help and support is significant. I would also like to thank Dr. Guangji Hu for his constructive criticism. I would like to extend my thanks to Mitacs Natural Gas Strategic Research Initiative for funding the research project. Special acknowledgement is due to Greg Dickie and Mike Pittman at Secure Energy Services for their valuable inputs and feedbacks to the project. Many thanks to John Nurkowski at the British Columbia Oil and Gas Commission for his help in data collection.  I would also like to thank Kelsey McAuliff for being an inspiration and a reliable research partner. I am extremely grateful for her friendship. Finally, I would like to offer enduring gratitude to my parents and my sister for moral and financial support and for encouraging me to pursue research.   xiv  DEDICATION  For my father1  CHAPTER 1. INTRODUCTION 1.1 Background The hydrocarbon production from shale or tight formation has been rising in Canada since 2005 and is expected to present 80% of Canada’s natural gas production by 2035 (NRCAN, 2016b; Rivard et al., 2014). For economic extraction of oil and gas from these formations, directional drilling, and hydraulic fracturing is used. These processes use water in huge quantities and produce a large volume of wastewater. The water produced from well immediately after hydraulic fracturing is known as flowback water (Jiang et al., 2014; Kondash & Vengosh, 2015; Ziemkiewicz et al., 2014). The flowback water may contain salts, metals, oil, greases, volatile and semi-volatile organic compounds (Gregory et al., 2011). Since a large volume of flowback water is generated and has high concentration of dissolved solids and a complex physicochemical composition, the flowback water creates potential risk to human health and the environment in an event of accidental or uncontrolled release into the environment (Gregory et al., 2011; Kargbo et al., 2010; Lester et al., 2015; Vengosh et al., 2014).  Flowback water is stored on site before being transported for reuse or disposal. Historical data show that majority of the flowback water release incidents constitute about 46% by a number of spills and 75% by volume of spills, have been reported during storage (US EPA, 2015). Various studies have pointed out that the accidental release of the hydraulic fracturing fluid chemicals and generated flowback water may affect surface water resources (Barbot et al., 2013; Becklumb, 2015; Burton et al., 2014; Ewen, Borchardt et al., 2012; Gagnon et al., 2016; Rozell & Reaven, 2012; Torres et al., 2015; Vengosh et al., 2014; Ziemkiewicz et al., 2014).  It has become a pressing concern to assess the integrity of the storage systems used for flowback water, investigate the causes to reduce the probability of its failure, and analyze the ecological consequences if it happens. The uptake of the flowback water constituents by the aquatic organisms can cause lethal and sub-lethal effects to the native species resulting in serious environmental consequences (Siegel, 2007; US EPA, 2007). Thus, it is critical to assess the risks 2  to the receiving aquatic system and to identify areas requiring mitigative actions in lieu of the growing unconventional oil and gas (UOG) industry. 1.2 Research Objectives The main objective of this study is to quantify the risk to the aquatic ecosystem caused by the accidental release of flowback water due to storage system failure. The objective of quantifying the risk is achieved through assessing the two major components of risk, the probability of failure and consequences of failure. They have been assessed using Backward Integrated Analysis (BIA) and Ecological Risk Assessment (ERA) frameworks respectively. These frameworks use techniques such as fault tree analysis (FTA), failure mode effect analysis (FMEA), regression based species sensitivity distribution (SSD) curves, dilution and distribution coefficient, risk quotient and Monte Carlo simulations to assess the risk as shown in Figure 1.1. BIA is used to calculate the probability of the failure of the storage system and rank the failure modes. ERA helped to assess the exposure and effect of the flowback water contaminants on the aquatic ecology. The probability is input in the ERA framework to estimate the risk. Uncertainty, sensitivity and scenario analysis is done to obtain the full range of risk estimate, to determine the parameters contributing to the uncertainty and derive the relationship between spill volume and receiving aquatic body discharge. This method is applied to a creek in Montney unconventional play in Northern BC, Canada to assess its effectiveness.  The results, thus computed, will help to identify the failure modes, determine the probability of failure of the system, rank the failure modes, assess the toxicity of the flowback water, quantify exposure to the aquatic ecosystem, and characterize the overall risk to the system. The framework will provide useful information to facilitate in devising guidelines for flowback water management to reduce the risk to the aquatic ecology.   3   Figure 1-1 Objectives, sub-objectives and methodology of the assessment  1.3 Thesis Outline This thesis consists of five chapters. This first chapter briefly introduces the problem and outlines the objectives of this research. Chapter 2 reviews the literature of UOG industry in Canada, flowback water quality and volume generated, adverse environmental effects, risk assessment frameworks and techniques, and flowback water storage regulations. Chapter 3 proposes the methodology to assess risk using BIA and ERA frameworks and discusses techniques used in each framework. Chapter 4 includes a hypothetical case study to demonstrate the developed methodology. The hypothetical case study includes data collection, site description, identification of failure modes and the assessment of the ecological risk. The results are discussed and assessed in lieu of the current provincial regulations. Chapter 5 provides limitations, conclusions, and recommendations for future research.    4  CHAPTER 2. LITERATURE REVIEW 2.1 Unconventional Gas Industry in Canada Canada ranks fifth in the production of natural gas on a global scale and accounts for five percent of the global production (NEB, 2017b). Current natural gas supply far exceeds the domestic demand allowing export of the surplus gas which makes Canada fourth largest exporter of natural gas in the world (NRCAN, 2015). It is the common fuel of choice in applications related to residential, commercial, and industrial as well as for electricity generation (NEB, 2009). Natural gas makes important contributions to the economy and labour force of Canada (CAPP, 2017).  Natural gas resources can be categorized into conventional resources and unconventional resources. Classification is made based on the permeability of the reservoirs in which the gas is entrapped. Reservoirs having a permeability of more than 1 millidarcy (>1 mD), can produce resources at economical flowrates and volumes using traditional techniques like vertical drilling into the pools of oil and gas and extracting them via pumping. The oil and gas resources extracted in such a manner are conventional resources (McGlade et al., 2013; Speight, 2013). While unconventional gas resources are those that are extracted from relatively low permeability reservoirs (<1 mD) requiring special techniques, such as hydraulic fracturing, horizontal drilling, and multilateral well bores, for economic extraction (McGlade et al., 2013). NEB (2017a) defines unconventional gas as: “Natural gas that can be produced at commercial rates only after the extensive use of technology. It may be that the gas is held by the matrix material such as coal, ice, or shale; or where the reservoir has an unusually low amount of porosity and permeability.” The examples of unconventional gas resources include tight gas, coalbed methane (CBM), and shale gas. Tight gas is commonly found in limestone or sandstone, CBM is found entrapped in coal seams, while shale gas is found in shale sedimentary rock (McGlade et al., 2013).   5  2.1.1 Global resource overview In 2015, the total unconventional oil and gas (UOG) global production accounted for 9% and 27% of the annual oil and gas production. These resources are having a profound impact on the global supply and demand structure and are becoming alternative to the conventional resources (Wang et al., 2016).  Wang et al. (2016) systematically estimated technically recoverable UOG to be 442.1 billion tons and 227 trillion cubic meters (tcm) respectively, globally. The estimate includes shale gas, tight gas, and CBM, which account for 71.1%, 7%, and 21.7% respectively. The top five countries with the largest unconventional resource are shown in Table 2.1. Technically recoverable is an estimate of the volume of oil or gas that could be recovered using current technology, without accounting for the economic conditions (NRCAN, 2016). Canada ranks fourth in terms of the recoverable unconventional natural gas in the world and contains 7% of the global resource, as shown in Table 2.1. Table 2-1 An estimate of top 5 countries with the highest recoverable unconventional natural gas resources (Wang et al., 2016) Country Recoverable unconventional natural gas (tcm) % of total global resource United States 39  17.4 China 31  13.9 Russia 29  12.6 Canada 16  7.0 Australia 16  6.4  2.1.2 Canada’s resource overview Canada has 16 trillion cubic meters (tcm) of recoverable unconventional gas reserves found in the form of tight gas, shale gas, and CBM (Cherry et al., 2014; Chong & Simikian, 2014; Wang et al., 2016). Canadian society of unconventional gas (CSUG) has provided an estimate of marketable gas in Canada for coalbed methane, tight gas, and shale gas. The estimate is shown in Table 2.2. They have defined marketable gas as (Heffernan & Dawson, 2010): “The volume of gas that can be sold to the market after allowing for removal of impurities and after accounting for any volumes used to fuel surface facilities” 6  Table 2-2 Total marketable gas from unconventional sources (Heffernan & Dawson, 2010) Unconventional gas resource type Estimate Range (Tcm) Coalbed Methane 1-3.6 Tight Gas 6.1-13.5  Shale Gas 3.6-9.7   Most of the estimated unconventional gas resources are tight gas and shale gas. Province wise distribution of the ultimate marketable recoverable gas in Canada is shown in Figure 2.1. The total unconventional gas in British Columbia consists of about 66% of the total estimated reserves, followed by Alberta, Northwest territories, Quebec and Yukon at 23%, 6%, 4% and 1% respectively (ARI, 2013; NEB, 2016). Major shale and tight gas plays include the Horn River Basin and Montney Play Trend in British Columbia, the Cretaceous Colorado Group in Alberta and Saskatchewan, the Utica Shale in Quebec, and the Horton Bluff Shale in New Brunswick and Nova Scotia (NEB, 2009).    Figure 2-1 Ultimate marketable recoverable shale and tight gas in Canada (ARI, 2013; NEB, 2016) British ColumbiaAlberta Northwest TerritoriesYukonSaskatchewan and ManitobaQuebecNova Scotia7  However, UOG extraction activities are not carried out in all provinces. It is banned in Nova Scotia, New Brunswick, Newfoundland, and Labrador until further research is done (Atherton & Macintosh, 2014; Gosine et al., 2016; NRCAN, 2016d). In Manitoba, there are thirteen active oil fields but the unconventional gas is not presently exploited (NRCAN, 2016c). Unconventional gas or oil are not being extracted anywhere in Ontario (McKinley, 2015; NRCAN, 2016f). In Quebec, 29 wells were drilled from 2006-2010 in Utica shale until a moratorium was imposed in 2013 (NRCAN, 2016f).   Alberta has drilled over 180,000 oil and gas wells using hydraulic fracturing completion. Apart from CBM and shale gas, a significant amount of natural gas liquids and oil are found in the formations. The complete unconventional oil and gas potential of Alberta is still being assessed. In 2013, CBM and shale gas contributed 8% to the total gas supply and are expected to contribute increasingly higher percentage in the coming years (AER, 2014).  2.1.3 British Columbia’s resource overview In British Columbia (BC), the unconventional portion has continued to increase since 2005. By the end of 2015, it accounted for about 80% of total gas production in BC (NRCAN, 2016a). The Montney and Horn River Basin are the most active unconventional gas plays that contributed to the daily production level of 3.4 billion cubic feet/day (Bcf/d) and 0.365 Bcf/d in 2015, respectively (NRCAN, 2016a). The Cordova Embayment and the Liard Basin shale plays in BC are in the early stages of development with predicted marketable natural gas production of 0.3106 Mm3/d (million cubic meters per day) from Cordova and 0.9106 Mm3/d from the Liard Basin by 2040 (NEB, 2016a).  These plays reside within the Western Canadian sedimentary basin in British Columbia and are chiefly gas-charged system, black oil being a smaller constituent of the overall hydrocarbon generation. Montney Basin is a mixture of tight and natural gas liquids play in the mid-Triassic, over-pressured siltstones (NRCAN, 2016b). Montney also has a marketable, technically recoverable natural gas liquid (NGL) resource potential estimated at 12.6 billion barrels (NRCAN, 2016b).   8  As of January 2017, Montney reached production levels of 127 Mm3/d represented one third of the total Canadian natural gas production. Also, due to continual application and experience gained, technological improvements have helped achieve more productivity at lower cost. These improvements include advances in geoscience and engineering, longer horizontal legs, number of fracture stages in a well, and the use of more fluid and/or proppant (NEB, 2017b).  The predicted increase in the unconventional gas production will help in filling the gap created due to decline in conventional resources. This can result in larger growth opportunities and lower natural gas prices. Despite potential economic benefits, unconventional oil and gas production has faced criticism and is banned in many provinces (Chong & Simikian, 2014). Concerns unique to the unconventional gas process circumscribe about the special completion techniques used for unconventional production like hydraulic fracturing and horizontal drilling technologies. These processes and related concerns are discussed in depth in the following sections.  2.2 Hydraulic Fracturing A typical shale formation can be a hundred metres to a few kilometres thick and may spread over extensive geographic areas (CCA, 2014). The resources are extensively spread over wide areas instead of being concentrated in definite places. Shale gas tends to remain where it was first generated and is often found as free gas trapped in pores, cracks or fissures, as adsorbed gas or as dissolved in organic matter (NEB, 2009). The pores in the shale formation can be up to 103 times smaller than in the most conventional formations. Hence, shale gas is unconventional gas form and requires special completion, stimulation, and/or production techniques, such as hydraulic fracturing, to be economically produced (King, 2012). The unconventional and conventional gas well stages are shown in Figure 2.2.  The upstream and downstream stages of the conventional and unconventional stages are similar. Site preparation, infrastructure construction, drilling, hydrocarbon production, processing, transportation and well closure stages are carried out in both conventional and unconventional production. The unconventional production required few additional steps. The drilling for unconventional gas extraction can go up to 4 km in depth.  9   Figure 2-2 Unconventional (UOG) and conventional (COG) well production stages Once the required vertical depth is reached through drilling production zone is connected to the surface, it is turned horizontal into the pay zone (also known as a kick off point). The horizontal drilling continues until 610 m to 1500 m or more. The shape of the horizontal segment can be lateral or continuously slanting (Cherry et al., 2014; King, 2012; US EPA, 2016c). The well construction is then completed by placing production casing which is cemented to the surface. In case of multi stage fracturing, the drilled hole is left open at the bottom of the well which is known as open-hole completion. On a multi well pad, the drilling rig is positioned next to the conductor pipe, and the entire process of horizontal drilling and well completion is carried out until all the wells are drilled (CCA, 2014; King, 2012).  Once the drilling is completed and casing installed, a perforation gun is inserted and brought in a predetermined position in the pay zone and small holes are shot through casing and cement. Then water, acid or propane based fluids are pumped at high pressure to create fracturing in the holes previously created within the rock surface. The tiny hair-like cracks are created, which facilitate in the flow of oil and gas through the rocks and is collected at the well head (CCA, 2014; King, 2012; US EPA, 2016c). 10  Hydraulic fracturing input fluids can be water-based, gel-based, acid-based, foams, emulsions of nitrogen, carbon dioxide, other such liquid or gas hydrocarbons (US EPA, 2016c). The most common type of hydraulic fracturing fluid is slick water formulation, which consists of a mixture of water and sand (98% - 99.9%) and three to twelve additive chemicals (2% - 0.1%) primarily friction reducers, biocides, stabilizers and corrosion preventers (GWPC, 2016). They help in preventing the corrosion of the pipe, scale downhole and in surface equipment, reduced friction, increased viscosity to suspend the proppants. The proppant sticks in between the fractures and facilitate the flow of the oil and gas to the surface. High purity quartz sand with very durable and very round grains are used as proppants (USGS, 2015). After the hydraulic fracturing stimulation, the pressure is released and the fluid is allowed to flow back to the surface. The plug is then set at the fractured location and the perforating gun shoots holes through casing and cement, the gun is then moved upwards and more holes are shot by repeating the process (Kargbo et al., 2010). These steps are unique to the unconventional process and have distinctive concerns related to it. Its impact on air, water, land, human health and society are discussed in the following sections. 2.2.1 Impacts on air Shale gas is mostly composed of methane, and its use can lead to the emission of methane and carbon dioxide, both these GHGs contribute to global warming. Methane is a potent GHG with a shorter lifetime in the atmosphere but more efficient at trapping radiation having an overall impact of more than 20 times than carbon dioxide on climate change over a 100-year period (Garvie et al., 2012). The major source of GHG emission in the shale gas well development process is the emissions caused by a large number of truck trips to deliver water, proppant, chemicals, cement, and transportation of huge volumes of flow back and produced water. A well with a single fracture needs nearly 2,000 one-way heavy commercial vehicle trips to deliver water and other supplies (CCA, 2014). Other sources of emissions include flaring or venting during drilling and well completion. However, as compared to the larger impact of the natural gas emissions during its life time, the contribution to GHG emissions from flaring completions gas is quite negligible. Methane emissions due to leakage during extraction, processing, and transport to market and methane emissions from well after abandonment also contribute to the 11  emissions. The deterioration of the cement leads to the escape of the buoyant gas from the annulus of the production well casing (CCA, 2014; Jiang et al., 2011).  2.2.2 Impacts on land The land and terrestrial ecosystem changes are caused primarily due to the development of energy resources (Northrup & Wittemyer, 2013). The related development involves extensive infrastructure for oil and gas energy source development. The infrastructure includes well pads, work camps, waste handling, pipelines, and roads. The associated infrastructure often take up more land than the well pads themselves (CCA, 2014). The oil and gas resources cover large geographical areas. However, in the case of unconventional shale development, multiple wells on a single well pad are developed. This is environmentally desirable as the land footprint is smaller (CCA, 2014).  Shale gas development needs proppant for hydraulic simulations, which has cause rise in the demand for high-quality silica, resulting in a large scale increase in sand mining. A 20,000 m3 of hydraulic fracturing injection fluid can use up to 1.5 million kilograms of proppants (King, 2012). The environmental impacts of mining contribute to dust, noise, and scarring of land (CCA, 2014). The shale gas development activities can also affect forest ecosystems by intersecting and sub-dividing the forests with infrastructure like roads, pipelines and forming transition zones between disturbed and undisturbed habitats. The high scale development in previously unexplored areas can disrupt the ecosystem and affect resource availability for wildlife (CCA, 2014; Northrup & Wittemyer, 2013; Oswald et al., 2012).  Other land use impacts include the decrease in biomass productivity due to displacement and compaction of soil, increase in erosion causing transfer of sediment and soil nutrients to streams and other water bodies, altered streamflow, loss of aquatic habitat, restricted animal passage due to road and bridge infrastructure, increased aquatic and terrestrial biota mortality and non-lethal effects because of increased hunting and fishing (Cherry et al., 2014; Northrup & Wittemyer, 2013). There is a potential threat of induced seismicity due to the fluid injection activities of the disposal wells and hydraulic fracturing processes (Rutqvist et al., 2013).  12  2.2.3 Human health and social impacts The impacts on human health can be due to the compromised groundwater, surface water quality or air quality. There are health risks to the workers due to inhalation of silica used as proppant which can cause lung cancer or silicosis. It has also been linked to diseases like tuberculosis, pulmonary and autoimmune diseases (CCA, 2014). Workers can be exposed to Naturally Occurring Radioactive Material (NORM) found in some of the flowback and produced water (CCA, 2014; King, 2012). The exposure to radioactive material like Radium-226 can cause lung, breast, thyroid, bone, digestive organs and skin cancer or leukaemia (Health Canada, 2017). A study was done in Pennsylvania to understand the impact of unconventional gas extraction activities on public health. The households within 1 km of a gas well reported higher health problems as compared to those residing more than 2 km away. Skin problems were the most common reported health related issues (Rabinowitz et al., 2015).  Some of the most common contaminants produced during unconventional activities and the potential impacts on human health due to inhalation, ingestion and/or direct consumption are shown in Table 2.3.  Often, a boomtown phenomenon occurrence is observed due to the large-scale development in a short period of time due to production activities. Skilled labours are imported to the place temporarily increasing the demands due to the growth of population putting a lot of stress on the available resources. This can have a negative impact on the human health and can adversely affect the quality of life. The nuisances to the public due to increased noise, dust, odour, traffic and visual impacts causes irritability and nausea (CCA, 2014). Large truck traffic increases congestion and disturbs the daily life activities of the people living there. There are invisible losses to the community having a cascading effect on their cultural practices leading to an identity crisis. Soil vibrations from fracturing can lead to change in water color and increased turbidity (CCA, 2014; King, 2012; Oswald & Bamberger, 2012). Psychologically there is a huge anxiety among the general population about the use of chemicals in the hydraulic fracturing process.  13  Table 2-3 Contaminants produced from unconventional activities and effects on human health  (Health Canada, 2017; US EPA, 2016c) Common Contaminants Potential Human Health Effects  Salt ions Chloride Physiological effects Bromide Renal cell tumors  Sulphate  Physiological effects Sodium Physiological effects Magnesium No evidence of adverse health effects Calcium No evidence of adverse health effects Metals Barium Increase in blood pressure, cardiovascular disease Manganese Undesirable taste, laxative effect Iron Haemochromatosis Strontium Thyriod, lung, breast, digestive system or skin cancer or leukaemia BTEX Benzene Bone Marrow changes, cancer, blood changes immunological responses Xylene Adverse neuromuscular effects Toluene Neurological disorders like vibration and auditory ailments,  colour discrimination, attention and memory disorder Radioactive material Radium 226 Thyriod, lung, breast, digestive system or skin cancer or leukaemia  2.2.4 Impacts on water Some of the major concerns related to hydraulic fracturing are the amount of water used and amount of wastewater generated. In 2014, 643 wells were fractured in British Columbia consuming 8,258,192 m3 of water (BCOGC, 2015d). Water is directly consumed in the hydraulic fracturing, drilling, cementing and site preparation stages. Other than the direct water consumption, indirect water use for the production of materials used in the extraction process involves water-intensive industries, for instance, proppant, additives, and cement (Jiang et al., 2014). A study of life cycle water footprint of hydraulic fracturing was carried out for the United States and the water use intensity was compared with other energy extraction techniques, to conclude that the unconventional water use is lower as compared to the conventional counterparts (Kondash & Vengosh, 2015). Similar conclusions were also reinforced by Kuwayama et al. (2015) and was suggested that the water quality concerns can be more severe than water quantity issues.  14  The wastewater generated during hydraulic fracturing, the fluid used for hydraulic fracturing purposes, the chemicals used as additives in the fluids can accidently spill and can potentially deteriorate the quality of groundwater and surface water besides causing a threat to aquatic/marine ecology. Various studies have been carried out to assess the impacts on water quality like Bergmann et al. (2014); Gagnon et al. (2016); Reagan et al. (2015) using toxicological risk assessment and numerical simulation techniques. The studies analyse the data gaps that does not allow a profound risk assessment to propose technical controls.  Groundwater contamination or surface water contamination can also occur due to upward migration of fracturing fluids or methane gas and pathways created by defective or missing cement seals. They can be caused by the abandoned wells due to increase in pressure, along well annulus or the wellbore. These sources of gas leaks can contaminate the fresh water aquifers (CCA, 2014; Vengosh et al., 2014a). Minor earthquakes simulated by the hydraulic fracturing or by natural causes may lead to the activation of upward gas leakage along the faults. The occurrence of fugitive gas in shallow drinking water wells could cause increase in salinity and other changes in quality of water (Vengosh et al., 2014) Spills and leaks of the wastewater generated contain high concentrations of potentially harmful organic and inorganic contaminants and need treatment before its reuse or discharge into the environment (Lester et al., 2015). The wastewater management includes activities like transportation, loading, unloading, treatment, storage, and disposal. There are challenges in the treatment of the wastewater produced as the quality of flowback water is highly variable across and within shale plays (Lester et al., 2015). The wastewater is often disposed of through deep injection well. This wastewater can percolate through the pervious layers contaminating the freshwater aquifers through leaking of injection wells (Lutz et al., 2013; Vengosh et al., 2014b). The generation of the wastewater, waste water volume and quality, its management and the potential modes of accidental release are discussed in the following sections. 15  2.3  Flowback Water Once the hydraulic fracturing simulation is completed, the water that flows back to the surface immediately after the pressure is released is known as flow back water. Jiang et al. (2014) defined flowback water as: “The water that returns from the well during the flowback period, immediately after hydraulic fracturing and before gas production” The flowback period was considered as the first 10-14 days approximately after the hydraulic fracturing completion process. Along with gas and oil, water is generated as a by-product of the life of service of well. This is known as produced water.  Produced water can be defined as (FracFocus, 2014): “The water extracted from the subsurface along with produced oil and gas, including water from the reservoir, water that has been injected into the formation, and any chemicals added during the production/treatment process”  2.3.1 Flowback water quantity About 5-75% of the injected water flows back to the surface (Jiang et al., 2011; Rivard et al., 2014; Vengosh et al., 2014; Ziemkiewicz et al., 2014). Table 2.4 gives a summary of the average water volume used for each well for hydraulic fracturing process and the total flowback and produced water recovered per well, from the unconventional plays in North America. The water recovered is expressed as the percentage of input water.  It can be seen from the table that the range of water used and water recovered is highly variable in various unconventional plays. The water use depends on factors like type of well (horizontal, lateral or vertical), length of horizontal or lateral, number of hydraulic fracturing stages and type of hydraulic fracturing fluid used (Scanlon et al., 2014). 16  Table 2-4 Water use for hydraulic fracturing per well and percentage total water recovered Unconventional Play Avg. Water use (m3) % Water recovered  References Montney, BC 14,241  50-100 (BCOGC, 2015c; Rivard et al., 2014) Horn River Basin, BC 88,634  - (BCOGC, 2014) Marcellus, PA 20,000  25 (Haluszczak et al., 2013; Jiang et al., 2011) Barnett Shale, TX 15,250  9-29 (Nicot & Scanlon, 2012; US EPA, 2016c) Woodford Shale, OK 16,000  - (Murray, 2013) Haynesville Shale, TX 21,500  5 (Nicot & Scanlon, 2012; US EPA, 2016c) Eagle Ford, TX 16,100  6-20 (Nicot & Scanlon, 2012; US EPA, 2016c) Niobrara, CO 13,000  30 (Vengosh et al., 2014)  The volume of water recovered from a well depends on the type and volume of hydraulic fracturing fluid injected, the type of liquid or gas hydrocarbon being extracted, the pressure in the formation, the reactions between the formation and injected fluid and the reactions within the reservoir. An unanticipated rise in the volume of water recovered can be due to the loss of mechanical integrity or miscommunication between well operators (US EPA, 2016c). Typically, higher rates of flowback water are recovered in the immediate weeks following hydraulic fracturing and reduce by almost an order of magnitude with time. Generally, it can be said that the volume of flowback recovered during the flowback period can be considered equivalent to the volume of produced water generated over the life of the well. This indicates large volumes of flowback water return over a period of several weeks (US EPA, 2016c). It becomes a challenge to manage the amount of water thus produced. 2.3.2 Flowback water quality The flowback water consists mainly of chemicals injected for fracturing and sometimes the formation chemicals (Vengosh et al., 2014). It mainly contains high concentrations of salts, metals, oil, grease, and organic compounds. Due to possible subsurface interactions, the composition of the flowback water depends on the injected fluid, the formation minerology, geochemistry, structure of the formation solids and flow patterns of water (Alley et al., 2011; Cherry et al., 2014).  Alley et al. (2011) conducted a literature review of wastewater constituents 17  generated from conventional and unconventional oil and gas extraction processes, as represented in Table 2.5.  Within the unconventional resources, shale and tight gas were similar, however, CBM contained a relatively low concentration of sulfate and magnesium. Also, shale gas flowback and produced water had a higher concentration of strontium (Sr), barium(Ba), and bromide(Br) (Alley et al., 2011). The variation in the quality of the generated water can be attributed to a mineralogical and geochemical characteristic of the formations. According to Health Canada (2001), total dissolved solids contains cations of calcium, magnesium, sodium and potassium and the anions of carbonate, bicarbonate, chloride, sulphate and sometimes nitrate. The concentration of the TDS is often expressed as the sum of its constituents. If the amount of TDS is less than 1000 mg/L, it is considered fresh or non-saline; 1001-3000 mg/L slightly saline; 3001-10,000 mg/L moderately saline or brackish; 10,001-100,000 saline and if greater than 100,000, then it is classified as brine (CCREM, 2008). Although there is a variation in the water salinity, flowback and produced water can be typically considered as saline (US EPA, 2016c). In addition to the factors mentioned, the quality of flowback water also depends on factors like the injected hydraulic fracturing composition, the properties of the formation, the type of hydrocarbon product to be extracted, the time of contact of the fluid with the formation, temperature, pressure and other such factors (US EPA, 2016c). Generally, the flowback and produced water constitutes salt ions and metal ions like bromide, chloride, sulfate, phosphate, nitrate, sodium, magnesium, calcium, barium, manganese, iron, strontium, zinc and radioactive materials like radium-226 and radium-228, oil, grease, BTEX, hydraulic fracturing input chemicals and produced water treatment chemicals (US EPA, 2016c). The quality of the wastewater generated also changes with time. It has been observed that the concentration of metals, TDS, organic and NORM increases with time (Barbot et al., 2013; Gregory et al., 2011; Haluszczak et al., 2013).  Thus, the constituents of the flowback water range in concentrations and type. Within the scope of this research, only inorganic constituents, salts and metals are studied.   18  Table 2-5 Comparision of range of concentration of constituents of flowback and produced water for Conventional Natural Gas, Coalbed Methane, Shale Gas and Tight Gas (Alley et al., 2011) Constituent (mg/L) Conventional NG CBM Gas Shale Gas Tight Gas Nitrate - 0.002-18.7 >2670 - Phosphate - 0.05-1.5 >5.3 - Sulfate 1-47 0.01-5590 >3663 12-48 Oil and Grease 2.3-60 - - >42 Ra226 (pCi/g) - - 0.65-1.031 - U - 0.002-0.012 - - HCO3 - - >4000 10-4040 Ammonium-N - 1.05-59 - >2.74 Al 0.4-83 0.5-5290 >5290 - As 0.002-11 0.0001-0.06 - >0.17 B >58 0.002-2.4 0.12-24 - Ba 0.091-17 0.01-190 >43700 - Br 0038-349 0.002-300 >10600 - Ca >51300 0.8-5870 0.65-83950 3-74185 Cd 0.02-1.21 0.0001-0.01 - >0.37 Cl 1400-190000 0.7-70100 48.9-212700 52-216000 Cr 0.002-0.231 0.001-0.053 - >0.265 Cu 0.02-5 >0.06 >15 >0.539 F  0.05-15.22 >33 - Fe >1100 0.002-220 >2838 >0.015 K 0.458-669.9 0.3-186 0.21-5490 5-2500 Li 0.038-64 0.0002-6.88 >611 - Mg 0.9-4300 0.2-1830 1.08-25340 2-8750 Mn 0.45-6.5 0.002-5.4 >96.5 >0.525 Na 520-120000 8..8-34100 10.04-204302 648-80000 Ni 0.002-0.303 0.0003-0.2 - >0.123 Sr 0.084-917 0.032-565 0.03-1310 - Zn 0.02-5 0.00002-0.59 >20 >0.076  2.3.3 Flowback water management and potential spill routes The flowback water, when recovered, is collected in storage systems like containment ponds, above ground walled storage systems or storage tanks before being transported via pipelines or heavy commercial vehicles. Depending upon the characteristic of the water, it might be treated for reuse for the hydraulic fracturing or drilling process or it can be recycled for crop irrigation,  or can be disposed of off-site using deep injection wells (Becklumb, 2015; Jiang et al., 2011; 19  Lutz et al., 2013; Rivard et al., 2014). Lester et al. (2015) have proposed a framework for assessing site-specific hydraulic fracturing wastewaters, techniques for their characterization and processes for guiding the selection of a custom treatment method.  The water thus treated can be reused for future fracturing operation or crop irrigation, livestock watering and indirect potable reuse. The use of flowback water outside the oil and gas industry will require a higher level of treatment (Lester et al., 2015). Closed loop systems to transfer produced water from well to the storage tanks via piping system might be used to decrease the number of heavy commercial vehicles transportation trips (GWPC and IOGCC, 2014). The water management activities for UOG extraction process is shown in Figure 2.3.  Figure 2-3 Water acquisition, use and wastewater management for UOG extraction process Failure of the pipes and connections during the transportation process or the failure of a flowback water storage tank can cause accidental spills or release. Spills and leaks of the flowback and produced water during transportation, loading, unloading, and storage possibly percolating through the pervious layers, contaminating the freshwater aquifers (Lutz et al., 2013; Vengosh et al., 2014). Majority of the flowback water release incidents have been reported during its storage. About 46% by number of spills and 75% by volume of spills were related to the failure of storage container integrity, according to the spill data for hydraulic fracturing flowback water from 2006-2011 (US EPA, 2015). Flowback water release volume ranging from 5000 to less than 1 m3 have been reported. An average spill volume of 288 m3 has been reported from 2008-2011 in Colorado, Texas, and Arkansas during storage (US EPA, 2016c). In addition, over 18 spills have been recorded in Montney in 2010 due to containment pond failure (D. Scheck, personal communication, November 29, 2016). 20  The types of storage container typically used for storing flowback water include above ground walled storage system(AWSS or C-rings), in-ground containment ponds and storage tanks. The former two are lined containment systems and are often lined using materials like high-density polyethylene (HSPE) liners or other synthetic material (BCOGC, 2015a). The liners prevent the stored waste or waste water from contaminating the soil and groundwater at or surrounding the site. The storage systems are often installed on site for a period of 2 months to 2 years (BCOGC, 2015a). Figure 2.4 shows an impounding containment pond with synthetic liners, drain and an inlet for flowback water loading. Figure 2.5 shows an above ground walled storage system which is a short cylindrical structure with open top and bottom with liner clamped on the structural wall, secondary containment system in form of a berm, inlet pipe, valve, and pump. These are the most commonly used storage systems in UOG industry currently (BCOGC, 2015b). Common modes of failure for this storage system are reported as liner failure, overflow, slope stability, use of inferior soil quality for construction of the berms, the presence of debris, erosion of the berm, field compaction and other (US EPA, 2016c). Ziemkiewicz et al. (2014) studied 15 containment ponds in West Virginia and listed slope instability and liner deficiencies as the most frequent failure modes. Olawoyin et al. (2013) statistically analyzed storage systems in Pennsylvania and ranked structural instability, insufficient capacity, and erosion as the most frequent types of failure. These analyses were focused on the containment ponds. In the literature reviewed, the analysis for the failure of AWSS does not seem to have been carried out.  Various studies have pointed out that the accidental surface releases of fracturing fluids and wastewater may affect surface water resources (Becklumb, 2015; Burton et al., 2014; Ewen et al., 2012; Gagnon et al., 2016; Rozell & Reaven, 2012; Torres et al., 2015; Vengosh et al., 2014; Vidic et al., 2013; Ziemkiewicz et al., 2014). Kuwayama et al. (2015) suggested that the water quality issues related to fracturing to be more severe than water quantity issues and have determined the more conclusive association between shale gas development and impacts on surface water quality.  21  Blewett et al. (2017) conducted toxicity studies to understand the sub-lethal and reproductive effects of flowback water on water fleas, a species that is likely to be found in environments subjected to flowback water spill. The organic fractions and salt components were identified as main mediators of toxicity. Chloride is required to maintain normal physiological functions, but when exposed to widely changing chloride concentration, the organisms are vulnerable to survival, growth and reproduction risks (Siegel, 2007). Essential metals like iron and manganese, typically found in flowback water, are required for the biological functions of an organism, however, superfluous concentrations of these metals can result in adverse effects if they overwhelm an organism’s homeostatic mechanisms (US EPA, 2007). Ziemkiewicz et al. (2014) evaluated the integrity of impoundments used to store fluids produced by hydraulic fracturing and offered recommendations to reduce environmental risk. However, the investigation does not characterize the effect of flowback water on the aquatic ecology. Chen et al. (2017) studied ecological impact of flowback water on soil ecosystem, conducted a preliminary human health risk assessment and pointed out the need for holistic environmental assessment of the implications of the flowback water release. Due to large volume generated, historically relatively higher tendency for spill, the high concentration of dissolved solids, and the complex physicochemical composition of the flowback water, there is potential for human health and environmental impact in an event of accidental release of flowback water into the environment (Gregory et al., 2011; Kargbo et al., 2010; Lester et al., 2015; Vengosh et al., 2014). From this, a need to conduct a holistic investigation of the adverse environmental effects on the aquatic ecology because of the possible failure of the storage system was realized. This assessment focuses on the assessment of the risk of an uncontrolled release of flowback water when stored near a surface fresh water body. 22    Figure 2-4 Containment pond and its components  (US EPA, 1988a)   Figure 2-5 Above ground walled storage system and its components 23 2.4 Risk Assessment Risk is most commonly defined as the possibility of adversely affecting something valued by humans from an event or an action (Pechan et al., 2011). Risk is related to the future and answers as to what can go wrong, the probability of that happening and the consequences arising out of it (Rausand, 2011d). The tool used for answering these questions is defined as risk analysis. When applied systematically, it can be used to assess the vulnerabilities of the system (Ostrom & Wilhelmsen, 2012). Risk analysis techniques can be qualitative or quantitative. When numerical estimates are provided for probabilities and consequences of the event, it is known as quantitative analysis (Rausand, 2011d).  Probability is defined as the likelihood that the event will occur (Ostrom & Wilhelmsen, 2012; Rausand, 2011e). A consequence is a specific damage to the asset(s). It can also be called adverse effects, impairment, impact or loss, while severity is the seriousness of the consequence (Rausand, 2011e). Risk can be defined as product of probability and severity of the consequence (Ericson II, 2005c)  𝑅𝑖𝑠𝑘 = 𝑃𝑟𝑜𝑏𝑎𝑏𝑖𝑙𝑖𝑡𝑦 ×  𝐶𝑜𝑛𝑠𝑒𝑞𝑢𝑒𝑛𝑐𝑒       (2.1) Risk analysis evaluates the probability of failure and severity of the consequence of that failure. The risk analysis results are used in the risk characterization stage. Risk characterization evaluates the risk and compares it with the risk acceptance criteria (Rausand, 2011d). It articulates major assumptions and uncertainty in the assessment and describes the results such that they can be used in the risk management decision-making process (US EPA, 1998). A complete risk assessment comprises of risk analysis and risk characterization. While proposing, implementing, control and assessing the effectiveness of the control measures would be the part of the overall risk management framework. The distinction and the components of each risk analysis, risk characterization, risk assessment and risk management for this study are shown in Figure 2.6. 24  Figure 2-6 Typical components of risk management Rausand (2011c) classified risk analyses based on three groups of hazards and their impacts on three groups of assets- the risk due to human activities or anthropogenic hazards; natural disasters; or hazardous material and their adverse effects on human health, environment and material or economy (Rausand, 2011d). Figure 2.6 is a generic framework which can be modified to assess the risk for each of the assets. The risk assessment carried out here is for assessing the impact of a flowback water storage system failure on the environment (aquatic ecology) due to uncontrolled release of flowback water caused by human, nature or toxicant. Some of the existing risk assessment studies for unconventional oil and gas activities are discussed here. The techniques specific to the assessing probability and adverse ecological effects are discussed.  2.4.1 Risk assessment studies for unconventional oil and gas development  Various risk assessment studies have been done in the past few years for analysing risks due to unconventional oil and gas development. Torres et al. (2016) reviewed risk assessment techniques that can be used for onshore UOG production to determine water quantity and quality risks related to hydraulic fracturing and flowback and produced water management. Techniques 25 and frameworks like environmental (ecological) risk assessment, barrier, and other operational risk analysis, hazard identification, layers of protection analysis and quantitative risk assessment were enlisted as often used in the oil and gas industry. Engineering techniques like probability bounds analysis and binomial distribution models have been used for assessing the likelihood of water contamination (Rozell & Reaven, 2012; Ziemkiewicz et al., 2014).  Ziemkiewicz et al. (2014) have evaluated the flowback water containment ponds by conducting field evaluations using event tree analysis and mapped likelihood of problem occurrence for each identified storage system integrity problem. Further, a probability analysis using the binomial distribution to identify construction and maintenance efforts to minimize accidental release of hydraulic fracturing fluids to the environment was carried out. The analysis specifically focused on the liner failure of the pond containment storage system of the data collected from 71 sites (Ziemkiewicz et al., 2014). Other storage systems like above ground walled storage system, widely used in British Columbia and Alberta for temporary storage, were not assessed (BCOGC, 2015b). Patterson et al. (2017) reviewed spills from 31,481 unconventional oil and gas wells in Colorado, New Mexico, North Dakota, and Pennsylvania. It was identified that 50% of spills were related to storage and moving fluids via flowlines and the spill volumes ranged from 0.5 m3 to 4.9 m3, the largest spills exceeding 100 m3. The need to report spills in order to prevent spills and mitigate potential environmental damage was recognized and an interactive spills data visualization tool was designed.  Ingraffea et al. (2014) collected data from 41,381 COG and UOG wells in Pennsylvania to assess well casing and cement impairment issues. Cox proportional hazards model was used to measure the risk of damage to the casing which can lead to methane migration into the atmosphere and/or into underground potable wells.  Some human health risk assessment studies have also been done for unconventional oil and gas development. Human health risks for exposures to air emissions were studied by McKenzie et al. (2012) using human health risk assessment framework. Chronic and sub chronic non-cancer hazard indices and cancer risks from exposure to hydrocarbons were estimated. The preliminary 26 results generated indicated the need for further studies. The studies carried out by Ferrar et al. (2013) affirmed this need. Interviews carried out with 53 participants attributed 59 unique health issues; stress was the most frequently-reported symptom. A need to address the identified health impacts and carry out exposure-based epidemiological studies were emphasized. A detailed exposure based human health risk assessment carried out by Bunch et al. (2014). Community-wide exposures to volatile organic compounds (VOCs) in the Barnett Shale region due to the unconventional oil and gas activities were studied to assess potential acute and chronic health effects. A probabilistic and deterministic human health risk assessment was carried out based on more than 4.6 million data points which demonstrated that the exposure levels of the VOCs were below the threshold of posed health concern.  A new study by Elliott et al. (2017) identified 20 known or suspected carcinogens and addressed the need for investigation into the relationship between unconvetional oil and gas development and cancer risk in general and childhood leukemia in particular. Similar study was carried out by Werner et al. (2015) to review the existing literature on to identify the evidence of environmental health impacts. The paper highlighted concerns related to air and water quality. The evidence gathered in the scientific research is unsure of the actual environmental health impacts, however, a clear gap is recognized in the scientific knowledge. Brittingham et al. (2014) reviewed the effects of unconvetional extraction activities on terrestrial and aquatic ecosystems in the United States. They identified the water quality and quantities as areas of concern for species like vernal pond inhabitants, and stream biota. Areas of futher research and monitoring were identified for developing effective policies to mitigate the adverse impacts on vulnerable species and ecosystems (Brittingham et al., 2014). Some of the above discussed studies on unconventional oil and gas were carried out to assess evidence or compute risk related to human health and environmental impacts. Few studies analyse the failure modes to identify and prioritize the mitigating actions. The techniques that have been used requires intensive region specific historic data, the collection of which is not always feasible. In addition, such a large-scale development of the unconventional oil and gas is not prevalent in most parts of the world, hence, the required experience and skills for such 27 analysis is straggling. Furthermore, the studies do not address the exposure in an event of accidental release while undertaking one or more oil and gas activities. The corresponding adverse effects on ecological or human health have also not been comprehensively studied. Overall, a need to undertake a holistic risk assessment to understand the impacts involved in all the stages of unconventional oil and gas production process and to prioritize them correctly was pointed out (Torres et al., 2015). While it is difficult to address all the stages, storage stage of the production process has been ranked one of the most critical in terms of failure. Two main objectives to be achieved through this assessment are to evaluate the probability of failure of the AWSS type of storage system and the potential ecological adverse effects of storage failure on the aquatic system. The scope of the assessment does not include proposing, controlling and assessing the risk reduction action. The tools and techniques used for assessing probability and aquatic ecological effect are discussed in the following sections.  2.4.2 Assessment of probability of failure To assess the probability of failure of an undesired event, it is required to identify and analyze the possible causes of the event. The cause of a system failure need to be identified and a logical sequence of the relations between the causes and the system failure needs to be established (Rausand, 2011c).  There are various techniques for frequency analysis, such as the cause and effect diagrams, fault tree analysis, a Bayesian network, Markov chains, and Petrinets. The Petrinets and Markov techniques are not suitable for identifying causes of the undesired events, while the cause and effect diagrams cannot provide quantitative estimates. Bayesian networks are flexible to use, but they are too complex and time consuming (Rausand, 2011b). Fault tree analysis (FTA) is the most commonly used technique for determining the root causes of an undesired event using a logical combination and graphical presentation of the various combinations of possible events occurring in a system (Ericson II, 2005b). FTA is a deductive analysis technique which transverses from the general problem to the specific causes known as basic events. It develops 28 logical fault paths from a single undesired event at the top to all the possible basic events at the bottom (Ericson II, 2005b; US NRC, 1981).  FTA has been often used to analyse failure modes for the storage system. Choi & Chang (2016) analyzed the reliability of the sea-bed storage systems tanks using fault tree analysis. Shahriar et al. (2012) studied pipe-line failure using fuzzy based bow-tie analysis and developed fault tree for pipeline failure. Wang et al. (2013) conducted a fuzzy fault tree analysis for crude oil tank explosion. US EPA (2013) proposed techniques for hazardous waste tank failure using FTA. It included underground and in-gound storage tank and closed and open storage tank analyses. However, bottomless lined tanks are not analysed, though the guidelines are supportive of mostly all types of tanks. An undesired event may be defined as an unwanted event, for example, an accident, a hazardous condition, or other such undesired failure modes. The graphical model can be interpreted as a mathematical model to calculate failure probabilities. FTA as an analytical technique uses logic theory, set theory, Boolean algebra, and reliability theory (CCPS, 2000b; Ericson II, 2005b). Table 2-6 Basic terms used in FTA and their definitions (CCPS, 2000b; Ostrom & Wilhelmsen, 2012; Rausand, 2011e; US NRC, 1981) Term Definition Failure The inability of a system, subsystem, or component to perform its required function Failure Modes The way the item or operation potentially fails to meet or deliver the intended function Probability Likelihood that the event will occur Undesirable/Top Event It is the complete or catastrophic failure and constitutes the top event in the fault tree. FTA focuses  on this particular event and the causes of this event are deduced logically Basic Event An initiating event requiring no further development Intermediate Event A failure event which can happen due to a combination of one or more basic events connected by logic gate  Minimal Cut Sets It is the smallest combination of basic events that will cause the top event to occur. A top event can have many minimal cut sets, and each minimal cut set may have a different number of basic or undeveloped events. Each event in the minimal cut set is necessary for the top event to occur, and all events in the minimal cut set are sufficient for the undesired event to take place.  Once the undesired events have been identified, it is required to prioritize them to assess the problem that can proactively be used for suggesting risk reduction actions that have the most 29 impact on the overall risk. The risk reduction actions can be based on either reactive or proactive approach. The actions taken to prevent or reduce the probability of an undesired event are proactive while the actions taken to reduce the severity of consequence are reactive (Rausand, 2011a). FTA requires minimal cut sets to rank and prioritize the identified failure modes. The number of cut sets can be large and can lead to complicated and tedious calculations. Also, the corrective actions thus prioritized does not take into consideration the severity of the effect of the failure mode or its detectability. The problem with high probability might have negligible severity on the overall system which affects the overall risk to the system.  The proactive actions use professional competences of the organization by a constructive approach to the problems and analysing the entire system to predict major adverse events and proactively implement changes to prevent them from occurring (Chiozza & Ponzetti, 2009).  Failure Mode Effect Analysis (FMEA) is a proactive error prevention system that is designed to identify problems in infrastructure and systems before adverse events occur (Duwe, et al., 2005). FMEA is a technique used to identify and fully understand potential failure modes, their causes, the severity of the effects and their detectability. It assesses the risk associated with the effects, causes, detectability of the failure modes and ranks them in decreasing order of their risk. It can also identify and carry out corrective actions to address the most serious concerns proactively (Carlson, 2012). The basic definitions of the term often used in FMEA are explained in Table 2.6. FMEA gives a systematic overview of failures in the system, thus, helping to assess the reliability of the system. FMEA provides less confidence that all the critical failure modes have been revealed, also human errors are often ignored. Accordingly, it is mostly used as a good basis for more comprehensive quantitative or qualitative analyses, such as fault tree analyses (Aven, 2015). Also in FMEA, all the failure modes are analysed and documented, including those having diminutive or insignificant consequences make it extremely demanding and extensive (Aven, 2015). Fault tree technique is often used in conjunction with FMEA as they both compliment each other. Such a system is known as Integrated Analysis. In Backward Integrated Analysis (BIA), FTA is considered as the main technique followed by FMEA as supplementary (Hong & Binbin, 30 2009). Ideally, FMEA is used for hazard identification and FTA is developed using the failure modes identified in FMEA. Such framework is known as Forward Integrated Analysis (FIA). However, some researchers have found the FIA approach to be labour intensive and difficult to apply as compared to BIA (Hong & Binbin, 2009). Thus, a BIA is proposed to identify failure modes logically using FTA and then prioritize failure modes for directing corrective actions to reduce the overall risk to the system. Table 2-7 Basic terms used in FMEA and their definitions (Carlson, 2012; Ericson II, 2005a)  Term Definition Occurrence (O) A ranking number associated with the likelihood that the failure mode and its associated cause will be present in the item being analyzed. The occurrence ranking considers the likelihood of occurrence during the design life of the product. Severity (S) A ranking number associated with the most serious effect for a given failure mode, based on the criteria from a severity scale. It is a relative ranking within the scope of the specific FMEA and is determined without regard to the likelihood of occurrence or detection. Detection (D) The detection ranking considers the likelihood of detection of the failure mode/cause, according to defined criteria. Risk Probability Number (RPN) RPN is a numerical ranking of the risk of each potential failure mode/cause, made up of the arithmetic product of the three elements: severity (S), occurrence (O), and detectability (D) of the cause Corrective Actions Actions required to prevent or control the cause  2.4.3 Assessing adverse environmental effects A typical environmental risk assessment framework would have four elements: hazard identification, toxicity assessment, exposure assessment, and risk characterization. The ecological risk assessment (ERA) framework proposed by US EPA (1998), provide detailed guidelines as to how to conduct each phase of the assessment. ERA evaluates the likelihood that adverse ecological effects might occur as a result of human activities and hence ecological risk assessment framework is used. It is conducted to offer information to the risk managers regarding the potential ecological effects of different management decisions (US EPA, 1998).  It is an iterative process which can continuously incorporate new information to improve environmental decisions. The assessment expresses deviations in ecological effects due to exposure to the contaminant. This capability may be useful to determine the degree of reduction to be applied to the contaminant to achieve required results. The ERA explicitly evaluates 31 uncertainty. They provide a basis for comparing, ranking, and prioritizing risks and can be used to analyse the cost-effectiveness of each decision (US EPA, 1998). ERAs are conducted to transform the scientific data into meaningful information regarding the anthropogenic risks to the ecology. The result of the ERA helps the risk assessor and managers to make informed decisions regarding the environment.  The ERA identifies problems in the first stage which is problem formulation. In the second stage, it assesses the toxicity profile for the contaminant to evaluate the effect or benchmark concentration for a range of species. The concentration of the contaminant in the water, sediment or soil media to which the organism is exposed to and at which it is expected to produce lethal or sub-lethal response and the concentration exceeds a particular limit but will have no effect below this limit, is termed as threshold effects concentration (US EPA, 2016a). It has been referred to as effect concentration in this document. Next, the amount of actual exposure concentration to the species is estimated. This phase describes the sources of contamination, exposure pathways, fate and transport of the contaminant, and estimated likelihood of exposure. Exposure is contact or co-occurrence between a stressor and a receptor (US EPA, 1998). Estimated exposure concentration (as the term used in the document) is the maximum concentration of the contaminant in the media (air, water, soil, sediment) that the species will be exposed to. The risk is then estimated by taking the ratio of the actual exposure concentration to the effect concentration. Risk calculation can be probabilistic or deterministic. A single point estimate of the ratio effect to exposure is taken for deterministic approach and for probabilistic approach, a range of possible environmental impacts and which ones are most likely to occur are derived so that they provide risk manager with a flexible tool for making decisions (US EPA, 2016h). Uncertainty analysis is carried out and the risk results are communicated for making informed decisions (US EPA, 1998). As enlisted by Torres et al. (2016) ecological risk assessment is one of the most widely used techniques in oil and gas industry. It has been used by Sadiq (2001b) to assess the effect of offshore drilling waste discharge on aquatic life. Fugacity and aquivalence based approaches were used to evaluate the contaminant fate and exposure concentration, toxicity profile for drilling waste discharge was created using empirical distribution functions and risk quotient was 32 derived by taking the ratio of exposure to effect. Nazir et al. (2008) conducted an ecological risk assessment of naphthalene and methane release in the marine environment. Xu et al. (2016) conducted an ecological risk assessment of the heavy metals in soils surrounding oil waste disposal areas by collecting soil samples. For conducting each phase of the ERA, various techniques are in place. The selection of techniques depends on factors like the characteristic of the contaminant, the representative species, assessment end, a measure of effect (lethal or sub lethal), exposure duration (acute and chronic) and other.  2.5 Regulatory Perspective The use and permit for the construction of flowback water storage systems are regulated under jurisdiction to avoid contamination events. Regulations1 are different in different provinces and they depend on the conditions in which storage systems are used, how they should be constructed, and their siting. For instance, regulations regarding the use of wastewater containment ponds in ecologically sensitive areas are in effect in almost all provinces. The role of the provincial and federal government, the regulatory structure, and current acts2, corresponding regulations and guidelines3 are studied.  2.5.1 National overview In Canada, the National Energy Board (NEB) has regulatory responsibilities for oil and gas exploration and production activities under the Canadian Oil and Gas Operations Act (COGOA). This includes the drilling, well completion, hydraulic fracturing and formation flow testing as well as production from onshore unconventional reservoirs. However, the NEB does not have jurisdiction over onshore hydraulic fracturing in any of the provinces. The onshore areas where this act is applied are the parts of the onshore that is under the administration of a federal minister, Nunavut and Sable Island (NEB, 2016; NEB, 2015, R.S.C., 1985, c. O-7). The purpose                                                           1 The regulation(s) under the Act provide the details to give effect to the policy. Not all Acts have regulations. Sometimes regulations are used to bring Acts into force. 2 Also called a statute. When a Bill (proposed law) passes third reading in the Legislative Assembly, and receives Royal Assent, it is thereby enacted and becomes an Act or law 3 the requirements and expectations 33 of the act is to promote safety, protection of the environment and the conservation of oil and gas resources (R.S.C., 1985, c. O-7). The Canadian oil and gas drilling and production regulations (COGDPR) under the COGOA, states (SOR/2009-315) that: “…the operator shall ensure that all chemical substances, including process fluids and diesel fuel, waste material, drilling fluid and drill cuttings generated at an installation, are handled in a way that does not create a hazard to safety or the environment…” In association with COGOA and COGDPR, NEB requires the use of Onshore Drilling Operations Involving Hydraulic Fracturing (Filing Requirements) for all cases where a proposed work or activity requiring an Operations Authorization (OA) involves hydraulic fracturing. These Filing Requirements focus on the unique elements of hydraulic fracturing (NEB, 2013). In addition, as a part of environmental protection, the NEB ensures that an environmental assessment (EA) is conducted for proposed activities in the Northwest Territories and Nunavut (NEB, 2013).  Furthermore, the federal government’s jurisdiction would be triggered on the provincial land if the proposed hydraulic fracturing activities were to occur within a wildlife area or migratory bird sanctuary (Atherton & Macintosh, 2014). The provincial hydraulic fracturing requirements and regulations are controlled by the appropriate provincial authority. They have a very significant decision-making role regarding the management, control, and exploitation of natural resources within the provincial jurisdiction (Atherton & Macintosh, 2014).   In Nova Scotia, the hydraulic fracturing activities are currently banned but the province had approved some operations in the past. New Brunswick regulates on-shore oil and gas activity through a series of Departments and statutes, with its Department of Energy and Mines being a central authority. New Brunswick recently developed a “Blueprint” stating the rules for industry to oversee all oil and gas activity in the province, including the extraction of shale gas through hydraulic fracturing (Atherton & Macintosh, 2014). In Manitoba, the Petroleum Branch under Manitoba Mineral Resources, develops, recommends, implements and administers policies and 34 legislation to provide for the sustainable development of Manitoba's oil and gas resources. The Branch deals with matters relating to well spacing, production allowable, pool designations, saltwater disposal, enhanced recovery projects and unitization (NRCAN, 2016d).  Hydraulic fracking for unconventional shale gas is currently not being used in Ontario. Shale gas or shale oil are not being extracted anywhere in Ontario (McKinley, 2015; NRCAN, 2016f).  In Quebec, oil and gas exploration activities require the obtaining of permits and authorizations issued by the Ministère du Développement durable, de l’Environnement et de la Lutte contre les changements climatiques (MDDELCC) and the Ministère de l’Énergie et des Ressources naturelles (MERN) (NRCAN, 2016g).  In Newfoundland and Labrador, Minister of Natural Resources has primary authority to regulate oil and gas activities under the Petroleum and Natural Gas Act and subordinate regulations. The Minister of Natural Resources appointed an independent panel to conduct a review of the socio-economic and environmental implications of hydraulic fracturing in western Newfoundland. The panel recommended that the hydraulic fracturing activity is paused in the western Newfoundland until the further examination is done (Gosine et al., 2016; NRCAN, 2016e).   Alberta and British Columbia are the most experienced with onshore oil and gas regulatory frameworks in Canada as they have a longer history of onshore oil and gas development. In Alberta, a single regulatory body, the Alberta Energy Regulator (AER), under the aegis of the Alberta Energy Ministry, is responsible for all aspects of oil and gas activities. In British Columbia, a single regulatory body, British Columbia Oil and Gas Commission (BCOGC), oversees all the oil and gas activities. (Atherton & Macintosh, 2014).  The regulations pertaining to a storage system for flowback water generated from hydraulic fracturing for the province of British Columbia are discussed in greater detail. The best practices voluntarily followed for the storage systems in U.S. are then discussed.  35 2.5.2 British Columbia overview The scope of the regulatory review includes acts, supporting regulations and guidelines for containment ponds, AWSS, secondary containment system and liner. Regulations pertaining to pipe, valves, and pumps are not considered in the scope. Table 2.7 describes the acts and their corresponding regulations applicable to the storage systems in British Columbia. Table 2.8 discusses the recommended guidelines or advisory by the Oil and Gas Commission.  Drilling and Production regulation, Environmental Protection and Management regulation, Oil and Gas Waste regulation, Hazardous Waste regulation, contaminated sites regulations and spill reporting regulations under Oil and Gas Activity Act and Environmental Management Act dictate directives about siting, design, construction, operation, decommissioning, dismantle and disposal of the AWSS, pits and liner (BCOGC, 2015b). These are general regulations, not specific to the unique needs of hydraulic fracturing. Management of Saline Fluids for Hydraulic Fracturing guidelines and information letter #OGC 09-07 dictate precise guidelines related to storage of hydraulic fracturing flowback water. Some of the points discussed in the regulations are omitted in the discussion of guidelines.  2.5.3 Voluntary best practices in the US A review was conducted by the Office of Resource Conservation and Recovery (ORCR) of the United States for the best voluntary management practices for oil and gas exploration and production waste. The review did not include provincial or federal regulations but was on such voluntary practices as they address pits, tanks, and land application/disposal (US EPA, 2014). Some of the best practices applicable to the flowback water storage guidelines are presented in Table 2.9. Best management guidelines recommend the use of AWSS storage system as compared to pits. Flowback water storage in pits is allowed in British Columbia, although the guidelines are rather strict, the regulations do not prevent the installation. The best practices suggest including the precipitation factor in the design of open tanks and considering storm run off for erosion safety of the storage containers. The closed loop systems have been given importance for reducing land footprint and environmental contamination. Alarms for overfill protection have been found to be beneficial in the event of the manual monitoring failure. The 36 restriction for storing wastewater based on its quality parameters like pH, salinity, hydrocarbon content, NORM is in place. The BC guidelines recommend the removal of liquids and gas before storing waste water, however, other quality parameters are not taken into account.37 Table 2-8 Acts and regulations pertaining to flowback water storage systems in BC (BCOGC, 2015b; Ernst and Young, 2015) Act Regulation Description Oil and Gas Activity Act [SBC 2008] (Chapter 36)  • Earthen Pits need to go through the permitting process under facilities approval • Prohibits spillage of harmful substances and requires the reporting, containment, elimination, and remediation in the event of a spill Drilling and Production Regulation (B.C. Reg. 282/2010)  • Should not contaminate any water supply well, usable aquifer, water body any land or public road, ice, or any water body merging in to any water body containing fish, aquatic plant or other aquatic biotas.  • Earthen pit used for storage must be designed by a professional engineer and should be installed under their supervision. The pit should be approved under Oil and Gas Activity Act facility permit.  • The pit should not be installed in the 100meters of the natural boundary of a water body unless it is a permitted location.  • The structure should not be less than 6600m3 capacity • It should be located and constructed so that the fluid will not cross the site boundary  • The structure must be able to bear the hydraulic pressure of the contained fluids at the designed capacity • The liner system must be installed such that the ground surface prep is satisfactory, the pond has a free board of 0.50m, is inspected and recorded daily for leaks, any sign of leakage is reported to the commission within 24 hours of discovery • The storage system must be uninstalled in after a year of first use. For the system to be used for a longer time, an impermeable secondary containment system having a capacity of 110% of the designed capacity of the primary system  Environmental Protection and Management Regulation (B.C. Reg. 200/2010 O.C. 435/2010) • Siting w.r.t water supply well or identified groundwater recharge area, watershed, or identified aquifer wildlife and wildlife habitat area, sensitive watershed, tree retention area, minimum riparian management and reserve distances • Maintain natural water flow in wetland • The waste materials or contaminants should not be dumped, accidently or otherwise, in the stream/lake Environmental Management Act [SBC 2003] (Chapter 53)  • Without the approval or permit, the waste should not be dumped into the environment unless it occurs under permit, approval, or in accordance with a regulation under EMA Oil and Gas Waste Regulations (B.C. Reg. 254/2005 O.C. 541/2005) • Any kind of unpleasant odor is prohibited outside the boundary of the site  Hazardous Wastes Regulation (B.C. Reg. 63/88 O.C. 268/88) • Make arrangements to allow for visual or any other form of manual inspection in an event of a leak  • Make available and maintain an impervious secondary containment system of sufficient capacity to provide overflow protection, must provide a water tight hose connection • The piping system, containment, and all  the equipment must be compatible with the type of the waste to be stored 38 Act Regulation Description • All waste transfer piping system must be equipped with automatic shutoff which can shut off the flow in an event of an accidental release if a secondary of containment is not provided • The waste water should be discharged into the environment, to the storm sewers or to a municipal or industrial treatment works, only if it meets the effluent safety criteria  Contaminated Sites Regulations (B.C. Reg. 375/96 O.C. 1480/96) • The site is contaminated only if the concentration of the contaminants exceed the background concentration of the contaminant at the site Spill Reporting Regulations (B.C. Reg. 263/90 O.C. 1223/90) • Provides who will report the spill, how it will be reported and reportable levels for certain substances Land Act [RSBC 1979] (Chapter 214)  • Needs to go through the permitting process for using the land for oil and gas activities • Siting. can be situated only with long term tenure Heritage Act [RSBC 1996] (Chapter 187)  • Permit process required for use of areas of land considered to be heritage property These materials contain information that has been derived from information originally made available by the Province of British Columbia at: http://www.bclaws.ca/ and this information is being used in accordance with the Queen's Printer License – British Columbia available at: http://www.bclaws.ca/standards/2014/QP-License_1.0.html. They have not, however, been produced in affiliation with, or with the endorsement of, the Province of British Columbia and THESE MATERIALS ARE NOT AN OFFICIAL VERSION.     39 Table 2-9 Guidelines for flowback water storage systems in BC (BCOGC, 2009, 2015b) Guideline Description Management of saline fluids, BCOGC, 2015 • Storage facilities must be further than 200m of supply well or groundwater capture zone or on top of aquifer or recharge zone, except with permission.  • The site investigation to ensure the geotechnical and global stability of the containment pond must be carried out or certified by a professional engineer • Synthetic liners should not be undamaged and must have a quality assurance/quality control report • It must be at least 30 mils thick and must have hydraulic conductivity of less than or equal to 10-7cm/s  • It must be designed for functional temperature • Proper ground surface preparation before installation.  • Use of geotextile cushion, if needed • Berm or dikes surrounding the entire site or AWSS must be constructed • An impermeable secondary containment system must surround the AWSS if the AWSS is being used for more than a year. It must have a capacity of 110% of the designed capacity of the largest AWSS,  • An AWSS must be situated at a site having a geologic unit of a minimum of 5m thickness and hydraulic conductivity of less than or equal to 10-6cm/s. Or it must have a barrier of compacted clay liner of thickness more than or equal to 30cm and hydraulic conductivity of less than or equal to 10-7cm/s; or any such liner of equal capabilities that act as a barrier between the ground and the contained fluid or waste  • The service life of an AWSS must be greater than the design life of the liner.  • The re-use of liners is no permitted and they are required to be recycled or disposed at permitted facility  • A survey of the soils below the AWSS should be conducted after dismantling to check for soil contamination • A site-specific response plan to be in place to ensure the protection of groundwater resources  • Containment pond will not be situated within a ravine, coulee, or gully, or within a 200-year floodplain or/and within 100m of normal high water mark of a natural water body  • It must have a minimum safety factor of 1.5 and must be certified by professional engineer • Earthwork construction must be carried out only under non-frozen conditions or measures should be taken during construction to meet the design criteria. • Primary and secondary synthetic liners for containment pond must be a minimum 1.5 mm thick, have a hydraulic conductivity of less than or equal to 10-7cm/s and have density, tensile strength appropriate to the site and must be chemical resistance, tear or puncture resistance. They must be separated by an engineered seepage system • A leak detection system must be installed within the engineered seepage pathway and a sub drainage system below secondary liner which allows for water sampling  • A minimum freeboard of 1.0 m must be always maintained in the pond. • The primary containment liner be regularly inspected and corrective actions be maintained and documented 40 Guideline Description  • A groundwater monitoring program must be developed and the samples must be collected from the leak detection system and sub-drain for analyzing its quality every week. A summary of the information collected must be submitted annually • If a spill occurs through the synthetic liner into the soil, action leakage rates, and flow though the holes must be calculated. The spill must be reported if the leakage occurs for 3 or more consecutive days, or the chloride levels are in excess of 250mg/L from the samples collected from the sub-drain • The water must be passed through the separator to  remove hydrocarbon liquids and gas Information Letter #OGC 09-07, BCOGC, 2009 • All types of liquid fracture fluid returns may be stored in closed top tanks. Only slick water fracture fluid returns may be stored in open top tanks or lined, earthen excavations. • Registration of all lined earthen excavations is required • The flowback water must not be stored in open or closed tanks for more than 90 days from the last day of completion unless otherwise approved by the OGC.  • The flowback water must be stored in  lined containment ponds, the liner must be used only until the design life is not exceeded • An impermeable synthetic liner must be provided certified by a professional engineer • If applicable, groundwater monitoring records must be maintained until reclamation has been completed • The lined containment ponds must be reported on decommissioning       41 Table 2-10 Voluntary best storage practices followed in the US  (US EPA, 2014) Document Section Description State Review of Oil and Natural Gas Environmental Regulations, Inc. - Guidelines for the Review of State Oil and Natural Gas Environmental Regulatory Programs, 2015 5.5.3 a The open tank system must consider the precipitation factor while accounting for the capacity of the tank 5.5.4 a Restrictions pertaining to the quality of waste stored for parameters like pH, salinity, liquid or gas hydrocarbon content, presence of NORM, or other content which can be harmful to the environment should be considered American Petroleum Institute (API) Guidance Document HF3 - Practices for Mitigating Surface Impacts Associated with Hydraulic Fracturing, January 2011 10.2  In order to avoid accidental spill, vital data must be maintained at the site regarding their site waste management and storage practices. The information must include capacity of each tank, capacity of secondary system, allowed accesses and restrictions and information about the liners  American Petroleum Institute (API) Recommended Practice 51R - Environmental Protection for Onshore Oil and Gas Production Operations and Leases, July 2009 6.1.7 Information about the drainage pattern of the site must be collected to avoid storage system failure due to runoff due to erosion of the base 8.1 The system and its operation must be set up so as to achieve minimum possible land footprint 8.3.1 j The facility must be situated away from the major transmission lines American Petroleum Institute (API) G00004 – Guidelines for Commercial Exploration and Production Waste Management Facilities, March 2001 4.2.3 The system must be protected from overflowing using a combination of one or more of the systems like backflow protection, automatic shut-off valves, visual check and/or loud audible alarms National Park Service (NPS) - Operators Handbook for Nonfederal Oil and Gas Development in Units of the National Park System, October 2006 Chapter 4, Pg 88 Use of closed loop system is desirable as a storage system to decrease the spill potential and the area of impact New Mexico Energy, Minerals, and Natural Resources Department - Pollution Prevention Best Management Practices for the New Mexico Oil and Gas Industry, 2000 Table 7.2 Use of closed-loop system with hydrotest pipelines is desirable  42 Document Section Description Railroad Commission of Texas - Surface Waste Management Manual, December 15, 2010 Chapter 4 The liner must be able to sustain impact during installation and transport. It should be tear and puncture resistant and must be able to withstand  thermal stress, weather conditions unique to the site, inflow, and outflow of the waste and must be compatible with it  High-temperature waste fluids discharged in the storage system over high pressure, accidental runover of the vehicles or equipment over the liner, burrowing animals must be considered for the compatibility of the liner U.S. Fish and Wildlife Service (Arkansas Field Office) - Best Management Practices for Fayetteville Shale Natural Gas Activities, April 2007 4.4 The capacity of the secondary containment system must be able to contain 1.5 times the total volume of all the storage tanks 4.19 Closed loop system is recommended near sensitive areas to decrease area of impact and avoiding environmental contamination 43 2.6 Summary In this chapter, the UOG resources have been defined and types of UOG have been described. The amount of UOG resources present in Canada at national and provincial level have been recognised. Hydraulic fracturing process and flowback water quantity, quality and management have been reviewed. Risks unique to hydraulic fracturing and the evidence of possible surface water contamination were summarized. Types of storage systems and the potential failure modes have been examined. The concepts of risk assessment have been discussed and the techniques to conduct a risk assessment of uncontrolled release of flowback water due to the failure of the storage container on aquatic ecology have been explored. Risk has been defined as a product of probability and consequence. The study proposes the ecological risk assessment for assessing the environmental effect and backward integrated analysis for identifying, evaluating probability and prioritizing failure modes of the storage system. Regulations, in effect, to ensure the integrity of the storage container and avoid environmental contamination have been studied for Canada at national and provincial level and have been examined in light of the voluntary best practices in the U.S. A detailed methodology, based on the frameworks discussed in this chapter, is proposed in the next chapter.  44 CHAPTER 3. RESEARCH METHODOLOGY 3.1 Overview of Methodology A methodology is developed to calculate risk to the aquatic ecology from accidental release of flowback water due to storage system failure as shown in Figure 3.1. It consists of two frameworks, Backward Integrated Analysis (BIA) and Ecological Risk Assessment (ERA). The probability of failure of the storage system is calculated using the BIA framework. The failure modes are identified and ranked to prioritize the areas of corrective action. The impact to the aquatic ecology is measured using ERA framework. The toxicity benchmark and exposure concentration of the flowback water constituents are computed. The probability number derived from BIA is input into the exposure component of ERA to incorporate the probability of accidental release for the contamination of aquatic ecology. The proposed methodology is developed for above-ground walled storage system (AWSS) type of storage system also known as C-rings as shown in Figure 2.5. It can also be applied to another type of storage system with minor modifications.     Figure 3-1 Proposed methodology for assessing risk to the aquatic ecology due to accidental release of flowback water 3.2 Backward Integrated Analysis (BIA) Framework Backward integrated analysis (BIA) is an integration of Fault Tree Analysis (FTA) and Failure Modes Effects Analysis (FMEA) techniques as shown in Figure 3.2. BIA uses FTA as the main 45 technique and FMEA as supplementary. The integration of these techniques is proposed to achieve a dual purpose of determining the probability of failure of the storage system using FTA and ranking the failure modes by means of FMEA. FTA and FMEA have widely used techniques in risk assessment as discussed in the previous chapter (Chiozza & Ponzetti, 2009; Duwe et al., 2005; Ericson, 2005; Lee, 1985; Mahmood et al., 2013; US NRC, 1981; Vesely, 2002).  In BIA framework, the first step is to examine the storage system, determine the scope of the analysis and identify the undesired top event. A logical fault tree is constructed in a top down manner to logically identify the basic and intermediate events, contributing factors for the system failure. Then the failure rates of the basic events are determined through historic data, literature data or expert advise. Based on the failure rates, top event failure probability is calculated using gate to gate analysis. This entire process is FTA. The details of each step are explained in Section 3.2.1. Once the basic events are identified, they can be evaluated through FMEA. Detectability and severity ranking are allocated to the basic events and risk probability number (RPN) is calculated by taking the product of occurrence (calculated from failure rate), severity and detectability. The basic events are ranked using the RPN and the areas to achieve the most effective risk reduction are assigned, however, corrective action evaluation for FMEA is not carried out. These steps partially define the FMEA technique which is further explained in Section 3.2.2. This framework is a modified version of the BIA as proposed by Hong & Binbin (2009). The BIA framework can be most effectively applied during the design phase of the system. The overall storage system is made of sub systems like the walled system, liner, piping, drainage system, secondary system. These sub-systems are made of numerous components and the failure for each of these components is to be analyzed. The number of failure modes increases rapidly for such a complex system with a large number of components, making it impracticable to analyze them item by item when done using the traditional FMEA technique (Hong & Binbin, 2009).  When the fault tree is constructed the entire system reduces to a number of logically derived basic events. With this information, it is reasonable to identify critical areas with great concern using FMEA for mitigation and risk management. For a complex system, the analysis process 46 can be simplified by logically determining basic events through FTA and considering them as critical areas to perform FMEA. BIA framework advocates efficiency by integrating FTA and FMEA (Hong & Binbin, 2009). FTA and FMEA techniques are discussed at length in the following sections.   Figure 3-2 Backward Intergrated Analysis (BIA) framework 3.2.1 Fault tree analysis (FTA) Fault tree analysis (FTA) is a hazard and root cause analysis technique where an undesired state of the system is analysed to find all the credible ways in which the undesired state can occur in the context of the operation and environment of the system (Ericson II, 2005b; US NRC, 1981). A fault tree is a graphic presentation of parallel and sequential combinations of failure events that will lead to the occurrence of the top or undesirable event. It portrays the logical interconnection of basic and intermediate events resulting in the undesired event (US NRC, 1981).   47   Figure 3-3 Steps required to perform fault tree analysis  (CCPS, 2000b; Ericson II, 2005b)  The main functions of FTA are to: 1. Determine the combinations of equipment failures, operating conditions, environmental conditions, human errors and other such root causes or hazards contributing to the top event 2. Identify the relationship between the root causes and determine their probability 3. Estimate the probability of occurrence of the top event There are six basic steps that are required to perform a complete and accurate FTA, as shown in Figure 3.3. FTA is the primary technique in the BIA framework and all the steps are carried out. The first step of FTA is to understand the system by studying the design drawings, schematics, and operation procedures. Then the scope and the top or undesired event of the fault tree is defined. A deductive logical fault tree model of the system is constructed (Ericson, 2005). The construction of the fault tree is an iterative process. The process begins at the top and moves down through the branches until all the events are defined in terms of the basic identifiable faults or human error (Ericson II, 2005b; US NRC, 1981). Often necessary, immediate and sufficient (N-I-S) concept is used to construct the fault tree, determining if the fault is necessary, immediate and sufficient to cause the top or intermediate events (Ericson II, 2005b; US NRC, 1981). In this manner, the fault tree proceeds down and continually approaches finer resolution until ultimately, the basic component failures or basic events are identified (CCPS, 2000b).  System DescriptionDefine top event and scopeConstruct fault treeQualitative examination of fault treeQuantitative evaluation of fault treeDocumenta--tion48 Once the fault tree is developed, it is inspected qualitatively. This can be done by inspection or using the rules of Boolean algebra. The combinations of the events can be expressed in terms of the Boolean equation. The evaluation can also be carried out using the minimal cut sets analysis (CCPS, 2000b). The minimal cut set analysis uses the rules of Boolean algebra to determine all the possible combinations of events in terms of basic and intermediate events that cause the top event to occur. These basic event combinations are called minimum cut sets. They represent the minimum set of events that are necessary and sufficient in order for the top event to occur. The minimal cut sets are more manageable for examining the fault tree qualitatively (CCPS, 2000a).  Once the fault tree is satisfactory in terms of logical correctness, completeness, effectiveness, the probability of the failure of the top event is computed and documented (CCPS, 2000b; Ericson II, 2005b; US NRC, 1981; Vesely, 2002). The symbols commonly used in the FTA are described in Table 3.1. The quantitative evaluation is done using the concepts of probability theory.  Table 3-1 Commonly used fault tree symbols (CCPS, 2000b; Ericson II, 2005b; US NRC, 1981) Symbol Name Description   Intermediate event It is a failure event that can happen due to one or more preceding event connected through the logic gates   Basic Event A basic event that does not need any additional development   OR Gate The resulting failure event will happen if at least one of the input events occur   AND Gate The resulting failure event will happen only if all the input events occur   Transfer In Symbolizes that the tree is further developed at the respective TRANSFER OUT   Transfer Out Symbolizes that the tree is to be connected to  the respective TRANSFER OUT  A basic assumption in FTA is that all failures in a system are binary in nature, i.e., a component or operator either performs successfully or fails completely (CCPS, 2000b). In order to calculate the probability of failure of a system from the failure rate of its sub-system and components, principles of reliability theory and probability distribution theory have been used. The Poisson distribution has numerous applications in describing the occurrence of the system failure under steady state condition (US NRC, 1981). Assuming that the system is in steady state does not 49 undergo any degradation, the failure is likely to occur equally over the time and the system either fails prior to time t or it does not, the reliability of a system will be defined by the probability of continuous successful operation for a time t (US NRC, 1981).  For the quantitative analysis, the probability model is selected based on the type of failure processes in the fault tree. For evaluating the probability of the storage system failure, longer exposure time period increases the probability of failure. As discussed above, the causes of failure can be due to human error, can be caused by operational, maintenance, and/or environmental factors. This model is most often used for quantification of the fault tree. The probability distributions are assumed exponential. The probability 𝑃 that the component fails in time period 𝑡, assuming that the component is initially working can be represented using equation 3.1 (US NRC, 1981):- 𝑃 = 1 − 𝑒−𝜆𝑇                                     (3.1) Where 𝑃 is the probability of failure, λ is the component failure rate per time 𝑡, and T is the component exposure time. In this case, the component is exposed over its operational life time, thus, 𝑡 = 𝑇. Equation 3.1 is an exponential distribution which has a constant failure rate. It arises out of Poisson distribution. . A ranking scale and corresponding probability of a failure is shown in Table 3.2. The probability of the failure of a top undesired event is calculated using “gate-to-gate” analysis. If, G is probability of event, and  𝑧1, 𝑧2, 𝑧3  are the probability of the events leading to G, the probability of AND gate would be: 𝐺 = 𝑧1 × 𝑧2 × 𝑧3                      (3.2) If the events are mutually exclusive or upper bound probability is considered, the equation for OR gate would be: 𝐺 = 𝑧1 + 𝑧2 + 𝑧3                      (3.3) 50 For example for Figure 3.4, the overflow of the storage system would occur if the control is lost over water level, AWSS is nearly full and the system fails to shut down. All the failure events occur at once. If the control is lost over water level but the AWSS is empty, then the overflow would not occur. All the three events have to occur at once for the system to overflow. So, the use of AND gate is appropriate. The lost control of water level occurs if either the inlet/outlet valve fails or water level was not monitored. Even if one of the two events occur, the control of water level is lost. Hence, the use of OR gate is appropriate. Also, “AWSS is nearly” full is a basic event, “overflow of the system” is an intermediate event and is transferring in, while “the system shut down failure” is transferring out. A scale is created to predict the occurrence of the basic events as shown in Table 3.2. This scale is similar to the traditional FMEA occurrence ranking. The failure rate values are relative. The corresponding probability of failure can be computed using equation 3.1 for a given ranking. A number from 1 to 10 can be allocated to each basic event. While the failure rate is not precise, for a new analysis like AWSS for which historic failure data is not available, data can be collected for the failure of basic events based on the given ranking. This occurrence rank is also used in the FMEA to calculate risk priority number (RPN). 3.2.2 Failure mode and effect analysis FMEA is a proactive fault prevention technique designed to identify problems in a system before they occur (Duwe et al., 2005). FMEA can be used as qualitative or semi-quantitative analysis. The main goals of the FMEA are to identify failure modes, assess causes and effects of failure modes, prioritize the failure modes by evaluating them and recognize actions to decrease the chance of the potential failures from occurring (Ben-daya, 2009). The FMEA has following major steps (Pillay & Wang, 2003):  i. Understand the system and identify and enlist the components of each sub-system and assembly ii. Determine the failure modes of the components   51  Figure 3-4 Example use of AND and OR gate  Table 3-2 Failure rate and correspoonding probability values and assigned ranking (Chang et al., 2001; Pillay & Wang, 2003) Ranking* Failure rate Probability 10 1/2 4E-01 9 1/10 1E-01 8 1/20 5E-02 7 1/100 1E-02 6 1/200 5E-03 5 1/1000 1E-03 4 1/2000 5E-04 3 1/10000 1E-04 2 1/20000 5E-05 1 1/100000 1E-06    52 iii. Analyse each failure mode to understand the severity of its effect on the system, its detectability, and its occurrence  iv. Estimate the occurrence of failure, severity of effect and detectability qualitatively (terms defined in Table 2.6) v. Calculate RPN, which is a ranking of the risk events and is a product of the three elements: severity (S), occurrence (O), and detectability (D) of the cause (Carlson, 2012; Ericson II, 2005a). vi. Rank the failure modes based on the RPN values. Develop recommendation and document the information in form of a table.  The purpose of ranking the failure modes in order of their criticality is to allocate the limited resources to the most critical failure modes and achieve the most effective results. A rank scale for occurrence, severity, and detectability of the FMEA is devised as shown in Table 3.3. This is the scales suggested by the traditional FMEA (Chang et al., 2001; Pillay & Wang, 2003). The table shows that the traditional FMEA uses the rank of 1-10 to measure the occurrence, severity, and detectability. This ranking is assigned to each failure mode. A risk priority number, a product of occurrence, severity, and detectability, is computed for each failure mode. The failure modes are then arranged in a decreasing order of their RPN to prioritize failure modes needing immediate attention. The FMEA and FTA have common occurrence scales. The FMEA is supplementary in the BIA framework. Step1 of the FMEA is common in both FTA and FMEA. Step 2, determination of failure modes is carried out using FTA. The failure modes in form of basic events are logically determined using the FTA. Steps 3 to 7 of FMEA are carried out in the BIA framework, however, the recommendations are not developed. The BIA framework gives two outputs: the probability of failure of the system and the ranking of the failure modes. The probability of failure derived is input into the ERA framework to calculate the final risk to the ecology. The failure modes ranked above are assessed in light of the current provincial regulations and guidelines reviewed in Chapter 2. In the next sections, the risk to the aquatic ecology would be assessed using the ERA framework. 53 Table 3-3 Severity, Occurrence and Detectability scale assigned for FMEA Rank Severity (S) Occurrence (O) Detectability (D) (%) 10 Complete system failure without warning 1/2 0-5 9 Complete system failure with warning 1/10 6-15 8 Serious damage 1/20 16-25 7 Major damage 1/100 26-35 6 Significant damage 1/200 36-45 5 Moderate damage 1/1000 46-55 4 Performance deterioration 1/2000 56-65 3 Slight deterioration 1/10000 66-75 2 Very slight deterioration 1/20000 76-85 1 No effect 1/100000 86-100  3.3 Ecological Risk Assessment (ERA) Framework The information about the failure of the storage system is available. The subsequent contamination of the surface water body is to be analysed. The aquatic ecological risk due to the accidental release of flowback water is estimated using the ecological risk assessment (ERA) framework. US EPA (1998) has defined ecological risk assessment as  “a process that evaluates the likelihood that adverse ecological effects occur or are occurring as a result of exposure to more or more stressors”. The ERA process is based on the characterization of ecological effects and characterization of the exposure of flowback water constituents to the aquatic ecology. The characterization of ecological effects evaluates the ability of a pollutant or contaminant, also known as a stressor(s), to cause adverse effects under a particular set of circumstances. While the characterization of exposure evaluates the interaction of the stressor with one or more ecological entities. These provide the focus for conducting the phases of risk assessment: problem formulation, analysis, and risk characterization (US EPA, 1998). The guidelines were proposed by the US EPA (1998) for conducting ERA. The ERA framework provides guidelines to avail consistency in carrying the risk assessment, however, they do not discuss the modeling tools to conduct the assessment (Sadiq, 2001). Figure 3.5 portrays the ecological risk assessment (ERA) framework for flowback 54 water and the proposed techniques to conduct each of the three phases (US EPA, 1998). These phases are developed step by step in the following sections. 3.3.1 Problem formulation Problem formulation is the first phase of ERA. This phase evaluates the primary hypothesis as to what kind of ecological effects might occur and refines the objectives of assessment. It consists of the integration of available data to produce assessment endpoints, conceptual model and analysis plan (US EPA, 1998). An accidental release of flowback water causes a release of contaminants into surface water through runoff, solute migration through groundwater or other pathways. The possible primary effect is related to the contaminants in flowback including increased mortality and behavioral, biochemical, growth, physiology, population and reproduction related adverse effect on the aquatic life (US EPA, 2016f). The problem identified in this assessment is to measure short-term acute adverse effects on the species found in an aquatic water body because of an accidental spill or release of flowback water into a surface water body. The three sub-steps of problem formulation process are discussed below. 3.3.1.1 Assessment endpoints The basic criteria to select appropriate assessment endpoints are ecological relevance, susceptibility to stressors, and relevance to management goals. Ecologically relevant endpoints reflect important characteristics of the system and are functionally related with other endpoints (US EPA, 2003). Aquatic organisms are vulnerable to salts, metals and organic compounds when exposed to concentrations above the benchmark levels. Selecting food web as assessment endpoint seems more accurate than choosing a single or a group of species (Nazir et al., 2008; Sadiq, 2001). As it is difficult to obtain toxicity data for all organisms of the aquatic ecosystem, the organisms representing taxonomic groups are used as surrogates (Nazir et al., 2008; Sadiq, 2001). In this assessment, the endpoints are chosen depending on the availability of the acute toxicity data of the aquatic organisms in the literature. 55 3.3.1.2 Conceptual model A conceptual model was developed to predict the relationships between ecological entities and the stressors to which they may be exposed, possible exposure pathways, resulting in ecological effects on the endpoints. The conceptual model can be complex, depending on the number of stressors, assessment endpoints, characteristics of effects and the receiving ecosystem. Conceptual models is an influential tool for risk assessors for a future application that also helps for adjustment as knowledge about the parameter improves (Suter II, 1996, 2007; US EPA, 1998). The conceptual model is shown in Figure 3.6. It shows the possible contamination pathways including soil, surface water, groundwater and potable aquifer contamination by mechanisms like leaching, run off release and groundwater transport. The contamination leads to lethal and sub-lethal effects on the aquatic and terrestrial biota, which are in-turn ingested by human causing risk to human health. The scope of the assessment is limited to the contamination due to run off. Flowback water is a mixture of organic compounds, salts, and metals. As defined in the scope of the assessment, only the inorganic constituents of the flowback water, metals and salts are considered. Toxicity of salts are often expressed as a combination of ions as their toxic effects might be more influenced by ion imbalance than the absolute concentration of the ions (Bright & Addison, 2002; Mount et al., 1997; Mount et al., 2016). However, this can lead to multiple single ion interactions among the ions in flowback, ions present in a surface water body as well as between the ions of flowback water and surface water. In addition, the ion concentrations in flowback water, surface water background concentration values, and available benchmark toxicity guideline values are measured for individual ions. Thus, it is practicable to consider single ion toxicity for the contaminants of flowback water.  The Canadian Water Quality Guidelines for Protection of Aquatic Life (CWQG-PAL) has recommended deriving acute toxicity values for short-term high intensity event like a spill (CCME, 2012). To access the lethal and non-lethal toxic effect of the inorganic ionic  56  Figure 3-5 Ecological Risk Assessment framework for flowback water,  adapted from (US EPA, 1998) 57  Figure 3-6 Conceptual model for flowback water release from storage and contamination pathways 4 contaminant on the species, LC50 (Lethal Concentration 50%) and EC50 (Effect Concentration 50%) values are considered. LC50 is the concentration of the contaminant at which 50% of the population is killed. EC50 is the concentration of the contaminant at which 50% of the population is adversely effected in terms of reproduction, mobility, and other factors. The LC50 and EC50 values for each contaminant of the flowback water are collected from the literature. The endpoint acute toxicity values are collected from literature for an exposure period of 96 hours or less. The geometric mean is taken if more than one LC50 or EC50 values are available for the same exposure period for the same organism. 3.3.1.3 Analysis plan This is the last stage of problem formulation. The stressor (flowback water) exposure to an aquatic organism could be through contaminated water and/or food. Physicochemical properties of the flowback water constituents and physical parameters of the surface water body are                                                           4 Contamination due to volatilization during accidental release not considered for open top storage system 58 required to be examined to conduct exposure assessment. The acute toxicity values of an organism for the stressor are derived in the effect analysis. 3.3.2 Analysis phase The two primary components of the analysis phase are exposure and effects analysis (US EPA, 1998). The analysis phase is an intermediate phase between problem formulation phase and risk characterization and links them. The analysis plan developed using assessment endpoint and conceptual model in the problem formulation phase act as the basis for the risk analysis. In the following subsections, the techniques to conduct exposure and effects assessment are discussed. The risk is assessed in the form of a risk quotient which is the ratio of exposure concentration to effect concentration. Uncertainty and sensitivity analyses are then performed to quantify the errors and improve the credibility of the analysis (Sadiq, 2001; Suter II, 2007; US EPA, 1998).  3.3.2.1 Analysis of effects An aquatic life is the assessment endpoint based on the availability of data. To assess short-term (acute) response, LC50 or EC50 values are used. Figure 3.7 represents the guidelines of CWQG-PAL to be used corresponding to the type of data available for deriving acute toxicity of the contaminant. Primary data refers to the data that are based on scientifically defensible toxicity tests whereas secondary data originate from literature studies that are of acceptable quality and documentation (CCME, 2007). In this analysis, only secondary type of data is considered. For Type A analysis, species sensitivity distribution (SSD) curves are generated. SSDs model the variation of the sensitivity of species to the defined stressor.  They are generated by fitting a statistical or empirical distribution function (EDF) to the proportion of species affected with respect to the stressor concentration (CCME, 2011; US EPA, 2016e). SSDs are created using secondary LC50 and EC50 data and they help understand the relative sensitivities of the species.  The empirical distribution of a random sample is the uniform discrete measure on the observation. For random samples, 𝑦1, 𝑦2, … , 𝑦𝑛, an EDF can be defined as (Van der Vaart, 2000): 59 𝐹𝑛(𝑡) =  1𝑛∑ 1(𝑦𝑖 ≤ 𝑡)𝑛𝑖=1              (3.5) The distribution curves were fitted by determining the goodness of fit of the model using Akaike Information Criterion (AIC). The AIC statistics are calculated from the log-likelihood function by the simple expressions and are generally recommended (Palisade Corporation, 2015b). A standard safety level was set at the 5th percentile on the fitted SSD curve to save 95% of the ecological community. This measure is defined as the contaminant effect concentration (Suter ll, 2006, Shaw-Allen & Suter II, 2016). Tools like SSD generator and @Risk can be used to develop the SSD curves (Palisade Corporation, 2010; US EPA, 2016e). For deriving toxicity value for using guideline Type B2, the LC50 value of the most sensitive species from the available dataset is divided by a safety factor of 10 (CCME, 2011). The derived effect concentrations are then compared with background concentrations of the contaminants naturally present in the water body. If the naturally occurring levels of the given contaminant are higher than the derived effect concentration value, then it is assumed that the aquatic life at this location has adapted to this condition. The natural background concentration is then taken as the site specific value (CCME, 2011).  3.3.2.2 Analysis of exposure Exposure assessment is used to understand the fate, transport through various exposure pathways and the bioavailable fraction (BAF) of the contaminant. It represents the final concentration available to the receiving ecology (CCME, 1997).  The surface water contamination can occur through groundwater, land contamination or runoff as shown in Figure 3.6. The exposure pathway considered for this assessment is surface water contamination through runoff. For exposure analysis of surface water body, fugacity and aquivalence based models are widely used. Fugacity models are applicable to contaminants which maintain vapour pressure, which is mostly valid for organic chemicals. Inorganics and metallic compounds do not maintain vapour pressure, hence a modified form of fugacity model, aquivalence is used for evaluating fate and transport of heavy metals (Diamond et al., 1994; Sadiq, 2001b).  60  Figure 3-7 Minimum secondary data requirement for short-term exposure in freshwater and applicable guidelines (Adapted from (CCME, 2012))61 Various exposure assessment models for organic contaminants are available. For surface water exposure assessment, the US EPA suggests the use of models like the AQUATOX, the GIS-based BASINS model, and the EXAMS for pesticide contamination, and the Visual PLUMES to evaluate dilution and dispersion behaviour of contaminants in surface water bodies (Suter II, 2007; US EPA, 2016g).  These models have very specific application purposes requiring explicit inputs. For instance, the AQUATOX evaluates the toxic effects of biologically available chemicals through processes such as sorption, hydrolysis, volatilization, and photolysis. It requires inputs of nutrient, sediment, and toxic chemical loadings, general characteristics of site, growth characteristics and sensitivity of the population to the contaminants (Suter II, 2007).  Another such tool is EFAST or exposure and fate assessing screening tool. It calculates human potential dose rates for a variety of exposure routes and estimates the number of days per year that an aquatic eco-toxicological benchmark concentration will be exceeded for organisms in the water (Tobias & Kwon, 2016). It requires inputs like the amount of chemical released; media of release; days per year of release; chemical properties; and, detailed release location data; if applicable (Tobias & Kwon, 2016). In addition, these models are designed for organic and pesticide fate and transport. For assessing the fate and transport of the metals, PHREEQC, MINTEQA2, Visual MINTEQ, MINEQL, WHAM models are available (Gustafsson, 2014; HydroGeoLogic Inc & Allison Geoscience Consultants Inc., 1998; US EPA, 2007). They consider the physicochemical properties unique to the inorganics like speciation, complexation, adsorption, indefinite persistence, bioavailability and bioaccumulation (US EPA, 2007). However, each of these models includes only some of the metal-specific capabilities and no single model is currently available for use, that includes all the features (US EPA, 2007). For instance, Visual MINTEQ calculates speciation, redox reactions, adsorption, and precipitation but it is difficult to calculate transport rate of the metal from contaminant source to water body (Gustafsson, 2014). Furthermore, due to the complexity and multiplicity of the processes and the interaction of multiple ions, recourse is made to the use of a single partition coefficient for this assessment. The exposure route to be assessed is a runoff of the flowback water from land into the surface water body from the source of the spill. The excess water that does not infiltrate is a runoff 62 (Findeling et al., 2003). The contaminants in the flowback water have different physicochemical properties, affecting their transport and fate in the environment. Contaminants having a higher tendency for soil adsorption, have higher adsorption coefficient value. They can potentially bind to soil and can become comparatively steady. The rest of the contaminants might travel with the transporting media, and get introduced into surface waters and ultimately to the aquatic organisms through ingestion, inhalation, direct consumption and biomagnification (US EPA, 1988). Metal ions are persistent in the environment and the rate of their movement through soils and their concentrations in groundwater are governed by the advection, dispersion, matrix diffusion, and retardation processes (US EPA, 2007). Advection and dispersion are functions of the transporting media. Matrix diffusion, which is a function of the contaminant, is relatively unimportant and is omitted in most transport models (US EPA, 2007). Retardation depends on a number of factors and may involve sorption, precipitation, colloid formation, and bio-fixation. Salt ions are highly soluble in water, and they are resistant to biodegradation, volatilization, and photolysis due to the inorganic nature (Bright & Addison, 2002). The environmental transport of these ions is being assessed for adsorption (US EPA, 2007). The assumption underlying the exposure assessment is that during run off, the flowback water constituents are adsorbed on to the surface of a solid and are in chemical equilibrium. Adsorption can be defined as the net accumulation of matter at the interface between a solid phase and an aqueous-solution phase (US EPA, 1999). The distribution coefficient or the adsorption coefficient is a measure of adsorption and can be defined as the ratio of the amount of the substance adsorbed (known as adsorbate) on the solid to the amount of adsorbate remaining in the solution at equilibrium (US EPA, 1999).  𝐾𝑑 =𝐴𝑖𝐶𝑖                                     (3.6) Where 𝐾𝑑  is the adsorption coefficient, 𝐴𝑖 is the amount of adsorbate adsorbed per unit mass of solid, and 𝐶𝑖 is the concentration of the chemical remaining in the solution per unit volume of liquid phase at equilibrium. Equation 3.6 is valid under the assumption that concentration of free 63 or unoccupied surface adsorption site on a solid phase is in great excess to 𝐶𝑖. The limitation of this model is that 𝐾𝑑 is constant for homogenous conditions and a new value of 𝐾𝑑  would be required if there is change in groundwater and soil properties.  There are various approaches available to measure 𝐾𝑑 values including laboratory or in-situ batch method, flow-through (or column) method, field modeling method, and 𝐾𝑜𝑐 method  (US EPA, 1999). Flow through method uses mean residence time and are helpful for non equilibrium conditions, field modeling is highly field specific and 𝐾𝑜𝑐  method is used for determining organic constituent partition. For further analysis the 𝐾𝑑 values are taken from literature and simple Batch method equations are used to establish 𝐴𝑖 (US EPA, 1999)- 𝐴𝑖 =𝑉𝑠 (𝐶𝑜−𝐶𝑖)𝑀𝑠𝑜𝑖𝑙          (3.7) Where, 𝑉𝑠 is the volume of spilled liquid, 𝐶𝑜 is the concentration of contaminant in the spilled liquid, and 𝑀𝑠𝑜𝑖𝑙 is the unit mass of soil contaminated by the spill. For the analysis, we need the amount of concentration remaining in the solution after passing through the land. Thus, the value of 𝐶𝑖 for given 𝐾𝑑 can be calculated using: 𝐶𝑖 =  𝑉𝑠𝐶𝑜𝐾𝑑𝑀𝑠𝑜𝑖𝑙+𝑉𝑠          (3.8) The exposure concentration of the contaminant in the surface water body is evaluated based on the volume of contaminant introduced in the stream and their dilution coefficients. Contaminants may be restricted and might remain concentrated, or might disperse and become diluted in the stream to insignificant concentrations (US EPA, 1988). A simple mass balance equation is proposed to calculate the contaminant concentration in the natural water body after the flowback water is discharged into it. The total contaminant concentration (𝐶∞) at equilibrium in the creek is obtained using the following equation (CCME, 1997; Leeuwen & Hermens, 2004; US EPA, 1988)- 𝐶∞ =  𝐶𝑤𝑄𝑤+𝐶𝑖𝑄𝑠𝑄𝑤+𝑄𝑠          (3.9) 64 Where, 𝑄𝑤 and 𝑄𝑠 are the discharge of the creek and spilled flowback water, respectively, and 𝐶𝑤  is the concentration of the contaminant in the surface water body. The dilution factor theory assumes that the blending of the contaminants in the creek water is uniform and complete. It also assumes that decay and removal processes like sorption, volatilization, leaching are negligible and the flow rates of the water and the contaminant remain constant (US EPA, 1988). If the spilled chemicals do not have a short biodegradation half-lives and high sorption potentials, the dilution factor model can be effectively used to predict the concentrations of the contaminants in the natural water body within few kilometers of the spill site as processes other than dilution would have relatively smaller effect (Leeuwen & Hermens, 2004). Moreover, the decrease in concentration due to adsorption are incorporated into the total contaminant concentration. Hence, the integration of these techniques should give satisfactory results. In order to take into consideration the movement of aquatic organisms about the contaminated water as well as the bioavailability of the contaminants, the estimated exposure concentration (𝐸𝐸𝐶), would be estimated using equation 3.10 (Nazir et al., 2008): 𝐸𝐸𝐶 = 𝑝 × 𝐶∞ × 𝐵𝐴𝐹         (3.10) Where 𝑝 is the probability of exposure. It is the ratio of the impact area to the total area under consideration. If the area considered for assessment is equal to the area under impact area, 𝑝  can be equal to the probability of contamination. The value of 𝑝 is derived from the BIA framework using FTA. 𝐵𝐴𝐹 is the bioavailable fraction which is approximately 1 for chemicals having a log of octanol water partition coefficient less than 5, that is, log 𝐾𝑜𝑤< 5 (US EPA, 1988, 2016f). The entire exposure assessment process is explained in Figure 3.8. The flowback water with initial concentration of its constituents, stored in the AWSS gets released accidently. Runoff of the flowback water occurs and some of the constituents get adsorbed by the soil. Remaining concentration enters the stream and dilutes, greatly reducing the concentration of the contaminants. After taking into consideration background concentration and 𝐵𝐴𝐹, the final concentration, 𝐸𝐸𝐶 is available to the aquatic organisms.  65 3.3.3 Risk characterization As the final phase of the risk analysis, risk characterization combines the results of exposure and effects analysis and the associated uncertainties. The risk quotient is used for risk characterization. A sum of the ratio of exposure and effects concentration for each contaminant is obtained from risk quotient (Sadiq, 2001; US EPA, 1998)-  𝑅𝑖𝑠𝑘 𝑄𝑢𝑜𝑡𝑖𝑒𝑛𝑡 = ∑(𝐸𝐸𝐶)𝑛(𝐸𝐶)𝑛𝑛         (3.11) Where n is the contaminant and 𝐸𝐶 is effect concentration for the contaminant to be a concern. If the risk quotient is greater than one, there is a possible adverse effect to the ecology. The quotient integrates risk due to multiple stressors. This approach assumes that the toxicities of the chemicals are additive. The synergistic or antagonistic effects between different chemicals contained in the flowback water are not considered (Suter II, 2007; US EPA, 1998). The full equation for calculating risk quotient (R) is shown as equation (8): 𝑅 =  ∑𝑝×𝐵𝐴𝐹𝐸𝐶𝑛(𝐶𝑤𝑛×𝑄𝑤𝑄𝑤+𝑄𝑠+𝑉𝑠×𝐶𝑜𝑛×𝑄𝑠((𝐾𝑑𝑛×𝐴𝑠𝑜𝑖𝑙×𝑑×𝜌)+𝑉𝑠)×(𝑄𝑤+𝑄𝑠))𝑛      (3.12) Where 𝑉𝑠 is the volume of spilled flowback water (L), 𝐸𝐶𝑛 is effects concentration of the nth contaminant (mg/L), 𝐴𝑠𝑜𝑖𝑙 is the area of the contaminated soil (m), d is the distance of the spill site from the water body (m), 𝐶𝑜𝑛 is the initial concentration of the nth contaminant in the flowback water (mg/L), 𝐶𝑤𝑛 is the background concentration of the nth contaminant in the creek (mg/L), 𝜌 is the soil bulk density (g/m3) and 𝐾𝑑𝑛 is the adsorption coefficient of the nth contaminant. A critical value of R as 1, would mean that the exposure concentration has reached the effect concentration value and 95% of the ecology is at risk to the lethal and sub-lethal effects caused by the increase in exposure concentration. 66     Figure 3-8 Exposure assessment model proposed for inorganic constituents of flowback water 67 3.3.4 Uncertainty, sensitivity and scenario analyses The risk quotient generates single point estimate of the risk. The single values used for the calculation of risk might be conservative to decrease the possibility of underestimating the true risk to the organism. However, in this approach, the possibility of overstating the risk might be high. Uncertainty analysis is used in the ERA framework to measure the confidence in the risk estimate. It can further help to fully characterize risk to wholly assess the consequences and limitations of the risk assessment (Hammonds et al., 1994).  Uncertainty analysis can be a valuable tool for ranking the parameters contributing to the uncertainty in risk estimate of the contaminants. The rankings give guidance regarding the acquisition of additional data pertaining to the parameter to reduce uncertainty in risk estimates (US EPA, 1998). In order to assess the full range of possible values, the usual approach is developing probability distributions for the uncertain parameters included in the computation of risk quotient. Conservative estimates of 𝐸𝐶 and 𝐸𝐸𝐶 are considered for risk assessment. Probability distributions are generated for a range of 𝐸𝐶 values for each contaminant derived from SSD generator and @Risk. The distributions for initial concentration, 𝐶𝑜  are produced for a range of concentration values of selected contaminants in flowback water collected from literature. The true values of 𝐸𝐶 and 𝐶𝑜 exists but are unknown. Such uncertainty is known as epistemic uncertainty (Regan et al., 2003). The distributions are fitted using the AIC test as discussed for SSD curve generation (Palisade Corporation, 2015a). For the assessment of uncertainty, Monte Carlo analysis is proposed. It is a computer-based simulation technique that is used to obtain a probabilistic approximation to the solution of a mathematical equation or model (US EPA, 1997). The analysis uses statistical sampling techniques to sample a random value from each distribution of the uncertain parameter and the process is repeated for the desired number of samples or iterations (Hammonds et al., 1994). The analysis uses the distributions generated for the uncertain parameters and the relationship established between the risk quotient and these parameters (as shown in Equation 3.12) act as inputs for uncertainty assessment of the risk estimate.  68 The analysis is performed using computerized mathematical Monte Carlo simulation programs to achieve a forecast of risk quotient in form of distributions which allow assignment of confidence intervals for the risk estimate (Hammonds et al., 1994; Palisade Corporation, 2015b; Regan et al., 2003). The confidence interval can be adjusted based on the required level of confidence in the risk estimate. A number of software packages can be used to conduct Monte Carlo analysis including @Risk, Crystal Ball and SimLab (Oracle, 2016; Palisade Corporation, 2015b; Refsgaard et al., 2007).  Sensitivity refers to the variation in the output parameter of the mathematical model or equation with respect to changes in the value of the input parameters. Sensitivity analysis provides an insight as to which input variables contribute most to the overall uncertainty to the output parameter (Hammonds et al., 1994; US EPA, 1997). The 𝐸𝐶 and 𝐶𝑜 of which contaminant are contributing the most towards the uncertainty in the risk estimation is examined. A reduction in the level of uncertainty of the most sensitive parameters would contribute to reduce an overall uncertainty of the output parameter. The sensitivity analysis be can be performed using Monte Carlo analysis. From the result of simulations, analytical techniques like change in output statistic, regression analysis, and rank correlation analysis are performed. The results are displayed as a tornado graph, with longer bars at the top representing the most significant input variables contributing to the uncertainty in the output (Palisade Corporation, 2015a). Scenario analysis is conducted to logically and internally explore alternative futures, as such, it is a tool to describe the future under different assumptions (Refsgaard et al., 2007). It is a powerful tool for asking “what if” questions to explore the consequences of uncertainty (Duinker & Greig, 2007). Scenario analysis is carried out for different flowback water spill volumes and the receiving aquatic body discharge values using the analytical technique of regression analysis. Regression analysis technique performs analysis using the least squares to fit a line through a set of observations. A change in single dependent variable affected by the values of one or more independent variables can be analyzed (Refsgaard et al., 2007). A relationship between a range of volume of the spill, seasonal discharge of aquatic body and the risk quotient is derived. The result obtained can help to derive the maximum volume of the spill for which the exposed ecology of the river, stream or creek can be considered safe.  69 3.3.5 Risk description The ERA is an iterative process. The tools or techniques used in each ERA phase might need assumptions due to unavailable information or missing data. As data is updated with newly available information, the risk should be assessed and quantified in light of the new data. The results generated in the risk assessment process are evaluated and interpreted. It assesses the assumptions of the risk assessment and sees if further evidence is required. The significance of the adverse effects as stated from the risk results are interpreted. The interpretation provides valuable information for decision-making. An acceptable risk, as described by US EPA, is one in a million. For risk quotient, it is binary, if greater than 1 there is a greater chance of risk and less than 1 the ecosystem is relatively safer. The risk results are evaluated as per the risk precepted by the stakeholders. Risk perception refers to an understanding of the risk among the general public and its acceptability as it is not possible to completely eliminate the risk.  The method proposes BIA and ERA framework to quantify the risk to the ecology due to the accidental release of flowback water. BIA is used to calculate the probability of the failure of the storage system and rank the failure mode. ERA helped to assess the exposure and effect of the flowback water contaminants on the aquatic ecology. The probability is input in the ERA framework to estimate the risk. Uncertainty, sensitivity and scenario analysis is carried to obtain the full range of risk estimate, to determine the parameters contributing to the uncertainty and derive the relationship between spill volume and receiving aquatic body discharge. This method is applied to a creek in Montney unconventional play in Northern BC, Canada to assess its effectiveness. The probability of failure of the AWSS system was assessed and the modes of failure studied. The study area, data collection, results of the assessment and conclusion are discussed in the following chapters.     70 CHAPTER 4. CASE STUDY 4.1 Study Area Montney unconventional play trend is the single most important unconventional gas producing horizon in Canada with over 3100 active gas wells (NRCAN, 2016a). The Montney is a massive resource play that reaches across northwest Alberta and into northeast British Columbia. Flatbed Creek located near Montney play within BC is selected for assessment. The creek orginates in the Rocky Mountains passes by 110 Heritage Highway in the Peace River regional district and flows North-West into Murray River (CEAA, 2012). The path of the creek is as traced in Figure 4.1. The creek is selected because it is located in an area of intensive unconventional gas productions and there are sufficient hydrogeological data available for the risk assessment including data for background concentration of contaminants and monthly discharge value. The selection was creek was done to determine the possible environmental effects on smaller discharge suface water body. The Flatbed Creek basin covers an area of 486 km2 and has an average annual discharge of 4.14 m3/s (BCOGC, 2016). An accidental spill or uncontrolled release of flowback water is assumed to have occurred from AWSS type of storage system located 200 m near the creek.  Figure 2.5 illustrates major components of the AWSS. The AWSS is an open, bottomless, insulated, short cylindrical system lined with heavy duty tarp liner clamped on the shell. It consists of structural walls system, a diked form of the secondary containment system, liner system, a pipeline system leading up to AWSS with an inlet/outlet pump and inlet/outlet valve. The flowback water is discharged into the AWSS via pipelines. There are no openings in the walled structure. The pipelines go over the shell. The AWSS, liner, pipeline, valve, and pump consist of the primary storage system. A secondary containment system in the form of a dike is constructed immediately surrounding the primary storage system.71     Figure 4-1 Flatbed Creek and AWSS located at 200m from the creek near Montney in BC5                                                          5 Source: iMapBC http://maps.gov.bc.ca/ess/sv/imapbc/. Copyright 2013 Province of British Columbia. All rights reserved. 72 4.2 Data Collection Variety of data was collected from literature in line with the study area described above and the method proposed to conduct the hypothetical case study in order to demonstrate the application of the assessment.  4.2.1 Data for BIA framework The most important and challenging component of BIA was understanding the failure modes of the system which is relatively new and does not seem to have been studied from the current literature review. The failure modes for the AWSS were determined by understanding the failure modes of its sub-systems and their components. The failure modes of shell like structural wall system and secondary system of AWSS were deduced by studying relevant failure modes and their mechanism of the tanks and containment ponds type of storage system from the existing literature including API (2014), Atherton et al. (2008), CCPS (2000b), Choi & Chang (2016), Cunat (2002), Kuwayama et al. (2015), Nakashima (2010), Trebuňa (2009), US EPA (2009, 1986a, 1986b, 2002). The liner failure was studied from sources like Cheng et al. (2009), Klimchuk et al. (2016), Peggs (2009), Robeson (2013), US EPA (1985, 1996). Pipeline and other components failure were investigated from Shahriar et al. (2012) and US EPA (1986b, 1986c).  Site specific information about AWSS failure modes as well as the occurrence, severity and detectability ranking was obtained from Secure Energy Services Inc. and visit their AWSS storage facilities in Grand Prairie, Alberta, Canada (C. Krauskopf and G. Dickie, personal communication, May 4, 2017).  4.2.2 Data for ERA framework The ERA will require data of the stressor, i.e., flowback water and data of the receving water body, i.e., Flatbed creek. The flowback water generated from the wells in Montney are sampled, tested and registered at IHS AccuMap as a database implemented by the BCOGC (IHS Accumap, 2016). The data for flowback of the wells in the Montney shale gas play from 2009 to 2016 were collected and cleaned. 212 wells were shortlisted based on the sampling date and well completion date to ensure the flowback period of 28 days. The contaminants and their 73 concentrations in the flowback water are as shown in Table 4.1. It is not a requirement to register flowback water quality during hydraulic fracturing operations; hence, it is difficult to determine if the enlisted contaminants in Table 4.1 represent all the constituents of flowback water. Circumstantial evidence shows the presence of NORMs, hydrocarbons, and grease in the flowback water; however, they were not included in the assessment due to lack of reliable data source.   The monthly average flow and the average background concentrations of serveral inorganic chemicals in the Flatbed creek are shown in Table 4.2. The monthly and average annual discharge rate of the creek is also reported. The information presented in Table 4.2 for Flatbed Creek was collected from BC Water Portal (BCOGC, 2016).  Table 4-1 Contaminants and their concentrations found in flowback water samples collected in Montney Contaminants  Mean (mg/L) Median (mg/L) Maximum (mg/L) Minimum (mg/L) TDS  93079 90037 260000 1458 Sodium 26852 25000 69700 96 Potassium 2148 1050 18200 16 Calcium 5282 4950 17164 80 Magnesium 797 708.5 8427 13.4 Barium 212 20.85 2470 0.61 Strontium 618 551 1573 3.8 Iron 33 26 240 0.2 Manganese 3 1.75 15 0.46 Chloride 57278 54869 177100 208 Bromide 259 196.5 1150 13.4 Iodide 12 10.4 41.1 0.6 Bicarbonate  450 190.5 3363 19.2 Sulfate 247 137 3499 2.4 Carbonate 121 77.6 416 14.1 Hydroxide 203 203.4 203.4 203.4 Hydrogen Sulfide 257 255.6 511.2 8.5  The impoundment is assumed to store flowback water with the mean concentration value of each contaminant (Table 4.1). Based on the historic spill volumes, a spill of 1000 L is assumed to occur over a period of 96 hours (US EPA, 2015).The surrounding soil is assumed to be loamy 74 clay with a bulk density of 1.5 g/cm3 and an effective porosity of 0.39 (Clark et al., 1962; Rawls et al., 1982; US EPA, 2005). The affected soil area is assumed 100 m wide and 5 cm deep. For deriving acute toxicity guidelines, an extensive literature review was carried out and secondary toxicity data was collected for lethal and sub lethal effects on fish, crustaceans, amphibians, molluscs, worms and aquatic plant (duckweed). These organisms were tested in the lab in freshwater media for an exposure duration of 96 hours or less for the endpoints LC50 and EC50. Considering the road salt toxicity studies, relatively larger amount of toxicity data exists for chloride ion. However, not enough data for sodium and calcium ions. Also, experimental studies show that the toxicity of the Na+ and Ca+ salt ions is negligible and can be attributed to the corresponding anion (Mount et al., 1997). The sufficient data for deriving acute freshwater toxicity of barium, potassium, magnesium, iodide, bicarbonate, carbonate and hydroxide were not available as required by the guidelines defined in Figure 3.7.  Table 4-2 Background concentrations and monthly discharge rate of the Flatbed Creek (BCOGC, 2016) Chemicals Average Concentration (mg/L) Month Discharge (m3/s) Month Discharge (m3/s) Month Discharge (m3/s) Barium  0.2 Jan/Feb 0.5 Jun 11.1 Oct 2.1  Iron  0.0843 Mar 0.6 Jul 7.1 Nov 1.5  Manganese 0.00399 Apr 4.8 Aug 3.3 Dec 0.8  Strontium  0.111 May 15.1 Sep 2.8 Annual Mean 4.1  Table 4-3 Physicochemical properties of the contaminants (Bright & Addison, 2002; US EPA, 1988, 1999, 2005, 2016f) Contaminant Kd(mL/g) log Kow Molecular weight (g/mol) Water Solubility Barium 4.1E+01 2.3E-01 1.3E+02 5.8E+03 Hydrogen Sulfide 9.9E+00 2.3E-01 3.4E+01 3.7E+03 Iron 2.5E+01 -7.7E-01 5.5E+01 3.5E+05 Manganese 6.5E+01 2.3E-01 5.4E+01 6.6E+04 Strontium 1.0E+00 2.3E-01 8.7E+01 6.7E+03 Chloride 0.0E+00  3.5E+01  Sulfate 0.0E+00  9.6E+01  Bromide 0.0E+00  7.9E+01  75 Appendix A lists the dataset used for the derivation of acute toxicity effect concentration values. These data are collected from ECOTOX (US EPA, 2016), CCREM (2008), McPherson, et al. (2014), Pacholski (2009), Hedayati, et al. (2015), Shuhaimi-Othman, et al. (2013) and other sources. The physicochemical properties of contaminants, including the values of adsorption coefficient, log octanol water partition coefficient, molecular weight and water solubility are shown in Table 4.3 (Bright & Addison, 2002; US EPA, 1988, 1999, 2005, 2016f). Chloride, sulfate, and bromide are usually reported to travel via soil columns at the same rate or faster than water in their dissolved form (US EPA, 1999). Thus, their adsorption coefficient was assumed zero (Bright & Addison, 2002). 4.3 Implementation of BIA Framework A fault tree entails a failure analysis of sub-system and components of the AWSS. The top failure event is identified as the uncontrolled release of flowback water. The failure modes for walled structure, liner, pipe, valve, pump and secondary containment system are assessed. Each failure event of the fault tree is described in Table 4.4 with the corresponding probability of failure values. It should be noted that the analysis is can have many variations to the fault tree and can be constructed in several ways. This is an example of the fault tree to the required level of detail for the analysis.  The basic failure events of the fault tree and the failure rates are used as inputs to the FMEA in BIA. Severity, occurrence, and detectability were ranked based on the scale given in Table 3.3 (C. Krauskopf and G.Dickie, personal communication, May 4, 2017). Detectability of the failure event is only through manual monitoring. Based on this assumption, Table 4.5 shows the FMEA table for failure modes as ranked using risk priority number (RPN). The failure modes are prioritized based on the calculated RPN and are arranged in decreasing order of their RPN. The higher the RPN, the higher the rank. For example, “Penetrations through liner” is a potential failure mode having the highest RPN of 30 and is ranked 1.   76 77  Figure 4-2 Fault tree for uncontrolled release of flowback water from AWSS failure6                                                          6 Secondary system is the berm surrounding the primary system. Primary system includes walled structure, lined system, piping system 78 Table 4-4 Description and probability of fault tree events Failure Event No. P7 Description Uncontrolled release of flowback water G1  Undesired top event for overall risk evaluation Failure of primary storage system G2  Failure of the primary storage system Failure of secondary containment system (dike) G3  Uncontrolled release of flowback water due to secondary containment (dike) failure Overflow of the AWSS G4  Overflow of flowback water from the AWSS Primary storage system develops leak or ruptures G5  Failure of the primary storage system by developing leak or rupturing External catastrophe G6  Failure of the primary storage system due to external catastrophe Spill during loading/unloading G7  Release of flowback water due to spill during loading/unloading Spill during thawing G8  Release of flowback water during thawing Improper construction G9  Failure due to improper construction of the dike Water level control error G10  Failure to control the overflow of flowback water System shutdown failure G11  Failure of the system to shutdown the inflow of water AWSS failure G12  C-ring tank leaks or ruptures Pipe failure G13  Flowback water release due to pipe failure Liner failure G14  Leak or rupture of storage system due to liner failure Operator error G15  Error of the operator in shutting down the system Mechanical failure  G16  Device failure to shutdown the system AWSS corrodes G17  Release of flowback water due to corrosion of C-ring AWSS ruptures G18  Release of flowback water due to rupture of C-ring Pipe puncture G19  Failure of pipe due to puncturing  Pipe rupture G20  Pipe failure due to rupture of pipe Stress Cracking on the liner G21  Liner failure by stress cracking at the break strength of the liner Mechanical Failure of the liner G22  Failure of liner by cracking and breaking Human error G23  Liner failure due to human error Corrosion of the bottom HSS G24  Corrosion of the hollow steel section in contact with the ground AWSS ruptures right after installation G25  C-ring fails as soon as installed Undetected faulty installation G26  Failure of the c-ring due to undetected faulty installation Corrosion thinning of piping G27  Puncture of pipe due to corrosion thinning Pipe defect G28  Puncture of pipe due to pipe defect Corrosion fatigue G29  Pipe failure due to corrosion fatigue of pipe External stress on the liner G30  Stress cracking of the liner due to excessive external stress in presence of acidic water Workmanship error G31  Liner failure due to poor or erroneous workmanship                                                           7 Probability of failure; applies to only basic events. Basic events are numbered xn 79 Failure Event No. P7 Description Adverse ground conditions G32  Corrosive ground characteristic  Piping external corrosion G33  Puncture of pipe due to external corrosion Piping internal corrosion G34  Puncture of pipe due to internal corrosion Operational defect G35  Puncture of pipe due to operational defect Alternate stress G36  Corrosion fatigue due to development of tensile stress Installation inefficiency G37  Inefficient installation of liners induces failure by stress cracking Traffic from people and equipment x1 1E-05 Secondary containment failure (dike) due to traffic from people and equipment Animal and plant interruption x2 1E-05 Burrows or holes created by animals and roots of plants voids in the embankment and weakening the dike Erosion of the berm x3 1E-04 Failure of the dike due to erosion Insufficient dike capacity x4 1E-05 Overflowing of flowback water due to insufficient dike capacity AWSS nearly full x5 5E-05 Water level in AWSS above freeboard Strong winds x6 5E-05 Failure of the primary storage system due to high winds Earthquake x7 1E-05 Failure of the primary storage system due to earthquake Flooding x8 1E-05 Failure of the primary storage system due to flooding Fire/Explosion x9 1E-05 Release of flowback water due to fire or explosion Flexible hose rupture x10 1E-03 Spill during filling or discharging due to rupture of flexible hose Loose flexible hose connection x11 1E-03 Spill during filling or discharging due to loose flexible hose connection Flexible hose rupture x12 1E-05 Spill during thawing due to rupture of flexible hose Loose flexible hose connection x13 1E-05 Spill during thawing due to loose flexible hose connection Bulking x14 5E-05 Dike failure due to bulking Improper material used x15 1E-05 Dike failure due to standing water at surface Incorrect slope angle x16 1E-05 Failure of dike due to construction at incorrect slope angle  Failure to monitor the water level x17 5E-05 Failure to monitor the flowback water level in AWSS Failure of inlet/outlet valve x18 1E-05 Failure of the inlet/outlet valve to shutdown allowing/restricting water inflow Commission error x19 1E-05 Error in commissioning the system shutdown Omission error x20 1E-05 Error in acting correctly to shutdown the system Failure of inlet/outlet pump x21 5E-05 Failure of the inlet/outlet pump to shutdown allowing/restricting water flow Corrosion under insulation of the wall x22 1E-05 Corrosion of C-ring plates under insulation Overloading of pipelines x23 1E-05 Pipe failure due to overloading of pipe Acidic water x24 1E-05 Stress cracking of the liner due to excessive external stress in presence of acidic water Inadequate ground preparation x25 5E-05 Failure to remove debris and smoothen the ground adequately 80 Failure Event No. P7 Description Clamps improperly secured x26 1E-05 Tear of liner due to improperly securing clamps Wear and tear due to aging* x27 1E-05  Inspection error x28 1E-05 Failure of liner due to error or omission in inspecting/reporting/acting upon areas of poor workmanship Failure of bottom HSS coating x29 1E-05 Protective coating of the hollow steel section fails Settlement of ground x30 1E-05 Settlement of soil due to c-ring installation Inadequate component strength x31 1E-05 Rupture of C-ring due to inadequate component strength Vehicle /fork lift collision x32 1E-05 C-ring ruptures due to accidental collision with the working equipment Tilting or displacement of AWSS* x33 1E-05 AWSS ruptures due to tilting or displacement C-ring plate or frame damaged during or before installation x34 1E-05 Failure of the c-ring due to damage to c-ring plate or frame during or before installation Piping material defect x35 1E-05 Puncture of pipe due to defect in pipe material Overburden Stress on the liner x36 1E-05 Liner failure by induced stress due to high overburden pressure  Wrinkles on the liner x37 1E-05 Wrinkles can fold over under load and can create weakened areas causing line failure Bridging of the liner x38 5E-05  Liner is lifted off its support at concave areas of the pond, toe, corners, etc. putting excessive stresses on liner Penetrations through liner x39 1E-03 Penetrations in a lined containment unit, increasing the potential for breaches Dropped tools on liner x40 1E-05 Accidental dropping of tools causing tear of liner due to impact, stress Adverse soil conditions x41 1E-05 Presence of salts of chloride in soil Water accumulation near bottom HSS x42 1E-05 Standing water near the HSS Failure of exterior coating of a pipe x43 1E-05 Corrosion of piping due to failure of external protective coating  Failure of interior pipe coating x44 1E-05 Corrosion of piping due to failure of internal coating  Bad piping installation x45 1E-05 Pipe failure due to bad installation of pipe Bad piping weld x46 1E-05 Pipe failure due to faulty welding of pipes Pressure surge in the pipe x47 1E-05 Development of tensile stress due to pressure surge in the pipe Excessive External load on the pipe x48 1E-05 Development of tensile stress due to external load on the pipe Local indentation on the liner x49 5E-05 Failure of liner due to local dents created while installing developing weak areas in the liner Waviness of the liner x50 1E-05 Liner failure due to waviness creating local stresses  Scratching on the liner x51 1E-05 Failure of liner due to lowered tensile resistance in the areas of scratch Impact by installation equipment x52 1E-03 Failure of liner due to impact by installation equipment 81 Table 4-5 FMEA table for basic failure events determined through FTA Potential Failure Modes Potential Effects of failure modes S Potential causes of failure O D RPN Penetrations through liner Weakened areas in the liner due to development of local stress 3 Debris and plants not cleared off site 5 2 30 Impact on the liner by installation equipment Liner susceptible to tear, breach, puncture 3 Inappropriate handling of the liner during installation 5 2 30 Bridging of the liner Increased stress on the liner due to absence of immediate support  4 Lifted off liner's support at concave areas of the c-ring like toe, corners, etc. putting excessive stresses on liner 2 2 16 Local indentation in the liner Causes stress cracking of the liner 3 Dragged debris, equipment, and hoses 2 2 12 Flexible hose rupture during filling and discharging Flowback water leakage through the ruptured hose 2 Wear and tear of the hose, Clogging of the hose 5 1 10 Corrosion under insulation Stress corrosion cracking 10 Corrosive environment and applied/residual tensile stress 1 1 10 Inadequate component strength of the shell AWSS rupture and deformation 10 Design error/defective material/improper use of the component 1 1 10 Vehicle /fork lift collision with the shell AWSS deformation 10 Equipment/vehicle operator carelessness 1 1 10 Wear and tear due to aging Tearing of liner 5 Reuse of liner 2 1 10 Earthquake Deformation and crack in shell, Spillover of the stored flowback water 10 Natural or induced seismicity 1 1 10 Flooding Spillover of the stored flowback water, Deformation of the shell 10 Heavy precipitation 1 1 10 Fire/Explosion Liner failure 10 Defective liner material 1 1 10 Fire/Explosion Pipe failure 10 Presence of H2S in the flowback water 1 1 10 Fire/Explosion Deformation, cracks, rupture of the shell  10 Presence of H2S in the flowback water 1 1 10 High winds Overflow of the flowback water 4 Wave action in the shell 2 1 8 Cring nearly full Flowback water level rising above the free board level 3 Increased inflow of water 2 1 6 Cring nearly full Flowback water level rising above the free board level 3 Operational error of inaction or incorrect action 2 1 6 Failure to monitor water level in C-ring Rising water level in the c-ring with a potential of overflow 3 Failure of water level detection system 2 1 6 Improper erection ground preparation Failure of liner due to excessive stress 1 Uneven ground surface 2 3 6 82 Potential Failure Modes Potential Effects of failure modes S Potential causes of failure O D RPN Improper erection ground preparation C-ring rupture right after installation 1 Inferior ground strength 2 3 6 Clamps loosely secured Flowback water release to the ground 3 Falling off the liner from c-ring shell walls 1 2 6 Dropped tools on liner Tearing of liner 3 Impact and stress caused by dropping of tools 1 2 6 Waviness of the liner Physical failure of the liner 3 Development of local stress 1 2 6 Scratching of the liner Lowered tensile resistance in the areas of scratch 3 Improperly installed liner 1 2 6 Loose flexible hose connection during filling and discharging Flowback water leakage through the hose connection 1 Failure to connect properly 5 1 5 Loose flexible hose connection during filling and discharging Flowback water leakage through the hose connection 1 Failure to inspect connection 5 1 5 Loose flexible hose connection during filling and discharging Flowback water leakage through the hose connection 1 Faulty hose fitting/adaptor 5 1 5 Traffic from people and equipment Disturbance to the dike structure 1 Insufficient compaction of the dike 1 3 3 Traffic from people and equipment Erosion of slope 1 Use of Low-density, saturated, cohesion-less soils 1 3 3 Animal and plant interruption Weakening of the structure 1 Formation of holes or dents 1 3 3 Erosion of the dike Reduced capacity of the dike 1 Wind and water action 3 1 3 Commission error Failure to stop inflow of flowback water 3 Failure to detect cause of overflow 1 1 3 Omission error Failure to stop inflow of flowback water 3 Failure to monitor water level 1 1 3 Bulking of dike Failure/weakening of the dike  1 Improper construction of the dike 2 1 2 Incorrect slope angle of the dike Instability of the dike 2 Construction/design error 1 1 2 Failure of inlet valve Lost control over inflow of flowback water 2 Wear and tear of the valve/fatigue 1 1 2 Failure of inlet valve Lost control over inflow of flowback water 2 Impact load 1 1 2 Failure of inlet valve Lost control over inflow of flowback water 2 Corrosion of the valve 1 1 2 Failure of inlet pump Leaky pumps 1 Failure of mechanical seal of the pump 2 1 2 83 Potential Failure Modes Potential Effects of failure modes S Potential causes of failure O D RPN Settlement C-ring rupture and deformation right after installation 2 Inferior ground strength/liquefaction 1 1 2 AWSS plate or frame damaged during or before installation Formation of a dent in the plate/frame 2 Impact during loading, unloading or installation 1 1 2 Adverse soil conditions Soil corrosion 2 Presence of chloride, adverse pH levels of the soils 1 1 2 Failure of exterior coating of a pipe Outer surface of the pipe susceptible to corrosion 2 Acidic water/corrosive soil attack 1 1 2 Failure of interior pipe coating Inner surface of the pipe susceptible to corrosion 2 Acidic water/corrosive soil attack 1 1 2 Insufficient dike capacity Overflow of the flowback water 1 Construction/design error 1 1 1 Flexible hose rupture during thawing Flowback water leakage through the ruptured hose 1 High temperature 1 1 1 Flexible hose rupture during thawing Flowback water leakage through the ruptured hose 1 High vapor pressure 1 1 1 Flexible hose rupture during thawing Flowback water leakage through the ruptured hose 1 Clogging of the hose 1 1 1 Loose flexible hose connection during thawing Flowback water leakage through the hose connection 1 Failure to connect properly 1 1 1 Loose flexible hose connection during thawing Flowback water leakage through the hose connection 1 Failure to inspect connection 1 1 1 Loose flexible hose connection during thawing Flowback water leakage through the hose connection 1 Faulty hose fitting/adaptor 1 1 1 Standing water on the top of dike Failure/weakening of the dike due to water seepage 1 Precipitation/melting of snow 1 1 1 Overloading of pipelines Rupture of pipe 1 Clogging of the pipelines 1 1 1 Acidic water Corrosion of the storage system components 1 Failure to take corrective action upon escape of acidic flowback water  1 1 1 Inspection error Faulty C-ring installation 1 Failure to conduct corrective actions before initiating operations 1 1 1 Failure of bottom HSS coating C-ring collapse or cracking 1 Failure of the c-ring shell support system 1 1 1 Piping material defect Puncture of pipe 1 Inappropriate material selection for the given quality of flowback water 1 1 1 Overburden Stress on the liner Increase in external stress on the liner 1 High overburden pressure 1 1 1 84 Potential Failure Modes Potential Effects of failure modes S Potential causes of failure O D RPN Wrinkles on the liner Weakened areas in the liner due to development of local stress 1 Liner installation inefficiency 1 1 1 Acidic water accumulation near bottom HSS Corrosion of the bottom HSS 1 Breakdown of the protective coating 1 1 1 Bad piping installation Erroneous operation of the pipe 1 Inexperienced piping installation personnel 1 1 1 Bad piping weld Weakened piping connection at the faulty weld leading to pipe puncture 1 Inexperienced piping welding personnel/Adverse welding conditions 1 1 1 Pressure surge in the pipe Increased pressure on the surface of the pipe 1 Change in flow velocity of the flowback water 1 1 1 Excessive External load on the pipe Rupture of pipe 1 Heavy equipment cut across pipeline 1 1 1 85 Table 4-6 Probability of failure calculation Gate Calculation Value Gate Calculation Value G37 x49+x50+x51+x52 0.001 G24 x29*G32 2E-10 G36 x47+x48 2E-05 G23 x28*G31 1.1E-08 G35 x45+x46 2E-05 G22 x2+x26+x26+x27 8E-05 G34 x24*x42 1E-10 G21 x24+G30 0.001 G33 x44*x41 1E-10 G20 x23+G29 8E-05 G32 x42+x43 2E-05 G19 G27+G28 3E-05 G31 x41+G37 0.001 G18 G25+G26 9E-05 G30 x37+x38+x39+x40 0.001 G17 x22+G24 1E-05 G29 x36+G36 7E-05 G16 x18+x21 6E-05 G28 x35+G35 3E-05 G15 x19+x20 2E-05 G27 G33+G34 2E-10 G14 G21+G22+G23 0.001 G26 x34*x28 1E-10 G13 G19+G20 1E-04 G25 x30+x31+x32+x26+x33 9E-05 G12 G17+G18 1E-04 G11 G15+G16 8E-05 G5 G12+G13+G14 0.001 G10 x17*x18 5E-10 G4 G10*x5*G11 2E-18 G9 x14+x15+x16 7E-05 G3 x1+x2+x3+G9+x4+G6 0.0002 G8 x12+x13 2E-05 G2 G4+G5+G6+G7+G8 0.003 G7 x10+x11 0.002 G1 G2*G3 1E-06 G6 x6+x7+x8+x9 8E-05     The probability of the uncontrolled release of flowback water is calculated using the gate to gate analysis. The detail calculation for each gate is shown in Table 4.6. It is to be noted that the 𝑝 value derived here is the upper bound value and that the probability cannot be higher than this. The value of 𝑝 for the identified top event is computed to be 1E-6. The present numbers represent the upper bound of the probability of failure. The BIA prioritizes failure of liner due to penetrations, impact by installation equipment, liner bridging and local indentations as top failure events having high severity but low detection rank. The failure of the liner is followed by flexible hose rupture during filling discharging. Next in ranking are corrosion of the shell under insulation, inadequate component strength, and collision with vehicles on site. The failure due to natural disaster follows right after. Even though the probability of a natural event is low, their effects are extremely severe.  86 The corrective actions, like providing two liners, providing rig-mats, monthly inspection of liners, providing extensions on AWSS shells to avoid bridging, taking extra measures to avoid splashing flowback water over insulated shell should be taken. These can provide cost benefit in long term by avoiding failure of the system. Regulations pertaining to the AWSS or above ground walled storage system, discussed in chapter 2, provide an insight to some of the issues related to the storage of flowback water. The failure modes are discussed in line with those regulations in Section 4.6. Although the corrective actions are not discussed here, the analysis gives valuable information to reduce risk. 4.4 Implementation of ERA Framework Based on the data availability as assessed in Appendix A, the toxicity of ferric ion, sulfate, manganese ion, bromide, chloride, and hydrogen sulfide was derived using Type A guidelines. Strontium toxicity was derived using Type B2 guidelines. Significant data gaps were identified for magnesium, potassium, sodium, and iodide. For Type A analysis, acute toxicity curves for each ion was derived using @Risk and US EPA’s species sensitivity distributions (SSD) generator and overlaid to compare the derived effect concentration values (Palisade Corporation, 2015b; US EPA, 2016e). SSD generator is an excel template to calculate and plot the proportion of species affected (US EPA, 2016e). The 5th percentile from the best fit distribution and lognormal distribution are derived from @Risk to compare with the SSD plots.  Table 4.7 and Figure 4.3 shows the results from the @Risk and SSD generator. The tables and curves in the graph represent the central tendency, the upper and lower 95% prediction interval, and the best-fit and lognormal distribution from @Risk. These values are compared with the established standard acute, chronic or drinking water concentrations in Canada, US, New Zealand and Australia. Based on the availability, guidelines for acute toxicity, chronic toxicity for aquatic or marine life or drinking water quality, are used.  The effects concentration of chloride is very close to the US Standard but much higher than the Canadian standard. This is because Northern riffle shell mussel,  the most sensitive species, was not used in this analysis as this species is resident to Ontario (GC, 2016b). The background concentration of the contaminant in the creek are safely below the derived effect concentration 87 values, hence they do not affect the derived values. The effect concentration of strontium is found to be 7.5 mg/L. The chronic toxicity threshold value of strontium in fresh water is 21 mg/L (McPherson et al., 2014; Pacholski, 2009). Lower prediction interval effect concentration values are used in risk quotient except for hydrogen sulfide. The lower PI value for hydrogen sulfide is too low as compared to the estimated concentration using the best-fit and lognormal values of @Risk and the toxicity threshold values in Canada, US, Australia and New Zealand. The threshold chronic toxicity value in Australia-New Zealand is the lowest (0.001), followed by the US (0.002) and they are comparable with the lognormal estimate of @Risk and a central estimate of the SSD generator. For this reason, the central estimate is considered for the analysis. SSD curves of bromide, chloride, hydrogen sulfide, iron, manganese, and sulfate, show a similar trend for central curve developed from SSD generator and log normal curve generated from @Risk. The best-fit distribution for chloride, iron, manganese, bromide, sulfate and hydrogen sulfide are gamma, kumaraswamy, pearson5, kumaraswamy, kumaraswamy, and levy, respectively. The central, best fit and log normal (@Risk curve) for chloride perfectly coincide. The highest number of data points are available for chloride, better curves can be generated for the larger data set.  Table 4-7 Effect concentrations of contaminants   Contaminant  (mg/L) SSD Generator 11 Result @Risk® Toxicity Threshold Values Upper PI Lower PI Central Best Fit Log Normal Canada Reference US Reference Aus-NZ1 Bromide 527.1 5.61 54.37 0.39 48.82 6 mg/L** (WHO, 2009) Chloride 916.1 553.5 712.42 579.1 720.2 640 (CCME, 2011) 860 (US EPA, 2016d) 400* Hydrogen Sulfide 0.02 0.0001 0.001 0.01 0.001 0.05** (CCREM, 2008) 0.002* (US EPA, 2016d) 0.001* Iron 3.74 0.56 1.45 0.02 2.25 1*** (Phippen, et al., 2008) 1* (US EPA, 2016d) 0.3* Manganese 8.88 1.46 3.60 6.42 4.95 1.6 (BC MOE, 2001) NA - 1.9 Sulfate 830.28 240.86 447.19 22.47 540.68 309* (BC MOE, 2013) 250* (US EPA, 1986c) 400* Notes:*Chronic toxicity value; **Drinking water threshold value; ***Province or State-level chronic toxicity values; 1(ANZECC & ARMCANZ, 2000) 88 a.   b.    00.10.20.30.40.50.60.70.80.9110.00 100.00 1000.00 10000.00 100000.00 1000000.00Proportion AffectedConcentration (mg/L)Bromide00.10.20.30.40.50.60.70.80.91100.00 1000.00 10000.00 100000.00Proportion AffectedConcentration (mg/L)Chloride89 c.   d.   00.10.20.30.40.50.60.70.80.910.000 0.000 0.001 0.010 0.100 1.000 10.000 100.000 1000.000 10000.000Proportion AffectedConcentration (mg/L)Hydrogen Sulfide00.10.20.30.40.50.60.70.80.910.01 0.10 1.00 10.00 100.00 1000.00 10000.00Proportion AffectedConcentration (mg/L)Iron90 e.   f.   Figure 4-3 a-f Cumulative distribution curves for flowback water constituents  00.10.20.30.40.50.60.70.80.911.00 10.00 100.00 1000.00Proportion AffectedConcentration (mg/L)Manganese00.10.20.30.40.50.60.70.80.9110 100 1000 10000 100000 1000000Proportion AffectedConcentration (mg/L)Sulfate91 Table 4-8 Effect of Kd and dilution coefficient on exposure concentration Contaminants Kd  (L/mg) Kd*Msoil Initial Conc. (mg/L) (Co) Remaining Conc. (mg/L) (Ci) Final Conc. (mg/L) (C∞) Bromide 0E+00 0E+00 259 259 0.00018 Chloride 0E+00 0E+00 57278 57278 0.040 Hydrogen Sulfide 1E-05 1E+07 257 103.42 0.0000012 Iron 3E-05 4E+07 33 6.94 0.08 Manganese 7E-05 1E+08 3 0.28 0.004 Sulfate 0E+00 0E+00 247 247 0.00017 Strontium 1E-06 2E+06 618 537.39 0.11  For a spill of 1000 L, the effect of adsorption coefficient on the initial concentration of the contaminant is shown in Table 4.8. The concentration of the contaminants will further decrease on dilution. As per Table 4.3, the log  𝐾𝑜𝑤 values are less than 5, hence,  𝐵𝐴𝐹 would be equivalent to 1. The reduction in the concentration of the contaminants is much higher due to dilution than due to adsorption. Table 4-9 Estimated exposure concentration and risk quotient of contaminants Contaminants Exposure Conc. (mg/L) (EEC) Background Conc. (mg/L) (Cw) Effect (mg/L) (EC) Risk Quotient R Bromide 0.00018 0 5.61 3.3E-05 Chloride 0.040 0 553.49 7.3E-05 Hydrogen Sulfide 0.0000012 0 0.001 1.1E-03 Iron 0.08 0.08 0.56 1.4E-01 Manganese 0.004 0.004 1.46 2.7E-03 Sulfate 0.00017 0 240.86 7.2E-07 Strontium 0.11 0.11 7.5 1.5E-02  For the spill of 1000 L over a period of 96 hours, the final exposure concentration of the contaminant in the stream for the mean annual discharge (4400 L/s) and the corresponding risk quotient for each contaminant would be as shown in Table 4.9.  The final risk quotient can be calculated by multiplying the cumulative risk quotient with the probability of failure calculated in section 4.3. For the risk equation 3.12, the probability of failure of the system was found to be 1E-6, the cumulative risk quotient is 0.16 and overall risk of the system would be 1.6E-7. This 92 value is less than one; hence, there is no significant risk to the system for the given set of assumptions.  The risk estimate value is computed using conservative lower PI effect concentration value and typical average initial exposure concentration data to yield a point estimate of exposure and risk that is different from a true but unknown value. Hence, uncertainty analysis for the estimated exposure concentration and effects concentration was carried out for the given volume of flowback water spill and creek discharge.  Using Monte Carlo simulations, ten thousand iterations were performed to investigate the possible risk quotient values. Probability distributions were defined for the uncertain input parameters. The effect concentration values obtained from SSD (upper prediction, lower prediction, central) and from @Risk (best fit, lognormal) were used to create distributions for effect concentration. Similarly, the distribution curves for an initial concentration of each contaminant was created from the Montney well dataset downloaded from IHS AccuMap using @Risk (IHS Accumap, 2016; Palisade Corporation, 2015b). Monte Carlo simulation was carried out for risk assessment assuming probability as 1, that the uncontrolled release of flowback water has occurred. The uncertainty analysis for the range of initial concentrations of the flowback water constituents and the effect concentration values derived would yield a probability distribution for the risk quotient. Figure 4.4 shows the result for the risk after 10,000 iterations. From this Monte Carlo simulation, a 90% confidence interval of [3.1E-2, 7.7E-1] is attained and is shown by the marks as shown on the graph provided in Figure 4.4. This indicates that after considering the uncertainty caused due to the EC and Co parameters, at a very confident subjective level of 90%, it can be said that the true value of RQ should lie between 3.1E-2 and 7.7E-1. The upper confidence limit, at 95%, is an RQ of 1, it can be said with high confidence that the most of the exposed aquatic organisms for this scenario would be exposed to a low level of risk. 93  Figure 4-4 Results of uncertainty analysis for risk quotient Sensitivity analysis results are displayed in Figure 4.5 using Tornado graphs. These results show the sensitivity of each output variable to the input distributions in the model. Since enough data points were not available for the effect concentration of strontium and initial concentration of hydrogen sulfide, they do not qualify for sensitivity analysis. The effect concentration of iron is the most sensitive (i.e., 0.04-3.9) followed by manganese, chloride, sulfate and hydrogen sulfide are cluttered near the risk values of (0.31-1.1). Moreover, the sensitivity of the initial concentration of manganese in flowback water is the highest (i.e., 0.31-1.1) and the rest are cluttered near the risk values of (0.27-1). The range of uncertainty in risk is not very high.   Scenario analysis for the risk quotient is carried out for the monthly discharge of the creek scenarios assuming probability to be 1, for a range of spill volume. A ratio of the volume of the spill (100 to 50000 L) and monthly creek discharge (500 to 15100 L/s) was plotted against risk quotient as shown in Figure 4.6. The polynomial trendline had R2 approaching 1, hence it is the best fit for the given scenario. It can be deduced that the relationship between the ratio volume of spill and creek discharge and risk quotient is non-linear. The highest (Vs/Qw) ratio achieved for risk quotient to be less than 1 is 4.5. 94  Figure 4-5 Sensitivity analysis of risk quotient for effect concentration and initial exposure concentration  Figure 4-6 Variation in cumulative risk quotient as a function of the ratio of spill volume to creek discharge  0.040.270.250.310.270.300.280.320.280.303.91.21.01.11.00.90.90.90.90.80 0.5 1 1.5 2 2.5 3 3.5 4EC of FeEC of MnEC of ClCo of MnEC of SO4EC of H2SCo of ClCo of FeCo of SrCo of SO4Risk Quotienty = 0.0009x2 + 0.2402x - 0.1153R² = 0.97480.001.002.003.004.005.000.00 5.00 10.00 15.00 20.00Risk QuotientVs/Qw95 4.5 Risk Description and Discussion The values of the effect concentration obtained from the central of the SSD curve and the lognormal @Risk curve were comparable with the available acute toxicity values of chloride and hydrogen sulfide. This analysis also brings out the data gaps in the available acute fresh water toxicity values of barium, potassium, calcium, magnesium, iodide, bicarbonates and hydroxide.  As per Table 4.8, the adsorption coefficient greatly reduces the exposure concentration of manganese and iron. The single point risk estimate for spill volume of 1000L and stream discharge of 4400 L/s suggested that the ecology would be potentially safe in an event of a spill. However, when a range of effect concentration and initial concentration values of the flowback water constituents is considered, the extreme values of risk suggest the adverse effect to the ecology in an event of spill cannot be completely overlooked.  Also, according to scenario analysis, a spill volume larger than 7000 L from a storage system at 200 m from the surface water body can possibly cause adverse effects to the ecology. A maximum ratio of 4.5 between spill volume and stream discharge is to be maintained to ensure the safety of the aquatic ecology. The flowback water systems are often temporary storage system and can be used anywhere from 2 months to 2 years. Risk analysis using monthly average creek flow is recommended as compared to annual mean average as the risk estimate value is very sensitive to spill and stream discharge value.  Through BIA, the probability of failure of the system was found to be 1E-6. The AWSS with a secondary containment system was found to be safe. The critical failure modes of the AWSS were ranked and liner failure, spill during loading and unloading, corrosion under insulation of the wall of AWSS were ranked the highest. A study conducted by Ziemkiewicz et al. (2014) for containment ponds found liner tear as a low likelihood event carrying a high severity as the tear can cause direct exposure pathway to the environment. FMEA within BIA also states the higher likelihood of the failure for this failure mode. The storage system was also assessed for failure due to a catastrophe like an earthquake, flooding, fire/explosion, and high winds. These failure modes are ranked 10,11 and 12 respectively out of 89 potential failure modes identified due to a combination of lowest occurrence value and highest severity and detectability value. 96 The overall risk of the system was assessed to be 1.6E-7. Assuming the probability of failure as 1, the ecological risk was found to be 0.16 with 90% confidence interval of [3.1E-2, 7.7E-1]. The inclusion of the probability number in the risk quotient made the analysis more practicable. The risk of the analysis is measured in terms of the risk quotient. The risk quotient acts as a binary function, if it is less than 1, then there is no potential risk to the system and for greater than 1, there is a potential risk to the system. For a risk quotient, 1.6E-7, the risk to the system is almost negligible, however, it depends on a lot of factors as depicted by the analysis, including, the type and components of the storage system and the failure rates as perceived. It should be noted that the hypothetical case study results were generated for spill volume of 1000L, for the assumed quality of flowback water, for a distance of 200m between AWSS and creek, given bulk density of the soil, assumed area of soil affected and creek discharge. The risk quotient rapidly changes with the change in the value of these parameters. 4.6 Regulations Regulatory review was carried out on for storage systems on a national level and for the province of BC in Chapter 2. The results of the assessment provide a scientific basis for proposing regulations specific to the storage systems for waste water generated from the hydraulic fracturing process. The US is the most experienced in hydraulic fracturing and has a large-scale industry, hence, best voluntary practices followed in the US were reviewed. As per the results of FMEA, the failure modes having rank 1,4,7 and 5 are related to liner failure and are regulated under Drilling and Production regulation (DPR) and Hazardous waste regulations. Proper control of these regulations would help to prevent the failure of most storage systems. The presence of the secondary containment system brought down the probability of failure from 3E-3 to 1E-06. However, according to the DPR, the secondary containment system is needed only if the structure is in use for more than one year. A stricter regulation for making the secondary storage systems mandatory would help if the primary storage structure integrity is compromised. The best practices suggest the inclusion of the precipitation factor in the design of open tanks and to consider storm run off for erosion safety of the storage containers. The failure mode due to flooding has high severity but low failure rate. However, for open top storage system regulating 97 the inclusion of precipitation factor in the design of the AWSS would be beneficial. In BC,  the quality of the waste water stored is not regulated. ERA proves the importance of type and concentration of chemicals in the flowback water and how they affect the risk quotient. Hence, the concentration and/or the quality of flowback water stored must be regulated.  Regulations also provide a restriction on the citing of the storage structure. DPR prohibits construction of pit or AWSS  within 100m of the natural boundary of a water body. Environmental Protection and Management Regulation prohibit the citing of the storage system on top of the water supply well or the groundwater capture zone identified, groundwater recharge area, watershed, or identified aquifer. ERA results prove that the there is an indirect relation between the risk and distance between receiving water body and storage system. From the analysis, a distance of 200 m can cause risk to the aquatic ecology for spills greater than 10 m3 under considered conditions and assumptions. Also, the allowable volume of water stored is 6600 m3. In an event of catastrophic occurrence, from the above assessment, the potential adverse effects to the aquatic ecology could be huge.  Hence, the volume and quality of water stored should be regulated with respect to the distance of AWSS from the surface water body.   98 CHAPTER 5. CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusions Through the proposed method, the consequence of the failure of the flowback water storage system on the aquatic ecology was assessed. The probability of failure of above ground walled storage system (AWSS) was calculated. The critical failure modes of AWSS were identified and ranked. The overall risk to the ecosystem was estimated and the results of the study were assessed in light of the current regulations. A new risk assessment method for flowback water integrating BIA and ERA frameworks was developed. BIA framework combined FMEA and FTA techniques. The probability of the failure of storage system was assessed using FTA of BIA. Basic controllable events leading to the system failure were identified logically identified though fault tree. These basic events were considered inputs as failure modes in FMEA. FMEA, as used in BIA, ranked the failure modes to help prioritize corrective actions. The prioritization achieved through FMEA help in decision making for improving the over-all integrity of the storage system. To access ecological consequence of the failure of AWSS, exposure assessment, and effect assessment was carried out. A new technique to assess contaminant exposure for flowback water constituents was proposed. A combination of dilution factor and adsorption coefficient were used to assess exposure concentration. A simple equation reduced the complexity and multiplicity of the system. The probability and bioaccumulation were factored in to exposure concentration.  Toxicity values were derived for bromide, chloride, hydrogen sulfide, iron, manganese, and strontium using Canadian Water Quality Guidelines for Protection of Aquatic Life (CWQG-PAL). Estimating toxicity through SSD is very flexible and can be adjusted with ease to any desired safety level based on the management goals and requirements, toxicity modifying factors, background concentration levels of the water body and available data sets. The 5th 99 percentile of the cumulative empirical distribution curves was considered as the effect concentration benchmark to save 95% of the ecology.  Risk quotient was calculated by taking a ratio of exposure concentration and effect concentration. Uncertainty and sensitivity analysis was assessed assuming the probability value to be one, for the range of derived effect concentration and available initial exposure concentration from data set of 212 wells using Monte Carlo simulation. Scenario analysis for monthly creek discharges and historical spill volume ranges was carried out and a ratio of the volume of the spill to creek discharge was calculated to 4.5 for the safety of the ecology from flowback water and an equation dictating the relationship between them was derived. The proposed method will be helpful in assessing the ecological risk posed by the spill of flowback water collected from hydraulic fracturing using well-established and site-specific models. The framework is flexible to be modified for a given quality of flowback or produced water. Depending upon the physicochemical properties of the flowback water constituents, more interactions like vaporization or sedimentation can be added to the exposure assessment.   The proposed risk assessment method can be effectively applied for assessing the ecological risk on the surface water body posed by the hydraulic fracturing flowback water contamination when limited information is available about the transport behaviour and fate of the contaminants in the flowback water. A single equation reduces the complexity and multiplicity of the number of contaminants and their interaction.  The results obtained by the application of framework provided valuable information, such as the safe distance between flowback water storage and surface water body, the effect of soil quality in retarding the flow of contaminants, and the effect of seasonal water discharge and the spill severity on the risk potential. Such information will facilitate decision-making for the construction of hydraulic fracturing flowback water storage facilities.  A review of the regulations, carried out in the previous chapter suggests stricter regulations pertaining to the citing of the storage system, the distance of the storage system versus the 100 volume and quality of flowback water stored, secondary containment system for temporary storage and application of more than one liner or more frequent monitoring of the liner system.  5.2 Limitations and assumptions This analysis is flexible and incorporates parameters important for decision making process related to the flowback water storage systems. However, the analysis has limitations. The scope of the study was limited to the aquatic ecology and only the inorganic constituents of the flowback water were assessed. The BIA and ERA framework are both limited by the data available. Only single input values from one source are used to evaluate the probability of failure of the storage system, there is a high input data uncertainty. The exposure model is true only for a system in equilibrium and does not consider all the behavioral aspects of metals and salts like speciation, precipitation, colloid matter, interaction with organic matter, bioaccumulation, bioavailability and sediment deposition. Salt ions and their interactions with the background water chemicals are not assessed. Homogenous and continues mixing is assumed for dilution and distribution coefficient. The analysis assumes the additive risk for the contaminant and does not take into consideration synergistic or antagonistic effects of the combined mixture.  5.3 Research Contributions This is the first study that applies Ecological Risk Assessment framework for flowback water storage system. A comprehensive assessment of the failure system and its consequence on the aquatic ecology is done. It is a unique study investigating AWSS or C-ring type of storage system failure. Flowback water quality for Montney unconventional play is studied and an exposure model was proposed for the inorganic constituents like salts and metals of the flowback water. The threshold values for freshwater toxicity are derived for chloride, bromide, hydrogen sulfide, iron, manganese, sulfate, and strontium. The acute toxicity values of aquatic organisms do not exist at a provincial or national level for most of these contaminants. The framework and its applications also highlight the existing data gaps. With the rise in the unconventional oil and gas industry, the developed methodology would prove to be useful in evaluating the risk of the accidental releases of flowback water and making informed decisions regarding storage systems. The review of acts, regulations, and guidelines for storage systems in British Columbia is 101 undertaken and recommendation for regulations and guidelines specific to the flowback water storage systems are provided. The risk assessment study provided a scientific basis for proposing regulations specific to AWSS and paved a way for risk based regulations. 5.4 Recommended future work Recommendations are valuable for establishing the directions for future research. The recommendations are described based on the limitations of this assessment. A more comprehensive multi-media exposure model is desired that includes soil, groundwater, air and sediment media. The exposure model that could incorporate the characteristics of organic and inorganic constituents of the flowback water should be devised. Lack of reliable data for inorganic constituents like potassium, barium, sodium was being a hindrance in assessing the overall flowback water toxicity. Primary data should be generated through field study to determine the site-specific effect concentration of the native aquatic ecology infested with flowback water.  Also, lack of understanding of interactions of these ions among themselves and with the media stalled a holistic risk assessment. A field based exposure concentration values must be collected for the assessment. A reliable method for obtaining failure rates of the system is desirable to acquire more dependable results for obtaining the probability of storage system failure. The framework is limited to the ranking of the failure modes and does not focus on proposing the corrective actions and measuring their effectiveness when implemented. A comprehensive assessment of the storage system that investigates the failure modes, proposes corrective actions and assesses the impact and cost effectiveness of the corrective action is recommended.      102 REFERENCES  Adelman Jr., I. R. (1970). Effect of Hydrogen Sulfide on Northern Pike Eggs and Sac Fry. Transactions of the American Fisheries Society, 99, 501–509. Adelman Jr., I. R. (1972). Toxicity of Hydrogen Sulfide to Goldfish (Carassius auratus) as Influenced by Temperature, Oxygen, and Bioassay Techniques. Journal of the Fisheries Research Board of Canada, 29(9), 1309–1317. AER. (2014). Alberta’s energy reserves 2013 and supply/demand outlook 2014-2023 (ST98-2014). Calgary, Alberta. Retrieved from http://www.aer.ca/data-and-publications/statistical-reports/st98 Al-Daham, N. K., & Bhatti, M. N. (1977). Salinity Tolerance of Gambusia affinis (Baird & Girard) and Heteropneustes fossilis (Bloch). Journal of Fisheries Biology, 11, 309–313. Alam, M. K., & Maughan, O. E. (1995). Acute Toxicity of Heavy Metals to Common Carp (Cyprinus carpio). J. Environ. Sci. Health. Part A, Environ. Sci. Eng. Toxic Hazard. Substance Control, 30(8), 1807–1816. Alley, B., Beebe, A., Rodgers Jr., J., & Castle, J. W. (2011). Chemical and physical characterization of produced waters from conventional and unconventional fossil fuel resources. Chemosphere, 85(1), 74–82. https://doi.org/10.1016/j.chemosphere.2011.05.043 ANZECC & ARMCANZ. (2000). Australian and New Zealand Guidelines for Fresh and Marine Water Quality. Book (Vol. 1). API. (2014). API recommended practice 583-Corrosion Under Insulation and Fireproofing. ARI. (2013). EIA/ARI World Shale Gas and Shale Oil Resource Assessment. Arlington, VA, USA. Retrieved from www.adv-res.com Atherton, F., & Macintosh, C. (2014). Report of the Nova Scotia Independent Review Panel on Hydraulic Fracturing. Atherton, W., Ash, J. W., & Alkhaddar, R. M. (2008). The modeling of spills from the catastrophic failure of above ground storage tanks and the develpment of mitigation. International Oil Spill Conference Proceedings, 2008(1), 949–956. https://doi.org/10.7901/2169-3358-2008-1-949 Aven, T. (2015). Risk analysis methods. In Risk Analysis: Assessing Uncertainties beyond Expected Values and Probabilities (Second, pp. 55–83). JohnWiley & Sons, Ltd. Barbot, E., Vidic, N. S., Gregory, K. B., & Vidic, R. D. (2013). Spatial and temporal correlation of water quality parameters of produced waters from Devonian-age shale following hydraulic fracturing. Environmental Science and Technology, 47(6), 2562–2569. https://doi.org/10.1021/es304638h Baudouin, M. F., & Scoppa, P. (1974). Acute toxicity of various metals to freshwater zooplankton. Bulletin of Environmental Contamination & Toxicology, 12(6), 745–751. https://doi.org/10.1007/BF01685925 BCMOE. (2001). Ambient Water Quality Guidelines for Manganese. Retrieved November 7, 2016, from http://www.env.gov.bc.ca/wat/wq/BCguidelines/manganese/manganese.html BCOGC. (2009). INFORMATION LETTER # OGC 09-07. Retrieved from http://tanzania.eregulations.org/Media/Editor_Repo/tanzania land act.pdf%5Cnpapers2://publication/uuid/59B81A13-1381-44BD-AA24-32324389F29E BCOGC. (2014). Water Use for Oil and Gas Activity 2014 Annual Report. BCOGC. (2015a). Management of Saline Fluids for Hydraulic Fracturing Guideline. 103 BCOGC. (2015b). Management of Saline Fluids for Hydraulic Fracturing Guideline. Victoria. Retrieved from https://www.bcogc.ca/node/12440/download BCOGC. (2015c). Water Management for Oil and Gas Activity 2015 Annual Report. BCOGC. (2015d). Water Use for Oil and Gas Activity 2014 Annual Report. Victoria. BCOGC. (2016). Water Portal. Retrieved December 6, 2016, from http://waterportal.geoweb.bcogc.ca/#13/56.4059/-124.4584 Becklumb, P. (2015). Shale Gas in Canada : Environmental Risks and Regulation. Ben-daya, M. (2009). Failure Mode and Effect Analysis. In A.-K. D. Ben-Daya M., Duffuaa S., Raouf A., Knezevic J. (Ed.), Handbook of Maintenance Management and Engineering (pp. 75–90). London. https://doi.org/10.1007/978-1-84882-472-0_4 Bergmann, A., Weber, F.-A., Meiners, H., & Müller, F. (2014). Potential water-related environmental risks of hydraulic fracturing employed in exploration and exploitation of unconventional natural gas reservoirs in Germany. Environmental Sciences Europe, 26(1), 10. https://doi.org/10.1186/2190-4715-26-10 Biesinger, K., & Christensen, G. (1972). Effects of various metals on survival, growth, reproduction, and metabolism of Daphnia magna. Journal of the Fisheries Board of Canada, 29(12), 1691–1700. Birge, W. J., Black, J. A., Westerman, A. G., Short, T. M., Taylor, S. B., Bruser, D. M., & Wallingford, E. D. (1985). Recommendations on numerical values for regulating iron and chloride concentrations for the purpose of protecting warm water species of aquatic life in the Commonwealth of Kentucky. Memorandum of Agreement No. 5429. Kentucky Natural Resources and Environmental Protection Cabinet. Lexington, KY. Blewett, T. A., Delompré, P. L. M., He, Y., Folkerts, E. J., Flynn, S. L., Alessi, D. S., & Goss, G. G. (2017). Sublethal and Reproductive Effects of Acute and Chronic Exposure to Flowback and Produced Water from Hydraulic Fracturing on the Water Flea Daphnia magna. Environmental Science & Technology, 51(5), 3032–3039. https://doi.org/10.1021/acs.est.6b05179 Bonn, E. W., & Follis, B. J. (1967). Effects of Hydrogen Sulfide on Channel Catfish, Ictalurus punctatus. Transactions of the American Fisheries Society, 96, 31–36. Boutet, C., & Chaisemartin, C. (1973). Specific Toxic Properties of Metallic Salts in Austro-potamobius pallipes pallipes and Orconectes limosus. C.R.Soc.Biol., 167(12), 1933–1938. Bowmer, C. T., Hooftman, R. N., Hanstveit, A. O., Venderbosch, P. W. M., & Hoeven, N. Van der. (1998). The Ecotoxicity and the Biodegradability of Lactic Acid, Alkyl Lactate Esters and Lactate Salts. Chemosphere, 37(7), 1998. Bright, D. A., & Addison, J. (2002). Derivation of Matrix Soil Standards for Salt under the British Columbia Contaminated Sites Regulation. Victoria. Bringolf, R. B., Cope, W. G., Eads, C. B., Lazaro, P. R., Barnhart, M. C., & Shea., D. (2007). Acute and Chronic Toxicity of Technical grade Pesticides to Glochidia and Juveniles of Freshwater Mussels (Unionidae). Environmental Toxicology and Chemistry, 26, 2086–2093. Brittingham, M. C., Maloney, K. O., Farag, A. M., Harper, D. D., & Bowen, Z. H. (2014). Ecological Risks of Shale Oil and Gas Development to Wildlife, Aquatic Resources and their Habitats. Environmental Science & Technology, 48(19), 11034–11047. https://doi.org/10.1021/es5020482 Broderius, S. J., Smith, L. L. J., & Lind, D. T. (1977). Relative Toxicity of Free Cyanide and Dissolved Sulfide Forms to the Fathead Minnow (Pimephales promelas). Journal of the 104 Fisheries Research Board of Canada, 34(12), 2323–2332. Bunch, A. G., Perry, C. S., Abraham, L., Wikoff, D. S., Tachovsky, J. A., Hixon, J. G., … Haws, L. C. (2014). Evaluation of impact of shale gas operations in the Barnett Shale region on volatile organic compounds in air and potential human health risks. Science of the Total Environment, 468–469, 832–842. https://doi.org/10.1016/j.scitotenv.2013.08.080 Burton, G. A., Basu, N., Ellis, B. R., Kapo, K. E., Entrekin, S., & Nadelhoffer, K. (2014). Hydraulic “fracking”: Are surface water impacts an ecological concern? Environmental Toxicology and Chemistry, 33(8), 1679–1689. https://doi.org/10.1002/etc.2619 Calleja, M. C., Persoone, G., & Geladi, P. (1994). Comparative acute toxicity of the first 50 multicentre evaluation of in vitro cytotoxicity chemicals to aquatic non-vertebrates. Archives of Environmental Contamination and Toxicology, 26, 69–78. CAPP. (2017). Canadian Economic Contribution. Retrieved June 6, 2017, from http://www.capp.ca/canadian-oil-and-natural-gas/canadian-economic-contribution Carlson, C. (2012). Effective FMEAs: Achieving safe, reliable, and economical products and processes using failure mode and effects analysis (Vol I). John Wiley & Sons. CCA. (2014). Environmental Impacts of Shale Gas Extraction in Canada. CCA Trade Contractors Council. (2011). A trade contractor ’s guide and checklist to construction contracts. CCME. (1997). A Framework for Ecological Risk Assessment: Technical Appendices. Winnipeg, Manitoba. CCME. (2007). A protocol for the derivation of water quality guidelines for the protection of aquatic life. Canadian Environmental Quality Guidelines. Retrieved from www.ccme.ca/files/Resources/supporting_scientific.../protocol_aql_2007e.pdf CCME. (2011). Canadian Water Quality Guidelines for the Protection of Aquatic Life: Chloride. CCME. (2012). Canadian Water Quality Guidelines for the Protection of Aquatic Life NITRATE ION. Nano, (3), 1–7. CCPS. (2000a). Appendix D: Minimal Cut Set Analysis. In Guidelines for Chemical Process Quantitative Risk Analysis (Second, pp. 661–670). Hoboken, New Jersey: John Wiley & Sons, Inc. https://doi.org/10.1128/AAC.03728-14 CCPS. (2000b). Event Probability and Failure Frequency Analysis. In Guidelines for Chemical Process Quantitative Risk Analysis (Second, pp. 297–393). Hoboken, New Jersey: John Wiley & Sons, Inc. https://doi.org/10.1017/CBO9781107415324.004 CCREM. (2008). Canadian Water Quality Guidlines. Ottawa, Ontario. CEAA. (2012). Archived - Hydrometric Clean-up 07FB009 Flatbed Creek at km 110 Heritage Highway. Retrieved August 23, 2017, from http://www.ceaa-acee.gc.ca/052/details-eng.cfm?pid=31819#desc Chang, C., Liu, P., & Wei, C. (2001). Failure mode and effects analysis using grey theory. Integrated Manufacturing Systems, 12(3), 211–216. https://doi.org/10.1108/09576060110391174 Chen, S. S., Sun, Y., Tsang, D. C. W., Graham, N. J. D., Sik, Y., Feng, Y., & Li, X. (2017). Science of the Total Environment Potential impact of fl owback water from hydraulic fracturing on agricultural soil quality : Metal / metalloid bioaccessibility , Microtox bioassay , and enzyme activities. Science of the Total Environment, The, 579, 1419–1426. https://doi.org/10.1016/j.scitotenv.2016.11.141 Cheng, J. J., Polak, M., & Penlidis, A. (2009). Polymer Network Mobility and Environmental Stress Cracking Resistance of High Density Polyethylene. Polymer-Plastics Technology 105 and Engineering, 48(12), 1252–1261. https://doi.org/10.1080/03602550903159085 Cherry, J., Ben-Ali, M., Bharadwaj, L., Chalaturnyk, R., Dusseault, M. B., Goldstein, B., … Molson, J. (2014). Environmental Impacts of Shale Gas Extraction in Canada. Ottawa. Chiozza, M. L., & Ponzetti, C. (2009). FMEA: A model for reducing medical errors. Clinica Chimica Acta, 404, 75–78. https://doi.org/10.1016/j.cca.2009.03.015 Choi, I.-H., & Chang, D. (2016). Reliability and availability assessment of seabed storage tanks using fault tree analysis. Ocean Engineering, 120, 1–14. https://doi.org/10.1016/j.oceaneng.2016.04.021 Chong, J., & Simikian, M. (2014). Shale Gas in Canada: Resource Potential, Current Production and Economic Implications. Clark, J. S., Brydon, J. E., & Hortie, H. J. (1962). The clay minerals in some British Columbia Subsoils. Canadian Journal of Soil Science, 42, 296–301. Retrieved from http://www.curvesafrica.com/media/pdf/30MinutesTo.pdf Clemens, H. P., & Sneed, K. E. (1959). Lethal Doses of Several Commercial Chemicals for Fingerling Channel Catfish (Vol. 316). Retrieved from https://babel.hathitrust.org/cgi/pt?id=mdp.39015086537894;view=1up;seq=4 Collins, S. J., & Russell, R. W. (2009). Toxicity of road salt to Nova Scotia amphibians. Environmental Pollution, 157, 320–324. COSEWIC. (2016). Committee on status of wildlife in Canada. Retrieved December 7, 2016, from http://www.cosewic.gc.ca/eng/sct4/printresult_e.cfm?SSGBox=4 Cowgill, U. M., & Milazzo, D. P. (1990). The sensitivity of two cladocerans to water quality variables: salinity and hardness. Archives of Hydrobiology, 120, 185–196. Cunat, P. (2002). Corrosion Resistance of Stainless Steels in Soils and in Concrete. Plenary Days of the Committee on the Study of Pipe Corrossion and Protection, (October 2001), 1–12. Dalzell, D. J. B., & Macfarlane, N. a a. (1999). The toxicity of iron to brown trout and effects on the gills: a comparison of two grades of iron sulphate. Journal of Fish Biology, 55(2), 301–315. https://doi.org/10.1111/j.1095-8649.1999.tb00680.x De March, B. G. E. (1988). Acute Toxicity of Binary Mixtures of Five Cations (Cu2+, Cd2+, Zn2+, Mg2+, and K+) to the Freshwater Amphipod Gammarus lacustris (Sars): Alternative Descriptive Models. Can. J. Fish. Aquat. Sci, 45(4), 625–633. Diamond, M. L., Poulton, D. J., Mackay, D., & Stride, F. A. (1994). Development of a Mass Balance Model of the Fate of 17 Chemicals in the Bay of Quinte. Journal of Great Lakes Research, 20(4), 643–666. https://doi.org/10.1016/S0380-1330(94)71184-9 Duinker, P. N., & Greig, L. A. (2007). Scenario analysis in environmental impact assessment: Improving explorations of the future. Environmental Impact Assessment Review, 27(3), 206–219. https://doi.org/10.1016/j.eiar.2006.11.001 Duwe, B., Fuchs, B. D., & Hansen-Flaschen, J. (2005). Failure mode and effects analysis application to critical care medicine. Critical Care Clinics, 21(1), 21–30. https://doi.org/10.1016/j.ccc.2004.07.005 Elliott, E. G., Trinh, P., Ma, X., Leaderer, B. P., Ward, M. H., & Deziel, N. C. (2017). Unconventional oil and gas development and risk of childhood leukemia: Assessing the evidence. Science of the Total Environment, 576, 138–147. https://doi.org/10.1016/j.scitotenv.2016.10.072 Elphick, J. R. F., Bergh, K. D., & Bailey, H. C. (2011). Chronic toxicity of chloride to freshwater species: effects of hardness and implications for water quality guidelines. Environmental Toxicology and Chemistry, 30, 239–246. 106 Ericson II, C. A. (2005a). Failure mode and effects analysis. In Hazard Analysis Techniques for System Safety (pp. 235–259). Hoboken, New Jersey: John Wiley & Sons, Inc. https://doi.org/10.1108/02656719610118151 Ericson II, C. A. (2005b). Fault Tree Analysis. In Hazard Analysis Techniques for System Safety (pp. 183–221). Hoboken, New Jersey: John Wiley & Sons, Inc. https://doi.org/10.1016/B978-0-08-100889-8.00004-0 Ericson II, C. A. (2005c). Hazards, Mishap, and Risk. In Hazard Analysis Techniques for System Safety (pp. 13–29). Hoboken, New Jersey. Ericson II, C. A. (2005d). System Safety. In Hazard Analysis Techniques for System Safety (pp. 1–12). Hoboken, New Jersey.: John Wiley & Sons, Inc. https://doi.org/10.1007/978-3-319-18344-2_8 Ernst and Young. (2015). Review of British Columbia’s hydraulic fracturing regulatory framework. Ewen, C., Borchardt, D., Richter, S., & Hammerbacher, R. (2012). Hydrofracking Risk Assessment. Berlin. Ferrar, K. J., Kriesky, J., Christen, C. L., Marshall, L. P., Malone, S. L., Sharma, R. K., … Goldstein, B. D. (2013). Assessment and longitudinal analysis of health impacts and stressors perceived to result from unconventional shale gas development in the Marcellus Shale region. International Journal of Occupational and Environmental Health, 19(2), 104–112. https://doi.org/10.1179/2049396713Y.0000000024 Findeling, A., Ruy, S., & Scopel, E. (2003). Modeling the effects of a partial residue mulch on runoff using a physically based approach. Journal of Hydrology, 275, 49–66. https://doi.org/10.1016/S0022-1694(03)00021-0 Flury, M., & Papritz, A. (1993). Bromide in the natural environment: occurrence and toxicity. Journal of Environmental Quality, 22(4), 747–758. FracFocus. (2014). Glossary of Terms. Retrieved June 13, 2017, from http://fracfocus.ca/glossary of terms Fung, D. K., & Bewick, P. H. (1980). Short-Term Toxicity of Aqueous Hydrogen Sulfide to Representative Fish Species of Lake Huron. ASTM Spec. Tech. Publ., 3, 377–396. Gagnon, G. A., Krkosek, W., Anderson, L., Mcbean, E., Mohseni, M., Bazri, M., & Mauro, I. (2016). Impacts of hydraulic fracturing on water quality : a review of literature, regulatory frameworks and an analysis of information gaps. Environmental Review, 10(November 2015), 1–10. https://doi.org/dx.doi.org/10.1139/er-2015-0043 Garibay, R., & Hall, S. (2004). Chloride Threshold Recommendations for the Protection of Aquatic Life in theUpper Santa Clara River. Garvie, K. H., Lowe, L., & Shaw, K. (2012). SHALE GAS DEVELOPMENT IN FORT NELSON FIRST NATION TERRITORY: Potential Regional Impacts of the LNG Boom. BC Studies, (184). Gillis, P. L. (2011). Assessing the Toxicity of Sodium Chloride to the Glochidia of Freshwater Mussels: Implications for Salinization of Surface Waters. Environmental Pollution., 159, 1702–1708. Gosine, R., Dusseault, M., Gagnon, G., Keough, K., & Locke, W. (2016). Unconventional Opportunities & Challenges: Results of the Public Review of the Implications of Hydraulic Fracturing Operations in Western Newfoundland. Retrieved from http://nlhfrp.ca/final-report/ Gregory, K. B., Vidic, R. D., & Dzombak, D. A. (2011). Water management challenges 107 associated with the production of shale gas by hydraulic fracturing. Elements, 7(3), 181–186. https://doi.org/10.2113/gselements.7.3.181 Gustafsson, J. P. (2014). Visual MINTEQ 3.1 user guide. Retrieved from c:%5CProgram Files (x86)%5CVminteq31%5Cvminhelp.htm GWPC. (2016). How Casing Protects Groundwater. Retrieved March 15, 2016, from https://fracfocus.org/water-protection/casing-process Haluszczak, L. O., Rose, A. W., & Kump, L. R. (2013). Geochemical evaluation of flowback brine from Marcellus gas wells in Pennsylvania, USA. Applied Geochemistry, 28, 55–61. https://doi.org/10.1016/j.apgeochem.2012.10.002 Hammonds, J. S., Hoffman, F. O., & Bartell, S. M. (1994). An Introductory Guide to Uncertainty Analysis in Environmental and Health Risk Assessment. https://doi.org/10.2172/10127301 Health Canada. (2001). Total Dissolved Solids (Vol. 80225). Retrieved from https://www.canada.ca/content/dam/canada/health-canada/migration/healthy-canadians/publications/healthy-living-vie-saine/water-dissolved-solids-matieres-dissoutes-eau/alt/water-dissolved-solids-matieres-dissoutes-eau-eng.pdf Health Canada. (2017). Guidelines for Canadian Drinking Water Quality- Summary Table. Ottawa, Ontario. Retrieved from https://www.canada.ca/content/dam/hc-sc/migration/hc-sc/ewh-semt/alt_formats/pdf/pubs/water-eau/sum_guide-res_recom/sum_guide-res_recom-eng.pdf Hedayati, A., Hoseini, S. M., & Ghelichpour, M. (2014). Acute toxicity of waterborne manganese to Rutilus caspicus (Yakovlev, 1870) – gill histopathology, immune indices, oxidative condition, and saltwater. Toxicological & Environmental Chemistry, 96(10), 1–11. https://doi.org/10.1080/02772248.2015.1028408 Heffernan, K., & Dawson, F. M. (2010). An Overview of Canada’s Natural Gas Resources. Retrieved from http://www.csur.com/sites/default/files/Canada’s_Resource_Base_Overview.pdf Hinton, M. J., & Eversole, A. G. (1979). Toxicity of ten chemicals commonly used in aquaculture to the black eel stage of the American eel. Proceedings of the World Mariculture Society, 10, 554–560. Hoke, R. A., Gala, W. R., Drake, J. B., Giesy, J. P., & Flegler, S. (1992). Bicarbonate as a Potential Confounding Factor in Cladoceran Toxicity Assessments on Pore Water from Contaminated Sediments. Canadian Journal of Fisheries and Aquatic Sciences, 1633–1640. Hong, Z., & Binbin, L. (2009). Integrated analysis of software FMEA and FTA. Proceedings - 2009 International Conference on Information Technology and Computer Science, ITCS 2009, 2, 184–187. https://doi.org/10.1109/ITCS.2009.254 Hughes, J. S. (1973). Acute toxicity of thirty chemicals to striped bass (Morone saxatilis). Presented at the Western Association of State Game and Fish Commissioners, Salt Lake City, Utah , LA. HydroGeoLogic Inc, & Allison Geoscience Consultants Inc. (1998). MINTEQA2/PRODEFA2, A Geochemical Assessment Model for Environmental Systems: User Manual Supplement for Version 4.0, 1998(June 1998), 76. IHS Accumap. (2016). IHS Accumap Dataset. Retrieved from https://www.ihs.com/products/oil-gas-tools-accumap.html Ingraffea, A. R., Wells, M. T., Santoro, R. L., & Shonkoff, S. B. C. (2014). Assessment and risk analysis of casing and cement impairment in oil and gas wells in Pennsylvania, 2000-2012. Proceedings of the National Academy of Sciences, 111(30), 10955–10960. 108 https://doi.org/10.1073/pnas.1323422111 Ishio, S. (1965). Behavior of Fish Exposed to Toxic Substances. Jackman, P. (2010). Email to M. Nowierski December 9. Reference toxicant data for NaCl tested on various aquatic organisms. Jiang, M., Hendrickson, C. T., & Vanbriesen, J. M. (2014). Life cycle water consumption and wastewater generation impacts of a Marcellus shale gas well. Environmental Science and Technology, 48(3), 1911–1920. https://doi.org/10.1021/es4047654 Jiang, M., Michael Griffin, W., Hendrickson, C., Jaramillo, P., VanBriesen, J., & Venkatesh, A. (2011). Life cycle greenhouse gas emissions of Marcellus shale gas. Environmental Research Letters, 6, 34014. https://doi.org/10.1088/1748-9326/6/3/034014 Kargbo, D. M., Wilhelm, R. G., & Campbell, D. J. (2010). Natural gas plays in the Marcellus Shale: challenges and potential opportunities. Environmental Science & Technology, 44(15), 5679–84. https://doi.org/10.1021/es903811p Khangarot BS, R., & PK, A. (1989). Investigation of Correlation Between Physicochemical Properties of Metals and Their Toxicity to the Water Flea Daphnia magna Straus. Ecotoxicology and Environmental Safety, 18, 109–120. King, G. E. (2012). Hydraulic Fracturing 101: What Every Representative, Environmentalist, Regulator, Reporter, Investor, University Researcher, Neighbor and Engineer Should Know About Estimating Frac Risk and Improving Frac Performance in Unconventional Gas and Oil Wells. S. Proceedings of the SPE Hydraulic Fracturing Technology Conference, 80 pp. https://doi.org/10.2118/152596-MS Klimchuk, B. A., Macdonald, R., & Ferris, W. (2016). Performance Analysis of Engineered Liner Systems Used to Store Saline Fluids in the Canadian Oil and Gas Industry : Physical and Environmental Influences. Kondash, A., & Vengosh, A. (2015). Water Footprint of Hydraulic Fracturing. Environmental Science & Technology Letters, 150918090943001. https://doi.org/10.1021/acs.estlett.5b00211 Kuwayama, Y., Olmstead, S., & Krupnick, A. (2015). Water Quality and Quantity Impacts of Hydraulic Fracturing. Current Sustainable/Renewable Energy Reports, 2(1), 17–24. https://doi.org/10.1007/s40518-014-0023-4 Kuwayama, Y., Roeshot, S., Krupnick, A., Richardson, N., & Mares, J. (2015). Pits versus Tanks : Options for On-site from Shale Gas. Resources for the future. Washington, DC. Laverock, M. J., Stephenson, M., & Macdonald, C. R. (1995). Toxicity of iodine, iodide, and iodate to Daphnia magna and rainbow trout (Oncorhynchus mykiss). Archives of Environmental Contamination and Toxicology, 29(3), 344–350. https://doi.org/10.1007/BF00212499 LeBlanc, G. A. (1980). Acute Toxicity of Priority Pollutants to Water Flea (Daphnia magna). Bull. Environ. Contam. Toxicol, 24(5), 684–691. Lee, W. S. (1985). Fault Tree Analysis , Methods , and Applications - A Review, (3), 121–123. https://doi.org/10.1109/TR.1985.5222114 Leeuwen, C. J. van, & Hermens, J. L. M. (2004). Risk Assessment of Chemicals: An Introduction. Springer Science. Lester, Y., Ferrer, I., Thurman, E. M., Sitterley, K. A., Korak, J. A., Aiken, G., & Linden, K. G. (2015). Characterization of hydraulic fracturing flowback water in Colorado: Implications for water treatment. Science of the Total Environment, 512–513, 637–644. https://doi.org/10.1016/j.scitotenv.2015.01.043 109 Lutz, B. D., Lewis, A. N., & Doyle, M. W. (2013). Generation, transport, and disposal of wastewater associated with Marcellus Shale gas development. Water Resources Research, 49(2), 647–656. https://doi.org/10.1002/wrcr.20096 Mahmood, Y. A., Ahmadi, A., Verma, A. K., Srividya, A., & Kumar, U. (2013). Fuzzy fault tree analysis: a review of concept and application. International Journal of System Assurance Engineering and Management, 4(1), 19–32. https://doi.org/10.1007/s13198-013-0145-x McGlade, C., Speirs, J., & Sorrell, S. (2013). Unconventional gas - A review of regional and global resource estimates. Energy. https://doi.org/10.1016/j.energy.2013.01.048 McKenzie, L. M., Witter, R. Z., Newman, L. S., & Adgate, J. L. (2012). Human health risk assessment of air emissions from development of unconventional natural gas resources. Science of the Total Environment, 424, 79–87. https://doi.org/10.1016/j.scitotenv.2012.02.018 McKinley, P. (2015). The evolution and future of hydraulic fracturing policy in Ontario. Ryerson University. McPherson, C. A., Lawrence, G. S., Elphick, J. R., & Chapman, P. M. (2014). Development of a strontium chronic effects benchmark for aquatic life in freshwater. Environmental Toxicology and Chemistry, 33(11), 2472–2478. https://doi.org/10.1002/etc.2696 Meays, C., & Nordin, R. (2013). Ambient Water Quality Guidelines For Sulphate : technical appendix. Mount, D., Erickson, R., Highland, T., Hockett, R., Hoff, D., Jenson, C. T., … Wisniewski, S. (2016). The acute toxicity of major ion salts to Ceriodaphnia dubia : I. influence of background water chemistry. Environmental Toxicology and Chemistry, 35(12), 3039–3057. https://doi.org/10.1002/etc.3487 Mount, D., Gulley, D., Hockett, J., Garrison, T. D., & Evans, J. M. (1997). Statistical models to predict the toxicity of major ions to Ceriodaphia dubia, Daphnia magna, and Pimephales promelas (Fathead minnows). Environmental Toxicology and Chemistry, 16, 2009–2019. Mukhopadhyay, M. K., & Konar, S. K. (1984). Toxicity of copper, zinc and iron to fish, plankton and worm. Geobios, 11(5), 204–207. Murray, K. E. (2013). State-scale perspective on water use and production associated with oil and gas operations, Oklahoma, U.S. Environmental Science and Technology, 47(9), 4918–4925. https://doi.org/10.1021/es4000593 Nakashima, M. (2010). Design Recommendation for Storage Tanks and Their Supports. Nazir, M., Khan, F., Amyotte, P., & Sadiq, R. (2008). Subsea release of oil from a riser: An ecological risk assessment. Risk Analysis, 28(5), 1173–1196. https://doi.org/10.1111/j.1539-6924.2008.01136.x NEB. (2009). A Primer for Understanding Canadian Shale Gas - Energy Briefing Note. Production. Calgary. NEB. (2013). Filing requirements for onshore drilling operations involving hydraulic fracturing. NEB. (2015). Frequently Asked Questions: Filing Requirements for Onshore Drilling Operations Involving Hydraulic Fracturing. Retrieved March 29, 2016, from https://www.neb-one.gc.ca/bts/ctrg/gnthr/flrqnshrdrllprtn/flrqnshrdrllprtnfq-eng.html#s4 NEB. (2016a). Canada’ s Energy Future 2016: Energy supply and demand projections to 2040. NEB. (2016b). The Unconventional Gas Resources of Mississippian-Devonian Shales in the Liard Basin of British Columbia , the Northwest Territories , and Yukon Energy Briefing Note ● March 2016. Retrieved from http://www.neb-one.gc.ca/nrg/sttstc/ntrlgs/rprt/ltmtptntlbcnwtkn2016/ltmtptntlbcnwtkn2016-eng.pdf 110 NEB. (2016c). Who we are. Retrieved May 25, 2017, from https://www.neb-one.gc.ca/bts/whwr/index-eng.html NEB. (2017a). Energy Information Program - Glossary. Retrieved June 6, 2017, from https://www.neb-one.gc.ca/nrg/tl/glssr-eng.html NEB. (2017b). Market Snapshot: Montney gas wells increasingly productive due to technological improvements. Retrieved June 13, 2017, from https://www.neb-one.gc.ca/nrg/ntgrtd/mrkt/snpsht/2017/01-08mntngswlls-eng.html NEB. (2017c). Natural Gas. Retrieved June 6, 2017, from https://www.neb-one.gc.ca/nrg/sttstc/ntrlgs/index-eng.html Nicot, J.-P., & Scanlon, B. R. (2012). Water use for shale gas production in Texas, U.S. U.S. Environmental Science and Technology, 46, 3580–3586. https://doi.org/10.1021/es204602t Northrup, J. M., & Wittemyer, G. (2013). REVIEW AND Characterising the impacts of emerging energy development on wildlife , with an eye towards mitigation. Ecology Letters, (16), 112–125. https://doi.org/10.1111/ele.12009 NRCAN. (2015). Natural Gas. Retrieved January 1, 2017, from https://www.nrcan.gc.ca/energy/natural-gas/5639 NRCAN. (2016a). British Columbia’s Shale and Tight Resources. Retrieved May 21, 2016, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17692 NRCAN. (2016b). Exploration and Production of Shale and Tight Resources. Retrieved May 29, 2017, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17677 NRCAN. (2016c). Geology of Shale and Tight Resources. Retrieved June 12, 2017, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17675 NRCAN. (2016d). Manitoba’s Shale and Tight Resources. Retrieved May 25, 2017, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17696 NRCAN. (2016e). Newfoundland and Labrador’s Shale and Tight Resources. Retrieved May 25, 2017, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17700 NRCAN. (2016f). Ontario’s Shale and Tight Resources. Retrieved May 25, 2017, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17709 NRCAN. (2016g). Quebec’s Shale and Tight Resources. Retrieved May 25, 2017, from http://www.nrcan.gc.ca/energy/sources/shale-tight-resources/17714 Olawoyin, R., Wang, J. Y., & Oyewole, S. A. (2013). Environmental safety assessment of drilling operations in the Marcellus-shale gas development. SPE Drilling and Completion, 28(2), 212–220. Retrieved from http://www.scopus.com/inward/record.url?eid=2-s2.0-84883729864&partnerID=tZOtx3y1 Oracle. (2016). Crystal Ball Version 11.1.2.3. Retrieved from http://www.oracle.com/technetwork/middleware/crystalball/overview/index.html Oseid Jr., D. M. (1974). Factors Influencing Acute Toxicity Estimates of Hydrogen Sulfide to Freshwater Invertebrates. Water Res., 8(10), 739–746. Ostrom, L. T., & Wilhelmsen, C. A. (2012). Risk Assessment: Tools, Techniques, and Their Applications (First). Wiley Publishing, Inc. Oswald, R. E., & Bamberger, M. (2012). Impacts of Gas Drilling on human and animal Health. New Solutions, 14585(1), 1–4. Pacholski, L. J. (2009). Toxicity of Stable Strontium in Surface Freshwaters of the Oil Sands region in Alberta, Canada-Deriving an aquatic guidelines and refined chronic effects benchmark. University of Ulster. Palisade Corporation. (2010). @RISK for Excel, Risk Analysis Add-in for Microsoft Excel. 111 Palisade Corporation. (2015a). @RISK: Risk analysis and simulation add-in for Microsoft® Excel- User’s guide. Ithaca, NY. Palisade Corporation. (2015b). @RISK version 7. Retrieved November 1, 2016, from http://www.palisade.com/risk/ Palmer, C. G. (2004). The Development of a Toxicity Database Using Freshwater Macroinvertebrates, and Its Application to the Protection of South African Water Resources. South African Journal of Science, 100, 643–650. Patterson, L. A., Konschnik, K. E., Wiseman, H., Fargione, J., Maloney, K. O., Kiesecker, J., … Saiers, J. E. (2017). Unconventional Oil and Gas Spills: Risks, Mitigation Priorities, and State Reporting Requirements. Environmental Science and Technology, 51(5), 2563–2573. https://doi.org/10.1021/acs.est.6b05749 Pechan, P., Renn, O., Watt, A., & Pongratz, I. (2011). Safe or Not Safe: Deiciding What Risks to Accept in Our Environment and Food. New York: Springer Science & Business Media. https://doi.org/10.1007/978-1-4419-7868-4 Peggs, I. D. (2009). Geomembrane Liner Durability: Contributing Factors and the Status Quo. Ecotoxicology and Environmental Safety, 72(3), 838–842. Peredo-Alvarez, V. M., Sarma, S., & Nandini, S. (2003). Combined effect concentrations of algal food (Chlorella vulgaris) and salt (sodium chloride) on the population growth of Brachionus calyciflorus and Brachionus patulus (Rotifera). Revista de Biología Tropical, 51, 399–408. Phippen, B., Horvath, C., Nordin, R., & Nagpal, N. (2008a). Ambient Aquatic Life Guidelines for Iron (No. 9780772659903). Phippen, B., Horvath, C., Nordin, R., & Nagpal, N. (2008b). Ambient Aquatic Life Guidelines for Iron- Overview Report (No. 9780772659897). Pillay, A., & Wang, J. (2003). Modified failure mode and effects analysis using approximate reasoning. Reliability Engineering and System Safety, (79), 69–85. Rabinowitz, P. M., Slizovskiy, I. B., Lamers, V., Trufan, S. J., Holford, T. R., Dziura, J. D., … Stowe, M. H. (2015). Proximity to natural gas wells and reported health status: Results of a household survey in Washington County, Pennsylvania. Environmental Health Perspectives, 123(1), 21–26. https://doi.org/10.1289/ehp.1307732 [doi] Rausand, M. (2011a). Barriers and barrier analysis. In Risk Assessment Theory, Methods and Applications (Edition 1, pp. 363–407). Hoboken, New Jersey. Rausand, M. (2011b). Causal and frequency analysis. In Risk Assessment Theory, Methods and Applications (Edition 1, pp. 265–336). Hoboken, New Jersey. Rausand, M. (2011c). Hazard identification. In Risk Assessment Theory, Methods and Applications (Edition 1, pp. 213–263). Hoboken, New Jersey. Rausand, M. (2011d). Introduction to Risk Assessment. In Risk Assessment: Theory, Methods, and Applications (Edition 1, pp. 3–28). Hoboken, New Jersey. Rausand, M. (2011e). The words of risk analysis. In Risk Assessment Theory, Methods and Applications (Edition 1, pp. 29–63). Hoboken, New Jersey: John Wiley & Sons, Inc. Rawls, W. J., Brakensiek, D. L., Saxton, K. E., & Asae, M. (1982). Estimation of Soil Water Properties. Transactions of the ASAE, 25(5), 1316–1320. https://doi.org/10.1143/JJAP.46.5964 Reagan, M. T., Moridis, G. J., Keen, N. D., & Johnson, J. N. (2015). Numerical simulation of the environmental impact of hydraulic fracturing of tight/shale gas reservoirs on near-surface groundwater: Background, base cases, shallow reservoirs, short-term gas, and water 112 transport. Water Resources Research, 51(4), 2543–2573. https://doi.org/10.1002/2014WR016086 Refsgaard, J. C., van der Sluijs, J. P., Højberg, A. L., & Vanrolleghem, P. A. (2007). Uncertainty in the environmental modelling process – A framework and guidance. Environmental Modelling & Software, 22(11), 1543–1556. https://doi.org/10.1016/j.envsoft.2007.02.004 Regan, H., Akcakaya, H., Ferson, S., Root, K., Carroll, S., & Ginzburg, L. R. (2003). Treatments of uncertainty and variability in ecological risk assessment of single-species populations. Human and Ecological Risk Assessment, 9(4), 889–906. https://doi.org/10.1080/713610015 Reynolds, F. A., & Haines, T. A. (1980). Effects of Chronic Exposure to Hydrogen Sulphide on Newly Hatched Brown Trout Salmo trutta L. Environmental Pollution, 22, 11–17. Rivard, C., Lavoie, D., Lefebvre, R., Sejourne, S., Lamontagne, C., & Duchesne, M. (2014). An overview of Canadian shale gas production and environmental concerns. International Journal of Coal Geology, 126, 64–76. https://doi.org/10.1016/j.coal.2013.12.004 Robeson, L. M. (2013). Environmental stress cracking: A review. Polymer Engineering & Science, 53(3), 453–467. https://doi.org/10.1002/pen.23284 Rozell, D. J., & Reaven, S. J. (2012). Water Pollution Risk Associated with Natural Gas Extraction from the Marcellus Shale. Risk Analysis, 32(8), 1382–1393. https://doi.org/10.1111/j.1539-6924.2011.01757.x Rutqvist, J., Rinaldi, A. P., Cappa, F., & Moridis, G. J. (2013). Modeling of fault reactivation and induced seismicity during hydraulic fracturing of shale-gas reservoirs. Journal of Petroleum Science and Engineering, 107, 31–44. https://doi.org/10.1016/j.petrol.2013.04.023 Sadiq, R. (2001). Drilling waste discharges in the marine environment: A risk based decision methodology. Memorial University of Newfoundland. https://doi.org/10.1017/CBO9781107415324.004 Sanzo, D., & Hecnar, S. . (2006). Effect of road de-icing salt (NaCl) on larval wood frogs (Rana sylvatica). Environmental Pollution, 140, 247–256. Scanlon, B. R., Reedy, R. C., & Nicot, J.-P. (2014). Comparison of water use for hydraulic fracturing for unconventional oil and gas versus conventional oil. Environmental Science & Technology, 48(20), 12386–93. https://doi.org/10.1021/es502506v Shahriar, A., Sadiq, R., & Tesfamariam, S. (2012). Risk analysis for oil & gas pipelines: A sustainability assessment approach using fuzzy based bow-tie analysis. Journal of Loss Prevention in the Process Industries, 25(3), 505–523. https://doi.org/10.1016/j.jlp.2011.12.007 Shaw-Allen, P., & Suter II, G. W. (2016). CADDIS Volume 4: Data Analysis. Retrieved December 5, 2016, from https://www3.epa.gov/caddis/da_advanced_2.html Shuhaimi-Othman, M., Yakub, N., Ramle, N.-A., & Abas, A. (2013). Comparative toxicity of eight metals on freshwater fish. Toxicology and Industrial Health, 31(January), 1–10. https://doi.org/10.1177/0748233712472519 Siegel, L. (2007). Hazard Identification for Human and Ecological Effects of Sodium Chloride Road Salt. Slabbert, J. L., & Venter, E. A. (1999). Biological Assays for Aquatic Toxicity Testing. Water Science Technology, 39(10/11), 367–373. Smith, L. L. J., & Oseid, D. M. (1974). Effect of Hydrogen Sulfide on Development and Survival of Eight Freshwater Fish Species. In: J.H.S.Blaxter (Ed.), The Early Life History of Fish, Proceedings of an International Symposium, Oban, Scotland, 417–430. 113 Smith, L. L., & Oseid, D. M. (1972). Effects of Hydrogen Sulfide on Fish Eggs and Fry. Water Res., 6(6), 71–720. Speight, J. G. (2013). Shale gas production processes. Elsevier Science. Stammer, H. A. (1953). The Effect of Hydrogen Sulfide and Ammonia on Characteristic Animal Forms in Saprobiotic System (Der Einfly von Schwefelwasserstoff und Ammoniak auf tierische Leitformen des Sparobiensystems). Vom Wasser, 20, 34–71. Stephan, C. E. (1978a). Chronic Screening Toxicity Test with Daphnia magna. Stephan, C. E. (1978b). Results of Toxicity Tests. Suter II, G. W. (1996). Guide for Developing Conceptual Models for Ecological Risk Assessments. Suter II, G. W. (2007). Ecological Risk Assessment. (G. W. Suter II, Ed.) (Second). Boca Raton: CRC Press. Tatara, C. P., Newman, M. C., McCloskey, J. T., & Williams, P. L. (1998). Use of ion characteristics to predict relative toxicity of mono-, di-and trivalent metal ions: Caenorhabditis elegans LC50. Aquatic Toxicology, 42(4), 255–269. Thornton, K. W., & Sauer., J. R. (1972). Physiological effects of NaCl on Chironomus attenuatus (Diptera: Chironomidae). Annals of the Entomological Social of America, 65, 872–875. Tobias, D. A., & Kwon, J. (2016). E-FAST-Exposure and Fate Assessment Screening Tool Version 2014. Retrieved December 6, 2016, from https://www.epa.gov/tsca-screening-tools/e-fast-exposure-and-fate-assessment-screening-tool-version-2014 Torres, L., Yadav, O. P., & Khan, E. (2015). A review on risk assessment techniques for hydraulic fracturing water and produced water management implemented in onshore unconventional oil and gas production. Science of the Total Environment, 539, 478–493. https://doi.org/10.1016/j.scitotenv.2015.09.030 Trebuňa, F., Šimčák, F., & Bocko, J. (2009). Failure analysis of storage tank. Engineering Failure Analysis, 16(1), 26–38. https://doi.org/10.1016/j.engfailanal.2007.12.005 US EPA. (1985). Assessment of synthetic membrane successes and failures at waste storage and disposal sites (No. EPA/600/2-85/100). Cincinnati, Ohio. https://doi.org/EPA/600/2-85/100 US EPA. (1986a). Hazardous Waste Tank Failure Model: Description of Methodology, Draft report. US EPA. (1986b). Probability of Failure of Hazardous Waste Tanks, Draft Appendices. US EPA. (1986c). Quality Criteria for Water 1986 (No. EPA/440/5-86-001). Washington. US EPA. (1988). Superfund Exposure Assessment Manual (No. EPA/540/1-881001). Washington, DC. US EPA. (1996). EPA Liner Study: Report to Congress- Section 4113(a) of the Oil Pollution Act of 1990 (No. EPA 540/R95/041). https://doi.org/OSWER 9380.0-24 US EPA. (1997). Guiding Principles for Monte Carlo Analysis (No. EPA/630/R-97/001). Washington, DC. https://doi.org/EPA/630/R-97/001 US EPA. (1998). Guidelines for Ecological Risk Assessment (No. EPA/630/R-95/002F). Washington, DC. https://doi.org/10.1897/1552-8618(1992)11[1663:AFFERA]2.0.CO;2 US EPA. (1999). Understanding Variation in Partition Coefficient, Kd, Values (No. EPA 402-R-99-004A) (Vol. I). Washington, DC. US EPA. (2002). Catastrophic Tank Failures: Highlights of Past Failures along with Proactive Tanks Designs. US EPA. (2003). Generic Ecological Assessment Endpoints (GEAEs) for Ecological Risk Assessment (No. EPA/630/P-02/004F). Washington, DC. https://doi.org/EPA/630/P-114 02/004F US EPA. (2005). Human Health Risk Assessment Protocol for Hazardous Waste Combustion Facilities-Final (No. EPA530-R-05-006). US EPA. (2007). Framework for Metals Risk Assessment (No. EPA 120/R-07/001). Washington, DC. Retrieved from https://www.epa.gov/sites/production/files/2013-09/documents/metals-risk-assessment-final.pdf US EPA. (2009). Rupture Hazard from Liquid Storage Tanks (No. EPA 550-F-09-004). https://doi.org/EPA 550-F-09-004 US EPA. (2013). Hazardous Waste Tank Failure Model: Description of Methodology, Draft report. BiblioBazaar, 2013. US EPA. (2014). Compilation of Publicly Available Sources of Voluntary Management Practices for Oil and Gas Exploration & Production (E&P) Wastes As They Address Pits , Tanks , and Land Application. US EPA. (2015). Review of State and Industry Spill Data: Characterization of Hydaulic Fracturing-Related Spills (No. EPA/601/R-14/001). Washington, D.C. US EPA. (2016a). Ecological Risk Assessment - Glossary of Terms. Retrieved July 11, 2017, from https://archive.epa.gov/reg5sfun/ecology/web/html/glossary.html US EPA. (2016b). ECOTOX User Guide: ECOTOXicology Knowledgebase System. Version 4.0. US EPA. (2016c). Hydraulic Fracturing for Oil and Gas: Impacts from the Hydraulic Fracturing Water Cycle on Drinking Water Resources in the United States (No. EPA-600-R-16-236Fa). Washington, DC. Retrieved from https://cfpub.epa.gov/ncea/hfstudy/recordisplay.cfm?deid=332990 US EPA. (2016d). National Recommended Water Quality Criteria-Aquatic Life Criteria Table. Retrieved November 7, 2016, from https://www.epa.gov/wqc/national-recommended-water-quality-criteria-aquatic-life-criteria-table US EPA. (2016e). Species Sensitivity Distribution Generator. Retrieved November 1, 2016, from https://www3.epa.gov/caddis/da_software_ssdmacro.html US EPA. (2016f). Superfund Chemical Data Matrix (SCDM) Methodology (No. 189740). Washington, D.C. Retrieved from https://www.epa.gov/superfund/superfund-chemical-data-matrix-scdm US EPA. (2016g). Surface Water Models. Retrieved December 6, 2016, from https://www.epa.gov/exposure-assessment-models/surface-water-models US EPA. (2016h). Technical Overview of Ecological Risk Assessment: Risk Characterization. Retrieved July 11, 2017, from https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/technical-overview-ecological-risk-assessment-risk US NRC. (1981). Fault Tree Handbook (No. NUREG-0492). Washington, DC: U.S. Government Printing Office. USGS. (2015). Trends in Hydraulic Fracturing Distributions and Treatment Fluids, Additives, Proppants, and Water Volumes Applied to Wells Drilled in the United States from 1947 through 2010— Data Analysis and Comparison to the Literature. https://doi.org/10.3133/sir20145131 Valenti, T. W., Cherry, D. S., Neves, R. J., Locke, B. A., & Schmerfeld, J. J. (2007). Case Study: Sensitivity of Mussel Glochidia and Regulatory Test Organisms to Mercury and a Reference Toxicant. Freshwater Bivalve Ecotoxicology. SETAC Press, Pensacola, FL, USA, 351–367. 115 Van Dam, R. ., Hogan, A. C., McCullough, C. D., Houston, M. A., Humphrey, C. L., & Harford, A. J. (2010). Aquatic toxicity of magnesium sulfate, and the influence of calcium, in very low ionic concentration water. Environmental Toxicology and Chemistry, 29(2), 410–421. Van der Vaart, A. (2000). Asymptotic statistics Vol. 3. Cambridge university press. Vedamanikam, V. J., & Shazilli, N. A. M. (2008). The Effect of Multi-Generational Exposure to Metals and Resultant Change in Median Lethal Toxicity Tests Values over Subsequent Generations. Bull. Environ. Contam. Toxicol., 80(1), 63–67. Vengosh, A., Jackson, R. B., Warner, N., Darrah, T. H., & Kondash, A. (2014). A Critical Review of the Risks to Water Resources from Unconventional Shale Gas Development and Hydraulic Fracturing in the United States. Environmental Science & Technology, 48(15), 8334–8348. https://doi.org/10.1021/es405118y Vesely, B. (2002). Fault Tree Analysis (FTA): Concepts and Applications. Fault Tree Handbook with Aerospace Applications. Vidic, R. D., Brantley, S. L., Vandenbossche, J. M., Yoxtheimer, D., & Abad, J. D. (2013). Impact of Shale Gas Development on Regional Water Quality. Science, 340(6134), 1235009–1235009. https://doi.org/10.1126/science.1235009 Viertel, B. (1999). Salt tolerance of Rana temporaria: spawning site selection and survival during embryonic development (Amphibia, Anura). Amphibia-Reptilia, 20, 161–171. Vinot, H., & Larpent, J. P. (1984). Water Pollution by Uranium Ore Treatment Works. Hydrobiologia, 112, 125–129. Vosyliene, M. Z., Baltrenas, P., & Kazlauskiene, A. (2006). Toxicity of Road Maintenance Salts to Rainbow Trout Oncorhynchus mykiss. Ekologija, 2, 15–20. Wallen, I. E., Greer, W. C., & Lasater, R. (2017). Toxicity to Gambusia affinis of certain pure chemicals in turbid water. Water Environment Federation, 29(6), 695–711. Retrieved from http://www.jstor.org/stable/25033364 Wang, D., Zhang, P., & Chen, L. (2013). Fuzzy fault tree analysis for fi re and explosion of crude oil tanks. Journal of Loss Prevention in the Process Industries, 26(6), 1390–1398. https://doi.org/10.1016/j.jlp.2013.08.022 Wang, H., Ma, F., Tong, X., Liu, Z., Zhang, X., Wu, Z., … Yang, L. (2016). Assessment of global unconventional oil and gas resources. Petroleum Exploration and Development, 43(6), 925–940. https://doi.org/10.1016/S1876-3804(16)30111-2 Wang, N., & Ingersoll, C. J. (2010). Email to M. Nowierski December 20. Reference toxicity test data for NaCl and various aquatic invertebrates. Wang, W. (1986). Toxicity tests of aquatic pollutants by using common duckweed. Environmental Pollution. Series B, Chemical and Physical, 11(1), 1–14. https://doi.org/10.1016/0143-148X(86)90028-5 Werner, A. K., Vink, S., Watt, K., & Jagals, P. (2015). Environmental health impacts of unconventional natural gas development: A review of the current strength of evidence. Science of the Total Environment, 505(x), 1127–1141. https://doi.org/10.1016/j.scitotenv.2014.10.084 WHO. (2009). Bromide in Drinking-Water. World Health Organization. https://doi.org/05.03.2015 Williams, D. D., Williams, N. E., & Y, C. (1999). Road salt contamination of groundwater in a major metropolitan area and development of a biological index to monitor its impact. Water Research., 34, 127–138. Williams, P. L., & Dusenbery, D. B. (1990). Aquatic toxicity testing using the nematode, 116 Caenorhabditis elegans. Environmental Toxicology and Chemistry, 9(10), 1285–1290. Xu, J., Wang, H., Liu, Y., Ma, M., Zhang, T., Zheng, X., & Zong, M. (2016). Ecological risk assessment of heavy metals in soils surrounding oil waste disposal areas. Environmental Monitoring and Assessment, 188(2), 125. https://doi.org/10.1007/s10661-016-5093-x Ziemkiewicz, P. F., Quaranta, J. D., Darnell, A., & Wise, R. (2014). Exposure pathways related to shale gas development and procedures for reducing environmental and public risk. Journal of Natural Gas Science and Engineering, 16, 77–84. https://doi.org/10.1016/j.jngse.2013.11.003 Ziemkiewicz, P., Quaranta, J. D., & McCawley, M. (2014). Practical measures for reducing the risk of environmental contamination in shale energy production. Environmental Science: Processes & Impacts, 16(7), 1692–1699. https://doi.org/10.1039/c3em00510k     117 APPENDIX A: Dataset for deriving SSDs of the contaminants Table A-1 Data points for deriving toxicity profiles of the contaminants Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Iron Ictalurus punctatus Channel Catfish Fish 3 LC50 500000 500.00 (Clemens & Sneed, 1959; US EPA, 2016b) Iron Cyprinus carpio Common Carp Fish 4 LC50 1220 1.20 (Alam & Maughan, 1995; US EPA, 2016b) Iron Cyclops Viridis Water Flea Crustacean 4 LC50 11800 26.54 (Mukhopadhyay & Konar, 1984; Burke et al., 2008) Iron Oreochromis niloticus Mozambique tilapia Fish 4 LC50 83200 83.20 (Mukhopadhyay & Konar, 1984; Burke et al., 2008) Iron Oreochromis mossambicus Mozambique tilapia Fish 4 LC50 118000 118.80 (Mukhopadhyay & Konar, 1984; Burke et al., 2008) Iron Crangonyx pseudogracilis Northern River Crangonyctid Crustacean 2 LC50 143000 143.00 (Mukhopadhyay & Konar, 1984; Burke et al., 2008) Iron Asellus aquaticus Waterlouse Crustacean 2.08 LC50 255900 348.20 (Burke et al., 2008) Iron Branchiura Soerbyi Tubificid Worm Worm 4 LC50 580000 580.00 (Mukhopadhyay & Konar, 1984; Burke et al., 2008) Iron Poecilia reticulata Rainbow fish Fish 1 LC50 13940 4.89 (Shuhaimi-Othman et al., 2013) Iron Gambusia affinis Mosquitofish Fish 4 LC50 26000 26.00 (Shuhaimi-Othman et al., 2013; Wallen et al., 2017) Iron Morone saxatilis Striped Bass Fish 4 LC50 6000 6.00 (Hughes, 1973; Shuhaimi-Othman et al., 2013) Iron Salmo trutta Brown Trout Fish 4 LC50 47000 47.00 (Dalzell & Macfarlane, 1999; Shuhaimi-Othman et al., 2013) 118 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Magnesium Gammarus lacustris Scud Crustaceans 4 LC50 64700 64.70 (De March, 1988; US EPA, 2016b) Magnesium Mogurnda mogurnda Northern Trout Gudgeon Fish 4 LC50 40000 40.00 (Van Dam et al., 2010) Magnesium Daphnia magna Water Flea Crustaceans 2 LC50 140000 204.60 (Biesinger & Christensen, 1972) Manganese Craterocephalus marjoriae Black Banded Rainbow fish Fish 4 LC50 10200 10.20 (Shuhaimi-Othman et al., 2013) Manganese Poecilia reticulata Rainbow fish Fish 4 LC50 354000 83.83 (Slabbert & Venter, 1999; US EPA, 2016b) Manganese Oncorhynchus kisutch Coho Salmon  4 LC50 26900 26.90 (Hedayati et al., 2014; US EPA, 2016b) Manganese Daphnia magna Water Flea Crustaceans 2 LC50 29000 34.06 (Stephan, 1978a; US EPA, 2016b) Manganese Daphnia magna Water Flea Crustaceans 2 EC50  40000 (C.T., Hooftman et al., 1998; US EPA, 2016b) Manganese Daphnia pulex Water Flea Crustaceans 2 LC50 32050 32.05 (Slabbert & Venter, 1999; US EPA, 2016b) Manganese Pimephales promelas Fathead Minnow Fish 4 LC50 28000 28.00 (Stephan, 1978b; US EPA, 2016b) Manganese Lemna minor Duckweed Flowers, Trees, Shrubs, Ferns 4 EC50 (Growth) 31000 31.00 (US EPA, 2016b; W. Wang, 1986) Manganese Chironomus plumosus Midge Insects/Spiders 4 LC50 9500 6.77 (US EPA, 2016b; Vedamanikam & Shazilli, 2008) Manganese Culicoides furens Little Gray Punkie Insects/Spiders 4 LC50 6100 6.44 (US EPA, 2016b; Vedamanikam & Shazilli, 2008) Manganese Caspian Roach Rutilus Caspicus Fish 4 LC50 300000 300.00 (Hedayati et al., 2014; US EPA, 2016b) Manganese Rasbora sumatrana Cyprinidae Fish 1 LC50 120800 26.72 (Shuhaimi-Othman et al., 2013; US EPA, 2016b) Potassium Daphnia magna Waterflea Crustaceans 2 LC50 93000 114.41 (Biesinger & Christensen, 1972; US EPA, 2016b) 119 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Potassium Gammarus lacustris Scud Crustaceans 4 LC50 53200 53.20 (De March, 1988; US EPA, 2016b) Sodium Daphnia magna Water Flea Crustaceans 2 EC50 1820000 1675.01 (Biesinger & Christensen, 1972; US EPA, 2016b) Strontium Cyclops abyssorum  Freshwater Copepods Crustaceans 2 LC50 300000 300.00 (Baudouin & Scoppa, 1974; Pacholski, 2009) Strontium Eudiaptomus padanus  Worm 2 LC50 180000 180.00 (Baudouin & Scoppa, 1974; Pacholski, 2009) Strontium Daphnia hyalina Water Flea Crustaceans 2 LC50 75000 75.00 (Baudouin & Scoppa, 1974; Pacholski, 2009) Strontium Daphnia magna Water Flea Crustaceans 2 LC50 125000 91.11 (Biesinger & Christensen, 1972; Pacholski, 2009)  Strontium Daphnia magna Water Flea Crustaceans 2 EC50 94000 (Khangarot BS & PK, 1989; Pacholski, 2009) Strontium Daphnia magna Water Flea Crustaceans 1 EC50 162930 Strontium Austropotamobius pallipes pallipes White-clawed crayfish Crustaceans 4 LC50 440000 440.00 (Boutet & Chaisemartin, 1973; Pacholski, 2009) Strontium Oncorhynchus mykiss Rainbow Trout Fish (salmonid) 21 LC50 286000 286.00 (Pacholski, 2009) Strontium Morone saxatilis Striped Bass Fish (Non Salmonid) 4 LC50 92800 92.80 (Pacholski, 2009) Strontium Orconectes limosus Spinycheek crayfish Crustaceans 4 LC50 110000 110.00 (Boutet & Chaisemartin, 1973; Pacholski, 2009) Strontium Tubifex tubifex Tubificid worm Worm 1 EC50 540000 346.52 (Khangarot BS & PK, 1989; Pacholski, 2009) Strontium   Worm 1 LC50 10806000 1409.61 (Pacholski, 2009; P. L. Williams & Dusenbery, 1990) Barium Pseudokirchneriella subcapitata Green Algae Algae, Moss, Fungi 4 EC50  28000 28.00 (US EPA, 2016b) Barium Daphnia magna Water Flea Crustaceans 1 LC50 530000 315.85 (LeBlanc, 1980; US EPA, 2016b) Barium Crustaceans 2 LC50 145000 (Biesinger & Christensen, 1972) 120 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Barium Caenorhabditis elegans Round worm Worm 1 LC50 2.8 383.60 (Tatara et al., 1998; US EPA, 2016b) Barium Lepomis macrochirus Bluegill Fish 4 LC50 198000 198.00 (US EPA, 2016b) Barium Lemna minor Duckweed Flowers, Trees, Shrubs, Ferns 4 EC50 26000 26.00 (US EPA, 2016b; W. Wang, 1986) Iodide Daphnia magna Water Flea Crustaceans 2 LC50 830 0.39 (Laverock et al., 1995; US EPA, 2016b) Iodide Oncorhynchus mykiss Rainbow Trout Fish 4 LC50 17000 39.28 (Laverock et al., 1995; US EPA, 2016b) Bromide Oryzias latipes Japanese Rice Fish Fish 4 LC50 24000000 24000. (Flury & Papritz, 1993) Bromide Gammarufsa sciatus  Crustaceans 4 LC50 67000 67.00 (Flury & Papritz, 1993) Bromide Salmo gairdneri  Fish (salmonid) 5 LC50 2200000 2200.00 (Flury & Papritz, 1993) Bromide Dugesia tigrina  Invertebrate 4 LC50 67000 67.00 (Flury & Papritz, 1993) Bromide Poecilia reticulata Guppy Fish 4 LC50 16000000 16000.00 (Flury & Papritz, 1993) Bromide Daphnia magna Waterflea Crustaceans 1 LC50 7200000 8510.64 (Flury & Papritz, 1993) Bromide Pimephales promelas  Fathead Minnow Fish 1 LC50 14300000 13529.23 (Flury & Papritz, 1993) Hydrogen Sulphite Asellus militaris Aquatic Sowbug Crustaceans 4 EC50 1700 1.35 (Oseid Jr., 1974; US EPA, 2016b) Hydrogen Sulphite Carassius auratus Goldfish Fish 4 LC50 53 0.06 (Adelman Jr., 1972; US EPA, 2016b) Hydrogen Sulphite Carassius auratus Goldfish Fish 4 LC50 22 (L. L. Smith & Oseid, 1972; US EPA, 2016b) Hydrogen Sulphite Catostomus commersoni White Sucker Fish 4 LC50 23 0.02 (L. L. Smith & Oseid, 1972; US EPA, 2016b) Hydrogen Sulphite Chironomus riparius Midge Insects/Spiders 4 EC50 10000 22.36 (Stammer, 1953; US EPA, 2016b) 121 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Hydrogen Sulphite Crangonyx richmondensis ssp. Laurentianus Amphipod Crustaceans 2 EC50 770 0.50 (Oseid Jr., 1974; US EPA, 2016b) Hydrogen Sulphite Dendrocoelum lacteum Turbellarian, Planarian Worms 4 LC50 50000 50.00 (Stammer, 1953; US EPA, 2016b) Hydrogen Sulphite Ecdyonurus venosus Mayfly Insects/Spiders 4 EC50 50000 22.36 (Stammer, 1953; US EPA, 2016b) Hydrogen Sulphite Erpobdella octoculata Leech Worms 4 EC50 10000 22.36 (Stammer, 1953; US EPA, 2016b) Hydrogen Sulphite Esox lucius Northern Pike Fish 3 LC50* 36 0.04 (Adelman Jr., 1970; US EPA, 2016b) Hydrogen Sulphite Fish 3 LC50* 40 (L. L. Smith & Oseid, 1972; US EPA, 2016b) Hydrogen Sulphite Gammarus pseudolimnaeus Scud Crustaceans 2 EC50 71 0.04 (Oseid Jr., 1974; US EPA, 2016b) Hydrogen Sulphite Hexagenia limbata Mayfly Insects/Spiders 3 EC50 240 0.24 (Oseid Jr., 1974; US EPA, 2016b) Hydrogen Sulphite Ictalurus punctatus Channel Catfish Fish 0.0069 LC50 620 0.80 (Bonn & Follis, 1967; US EPA, 2016b) Hydrogen Sulphite Lepomis macrochirus Bluegill Fish 4 LC50 14 0.03 (L. L. Smith & Oseid, 1972; US EPA, 2016b) Hydrogen Sulphite Micropterus salmoides Largemouth Bass Fish 4 LC50 63 0.05 (Fung & Bewick, 1980; US EPA, 2016b) Hydrogen Sulphite Oncorhynchus mykiss Rainbow Trout Fish 4 LC50* 49 0.04 (L. L. Smith & Oseid, 1972; US EPA, 2016b) Hydrogen Sulphite Osteichthyes Bony Fishes Fish 0.3889 EC50 920.7 0.92 (Ishio, 1965; US EPA, 2016b) Hydrogen Sulphite Perca flavescens Yellow Perch Fish 4 LC50 8 0.03 (Fung & Bewick, 1980; US EPA, 2016b) Hydrogen Sulphite Perla sp. Stonefly Insects/Spiders 4 EC50 10000 22.36 (Stammer, 1953; US EPA, 2016b) Hydrogen Sulphite Pimephales promelas Fathead Minnow Fish 4 LC50 806 0.07 (Broderius et al., 1977; US EPA, 2016b) 122 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Hydrogen Sulphite Planaria gonocephala Planarian Worms 4 LC50 10000 22.36 (Stammer, 1953; US EPA, 2016b) Hydrogen Sulphite Salmo trutta Brown Trout Fish 4 LC50 7 0.01 (Reynolds & Haines, 1980; US EPA, 2016b) Hydrogen Sulphite Salvelinus fontinalis Brook Trout Fish 4 LC50 54 0.03 (L. L. J. Smith & Oseid, 1974; US EPA, 2016b) Hydrogen Sulphite Sander vitreus Walleye Fish 3 LC50* 60 0.03 (L. L. Smith & Oseid, 1972; US EPA, 2016b) Hydrogen Sulphite Stylaria lacustris Oligochaete Worms 4 EC50 10000 22.36 (Stammer, 1953; US EPA, 2016b) Hydrogen Sulphite Tubifex tubifex Tubificid Worm Worms 4 EC50 50000 22.36 (Stammer, 1953; US EPA, 2016b) Chloride Scenedesmus subspicatus Green Algae Algae, Moss, Fungi 5 LC50 4200000 4200.00 (US EPA, 2016b; Vinot & Larpent, 1984) Chloride Daphnia magna Water Flea Crustaceans 1 LC50 2600000 2600.00 (US EPA, 2016b; Vinot & Larpent, 1984) Chloride Danio rerio Zebra Danio Fish 1 LC50 6700000 6700.00 (US EPA, 2016b; Vinot & Larpent, 1984) Chloride Pimephales promelas Fathead minnow Fish 4 LC50 4223000 4223.00 (D.R. Mount et al., 1997; US EPA, 2016b) Chloride Lepomis macrochirus Bluegill sunfish Fish 4 LC50 5272000 5272.00 (Birge et al., 1985; CCME, 2011) Chloride  LC50 (CCME, 2011) Chloride Cyprinella leedsi Bannerfin shiner  Fish 4 LC50 6070000 6070.00 (CCME, 2011) Chloride Oncorhynchus mykiss Rainbow trout Fish 4 LC50 8634000 8634.00 (CCME, 2011; J.R.F., Bergh, & Bailey, 2011; Vosyliene et al., 2006) Chloride  LC50 Chloride Gambusia affinis Mosquito fish  Fish 4 LC50 9099000 9099.00 (Al-Daham & Bhatti, 1977; CCME, 2011) Chloride Gasterosteus aculeatus Threespine stickleback  Fish 4 LC50 10200000 10200.0 (CCME, 2011; Garibay & Hall, 2004) Chloride Anguilla rostrata American eel  Fish 4 LC50 13012000 13012.0 (CCME, 2011; Hinton & Eversole, 1979) 123 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Chloride Ambystoma maculatum Spotted salamander Amphibians 4 LC50 1178000 1178.00 (CCME, 2011; Collins & Russell, 2009) Chloride Pseudacris triseriata feriarum Chorus frog Amphibians 4 LC50 2320000 2320.00 (CCME, 2011; Garibay & Hall, 2004) Chloride Lithibates sylvatica Wood frog Amphibians 4 LC50 2716000 2716 (CCME, 2011; Collins & Russell, 2009; Jackman, 2010; Sanzo & Hecnar, 2006) Chloride LC50 Chloride LC50 Chloride Pseudacris crucifer Spring peeper  Amphibians 4 LC50 2830000 2830.00 (CCME, 2011; Collins & Russell, 2009) Chloride Rana clamitans Green frog  Amphibians 4 LC50 3109000 3109.00 (CCME, 2011; Collins & Russell, 2009) Chloride Rana temporaria Common frog Amphibians 4 LC50 3140000 3140.00 (CCME, 2011; Viertel, 1999) Chloride Lithibates pipiens Leopard frog Amphibians 4 LC50 3385000 3385.00 (CCME, 2011; Jackman, 2010) Chloride Bufo americanus American toad Amphibians 4 LC50 3926000 3926.00 (CCME, 2011; Collins & Russell, 2009) Chloride Rana catesbeiana Bullfrog  Amphibians 4 LC50 5846000 5846.00 (CCME, 2011) Chloride Sphaerium simile Fingernail clam  Invertebrate 4 LC50 902000 902.00 (CCME, 2011) Chloride Ceriodaphnia dubia Water flea  Invertebrate 2 LC50 1080000 1080 (CCME, 2011; Cowgill & Milazzo, 1990; Hoke et al., 1992; J.R.F. et al., 2011; D.R. Mount et al., 1997; Valenti et al., 2007) Chloride 2 LC50 Chloride 2 LC50 Chloride 2 LC50 Chloride 2 LC50 Chloride 2 LC50 Chloride Daphnia pulex Water flea Invertebrate 2 LC50 1248000 1248.00 (Birge et al., 1985; CCME, 2011; Palmer, 2004) Chloride 2 LC50 Chloride Elliptio lanceolata Yellow lance mussel  Invertebrate 4 LC50 1274000 1274.00 (CCME, 2011) Chloride Brachionus patulus Rotifer Invertebrate 1 LC50 1298000 1298.00 (CCME, 2011; Peredo-Alvarez et al., 2003) Chloride Hyalella azteca Amphipod Invertebrate 4 LC50 1382000 1382.00 (CCME, 2011; Elphick et al., 2011) 124 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Chloride Musculium transversum Fingernail clam Invertebrate 4 LC50 1930000 1930.00 (CCME, 2011) Chloride Brachionus calyciflorus Rotifer Invertebrate 1 LC50 2026000 2026.00 (Calleja et al., 1994; CCME, 2011; J.R.F. et al., 2011; Peredo-Alvarez et al., 2003) LC50 LC50 Chloride Physa gyrina Snail Invertebrate 4 LC50 2540000 2540.00 (Birge et al., 1985; CCME, 2011) Chloride Lirceus fontinalis Isopod Invertebrate 4 LC50 2950000 2950.00 (Birge et al., 1985; CCME, 2011) Chloride Gyraulus parvus Snail Invertebrate 4 LC50 3043000 3043.00 (CCME, 2011) Chloride Chironomus dilutus / tentans Midge Invertebrate 4 LC50 3761000 3761.00 (CCME, 2011; N. Wang & Ingersoll, 2010) Chloride Lumbriculus variegates Oligochaete Invertebrate 4 LC50 4094000 4094.00 (CCME, 2011; Elphick et al., 2011) Chloride LC50 Chloride Nephelopsis obscura Leech  Invertebrate 4 LC50 4310000 4310.00 (CCME, 2011) Chloride Hexagenia sp. Mayfly Invertebrate 2 LC50 4671000 4671.00 (CCME, 2011; N. Wang & Ingersoll, 2010) Chloride Chironomus attenatus Midge  Invertebrate 2 LC50 4850000 4850.00 (Thornton & Sauer., 1972; US EPA, 2016b) Chloride Daphnia hyalina Water flea Invertebrate 2 LC50 5308000 5308.00 (Baudouin & Scoppa, 1974; CCME, 2011) Chloride Lepidostoma sp Caddisfly Invertebrate 4 LC50 6000000 6000.00 (CCME, 2011; D. D. Williams et al., 1999) Chloride Tubifex tubifex Oligochaete  Invertebrate 4 LC50 6119000 6119.00 (CCME, 2011; Elphick et al., 2011; N. Wang & Ingersoll, 2010) Chloride 4 LC50 Chloride 4 LC50 Chloride Chironomus riparius Midge Invertebrate 2 LC50 6912000 6912.00 (CCME, 2011) Chloride Eudiaptomus padanus padanus Copepod Invertebrate 2 LC50 7077000 7077.00 (Baudouin & Scoppa, 1974; CCME, 2011) Chloride Cyclops abyssorum prealpinus Copepod Invertebrate 2 LC50 12385000 12385.00 (Baudouin & Scoppa, 1974; CCME, 2011) Chloride Epioblasma torulosa rangiana Northern riffleshell mussel Invertebrates 1 EC50 244000 244.00 (CCME, 2011; Gillis, 2011) Chloride Lampsilis siliquoidea Freshwater mussel Invertebrates 1 EC50 709000 709.00 (Bringolf et al., 2007; CCME, 2011) 125 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Chloride  Lampsilis fasciola Wavy-rayed lampmussel Invertebrates 1 EC50 746000.0 746.00 (Bringolf et al., 2007; CCME, 2011; Gillis, 2011; Valenti et al., 2007) 1 EC50 1 EC50 Chloride Lampsilis cardium Plain pocketbook Invertebrates 1 EC50 817000 817.00 (CCME, 2011; Gillis, 2011) Chloride Daphnia ambigua Water flea Invertebrates 2 EC50 1213000 1213.00 (CCME, 2011) Chloride Elliptio complanata Freshwater mussel Invertebrates 1 EC50 1620000 1620.00 (Bringolf et al., 2007; CCME, 2011) Chloride Epioblasma brevidens Cumberlandian combshell Invertebrates 1 EC50 1626000 1626.00 (CCME, 2011; Valenti et al., 2007) Chloride Epioblasma capsaeformis Oyster mussel Invertebrates 1 EC50 1644000 1644.00 (CCME, 2011; Valenti et al., 2007) Chloride Villosa constricta Freshwater mussel Invertebrates 1 EC50 1674000 1674.00 (Bringolf et al., 2007; CCME, 2011) Chloride Villosa iris Rainbow mussel Invertebrates 4 EC50 1815000 1815.00 (CCME, 2011) Chloride Villosa delumbis Freshwater mussel Invertebrates 1 EC50 2008000 2008.00 (Bringolf et al., 2007; CCME, 2011) Sulfate Scenedesmus subspicatus Green Algae Algae, Moss, Fungi 5 LC50 5600000 5600.00 (CCME, 2011; Vinot & Larpent, 1984) Sulfate Daphnia magna Water Flea Crustaceans 1 LC50 7000000 7000.00 (CCME, 2011; Vinot & Larpent, 1984) Sulfate Danio rerio Zebra Danio Fish 1 LC50 9600000 9600.00 (CCME, 2011; Vinot & Larpent, 1984) Sulfate Ceriodaphnia dubia Water Flea Invertebrate 2 LC50 2050000 2050.00 (Meays, C. & Nordin, 2013) Sulfate Chironomus tentans Midge Invertebrate 2 LC50 14134000 14134.0 (Meays, C. & Nordin, 2013) Sulfate Hyalella azteca Scud Crustaceans 4 LC50 512000 512.00 (Meays, C. & Nordin, 2013) Sulfate S. simile   4 LC50 2078000 2078.00 (Meays, C. & Nordin, 2013) Sulfate Lepomis macrochirus Blue gill Non-salmonid 4 LC50 13500000 13500.0 (Meays, C. & Nordin, 2013) Sulfate Pimephales promelas Fathead Minnow Non-salmonid 4 LC50 7960000 7960.00 (Meays, C. & Nordin, 2013) 126 Chemical Name Species Scientific Name Species Common Name Species Group Time (Days) End  point Conc (ug/L) Geo Mean (mg/l) References Sulfate Brachionus calyciflorus Rotifer  4 LC50 1701000 1701.00 (Meays, C. & Nordin, 2013) Sulfate Pseudacris regilla Pacific Tree Frog  4 LC50 1784500 1784.50 (Meays, C. & Nordin, 2013) Sulfate Raphidocelis subcapitata Microalga  4 LC50 1939000 1939.00 (Meays, C. & Nordin, 2013) Sulfate Oncorhynchus mykiss Rainbow Trout Fish (Salmonid) 4(assumed) LC50 889000 1137.39 (Meays, C. & Nordin, 2013) Sulfate Lemna minor Duckweed Flowers, Trees, Shrubs, Ferns 4 EC50 1000000 1000.00 (Meays, C. & Nordin, 2013)  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0357977/manifest

Comment

Related Items