Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Multi-hazard risk assessment : an interdependency approach Juarez Garcia, Hugon 2010

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

Item Metadata

Download

Media
24-ubc_2010_fall_juarezgarcia_hugon.pdf [ 5.69MB ]
Metadata
JSON: 24-1.0062902.json
JSON-LD: 24-1.0062902-ld.json
RDF/XML (Pretty): 24-1.0062902-rdf.xml
RDF/JSON: 24-1.0062902-rdf.json
Turtle: 24-1.0062902-turtle.txt
N-Triples: 24-1.0062902-rdf-ntriples.txt
Original Record: 24-1.0062902-source.json
Full Text
24-1.0062902-fulltext.txt
Citation
24-1.0062902.ris

Full Text

MULTI-HAZARD RISK ASSESSMENT: AN INTERDEPENDENCY APPROACH by HUGÓN JUÁREZ GARCÍA  B.S., Universidad Autónoma Metropolitana, 1987 M.S., Universidad Nacional Autónoma de México, 1991  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Civil Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  August 2010 © Hugón Juárez García, 2010  Abstract This research began with the Joint Infrastructure Interdependencies Research Program (JIIRP). JIIRP was part of an effort by the Government of Canada, to fund research to develop innovative ways to mitigate large disaster situations. An interdependency simulator was developed through the UBC-JIIRP project. This tool was developed to take into account the dynamic changes of the functional conditions of any given system. Interdependencies cannot be evaluated without assessing all the components of the system. This thesis makes two major contributions to the capability of the simulator, to handle seismic events and events that affect dense concentrations of people. The distinguishing characteristic of an earthquake event can affect the city and all the surrounding regions, causing damage to all lifeline systems, such as water, power, transportation and medical services. In its original form, I2Sim could model the damage and impact of each system on its own, but was unable to account for the effects of all other systems. The interdependency between systems is a crucial element for determining the impact of an earthquake and the time for recovery. The methodology proposed here can be used to measure Interdependencies and Resiliency in a region; and it was used in the simulator. Two cases were studied and implemented to test the methodology and the simulator. The first one was an earthquake hazard in a relative ―small‖ region (UBC Campus), this type of hazard will impose severe functional conditions to almost every infrastructure in a region, and the interdependencies would be revealed to the managers of the region; the second one, was a localized event in a massive sporting event (Winter Olympics in Vancouver), a black out in a Football Stadium that caused an uncontrolled egress, and related casualties due to a collapsing stage and the evacuation process was modelled. With the methodology and I2Sim it is possible to build up Region models, Disaster Scenarios, Objective Functions and Emergency Planning. The methodology, interdependency  ii  mapping, resiliency graphs and I2Sim will help in three phases of disaster: planning, response and recovery.  iii  Table of Contents Abstract ........................................................................................................................ ii Table of Contents ..........................................................................................................iv List of Tables ............................................................................................................... viii List of Figures ................................................................................................................ ix Acknowledgements ....................................................................................................... xi Dedication ................................................................................................................... xiii 1  2  Introduction .......................................................................................................... 1 1.1  Problem ................................................................................................................... 1  1.2  Purpose .................................................................................................................... 1  1.3  Motivation ............................................................................................................... 1  1.4  Introduction ............................................................................................................. 3  1.4.1  Functionality Definition ..................................................................................................... 4  1.4.2  Objectives......................................................................................................................... 5  1.5  Organization of the Thesis ....................................................................................... 5  1.6  Contributions to a Multidisciplinary Team ............................................................... 7  Literature Review ................................................................................................ 10 2.1  Introduction ........................................................................................................... 10  2.2  Natural and Man-Made Hazards ............................................................................ 11  2.2.1  Terrorist Attacks, September 11th, 2001 in the USA ......................................................... 11  2.2.2  The North-Eastern Blackout of 2003 in USA and Canada .................................................. 13  2.2.3  Hurricane Katrina, August 29th, 2003 in New Orleans....................................................... 14  2.2.4  Beginnings of Seismic Risk ............................................................................................... 15  2.2.5  The 1971 San Fernando Earthquake ................................................................................ 15  2.2.6  ATC-13 Earthquake Damage Evaluation Data for California .............................................. 15  2.2.7  The 1985 México City Earthquake ................................................................................... 16  2.2.8  The 1989 Loma Prieta Earthquake ................................................................................... 19  iv  2.2.9  Federal Emergency Management Agency in the US ......................................................... 20  2.2.10  The 1994 Northridge Earthquake .................................................................................... 21  2.2.11  The 1995 Kobe Earthquake ............................................................................................. 22  2.2.12  HAZUS ............................................................................................................................ 24  2.3  3  4  Interdependency Studies ....................................................................................... 24  UBC-JIIRP General Project ................................................................................... 37 3.1  Introduction ........................................................................................................... 37  3.2  Integration of Various Methodologies - Interdisciplinary Work ............................. 39  3.3  Motivation ............................................................................................................. 40  3.4  Ontology for an Infrastructure Interdependency Simulator (I2Sim) ....................... 41  3.4.1  Resources and Human Readable Tables (HRTs)................................................................ 43  3.4.2  Cell Model ...................................................................................................................... 45  3.4.3  Channel Model ............................................................................................................... 45  3.4.4  Scalability ....................................................................................................................... 46  3.4.5  System Model ................................................................................................................. 46  3.4.6  Hazards........................................................................................................................... 46  3.4.7  Events............................................................................................................................. 47  3.4.8  Scenarios (Simulation Cases) ........................................................................................... 47  3.4.9  Decisions ........................................................................................................................ 48  3.5  Interdependency among Critical Infrastructures .................................................... 49  3.6  I2Sim Capabilities ................................................................................................... 56  3.6.1  Damage to Stadia and Crowd Egress ............................................................................... 56  3.6.2  Traffic and Street Crowds ................................................................................................ 57  3.6.3  Damage to Physical Infrastructure................................................................................... 57  3.6.4  Damage to Human Infrastructure .................................................................................... 57  3.6.5  Egress and Traffic Models for I2Sim ................................................................................. 58  3.7  I2Sim Functionality................................................................................................. 60  3.8  Final Remarks on UBC-JIIRP.................................................................................... 60  Interdependencies ............................................................................................... 69 4.1  Introduction ........................................................................................................... 69  4.2  Interdependencies ................................................................................................. 70  v  5  4.2.1  Directionality and Connectivity for Lifeline Systems ......................................................... 70  4.2.2  Single and Global Interdependencies............................................................................... 72  4.2.3  Single Interdependency Functionality (SIF) ...................................................................... 73  4.2.4  Example on SIF of the Water System ............................................................................... 79  4.2.5  Global Interdependency Functionality (GIF)..................................................................... 87  4.2.6  Resources and Human Readable Tables (HRTs)................................................................ 91  4.2.7  Example of Resiliency and GIF of Region A12 ................................................................... 92  4.2.8  Interdependency Matrix ................................................................................................. 96  4.2.9  Interdependency Index and Importance Factors ............................................................ 101  UBC Test Case .....................................................................................................103 5.1 5.1.1  6  Problem Definition: UBC Test Case ...................................................................... 103 Description ................................................................................................................... 103  5.2  Methodology: UBC Test Case ............................................................................... 105  5.3  Damage Assessment: UBC Test Case .................................................................... 106  5.3.1  Identification of Critical Lifeline Systems ....................................................................... 107  5.3.2  Information (Required vs Outcomes)............................................................................. 108  5.3.3  Hazards......................................................................................................................... 110  5.3.4  Results .......................................................................................................................... 110  5.3.5  Human Functionality ..................................................................................................... 129  Downtown Vancouver Scenario .........................................................................134 6.1  Introduction ......................................................................................................... 134  6.2  Sample Complex Scenario .................................................................................... 134  6.2.1  Scenario Events............................................................................................................. 137  6.2.2  I2Sim-ETran Model ....................................................................................................... 139  6.3  Scenario Outcomes .............................................................................................. 142  6.4  Effectiveness of Response .................................................................................... 144  6.5  Simulation Results ............................................................................................... 145  6.5.1  Road Condition Results ................................................................................................. 154  6.5.2  Number of Casualties .................................................................................................... 156  6.5.3  High Casualty Case ........................................................................................................ 157  6.6  Implications of the Study ..................................................................................... 169  vi  7  Conclusions .........................................................................................................171 7.1  Future Work ......................................................................................................... 173  7.1.1  Disaster Response Network........................................................................................... 173  7.1.2  Design of Sustainable and Resilient Communities .......................................................... 175  7.1.3  Extended Objective Functions for an Infrastructure Interdependent Simulator (I2Sim) .. 176  7.2  Final Remarks ....................................................................................................... 177  References..................................................................................................................179  vii  List of Tables Table 3.1 Percentages and Colors for Physical and Resource Modes ..................................................................... 43 Table 3.2 HRTs for a Hospital ............................................................................................................................... 45 Table 4.1 General Characteristics of the Distribution Pipeline (PH to H) ................................................................ 81 Table 4.2 Functionality Conditions of Power House Intensities VI to IX .................................................................. 81 Table 4.3 Functionality Conditions of Power House, II X to XII ............................................................................... 81 Table 4.4 Losses in the Water Pipelines (Water Station to Hospital Distribution Lines) .......................................... 82 Table 4.5 Functionality Conditions at the Power House, after Equation 1 .............................................................. 83 Table 4.6 Hospital Performance Table .................................................................................................................. 91 Table 4.7 Power House Performance Table .......................................................................................................... 91 Table 4.8 Electrical Substation Performance Table ............................................................................................... 92 Table 4.9 Interdependency Matrix ..................................................................................................................... 100 Table 4.10 Interdependency Indices and Importance Factors .............................................................................. 101 Table 5.1 Comparison between Survival Tokens vs Critical Sectors, ..................................................................... 107 Table 5.2 General information used for Earthquake Damage Estimation ............................................................ 109 Table 5.3 Information Used and Outcomes for Earthquake Damage Estimation.................................................. 109 Table 5.4 Part I, Portion North-East ................................................................................................................... 122 Table 5.5 Repair Rates of Part I, Portion North-East, See Table 5.1 for Details..................................................... 123 Table 5.6 Accumulated and Segmented Loss in the Pipeline................................................................................ 123 Table 5.7 Population, Casualties and Level of Injuries after an Intensity IX Earthquake ....................................... 130 Table 5.8 Load Calculation of Patients and People to be Served by the Hospital after an Intensity IX Earthquake 132 Table 6.1 Variables for Object Function Simulations ........................................................................................... 147 Table 6.2 I2Sim Results for Road Conditions ....................................................................................................... 158 Table 6.3 Results from Road Conditions ............................................................................................................. 159 Table 6.4 Results from I2Sim Simulation with 30, 50 and 100 Casualties ............................................................. 160 Table 6.5 283 Yellow-Coded Casualties and 7 Hospitals ...................................................................................... 161 Table 6.6 864 Yellow-Coded Casualties and 7 Hospitals ...................................................................................... 162 Table 6.7 2,591 Yellow Coded Casualties and 7 Hospitals ................................................................................... 163  viii  List of Figures Figure 2.1 System Resiliency ................................................................................................................................ 31 Figure 3.1 Cell, Channels, Aggregators and Distributors ....................................................................................... 43 Figure 3.2 Cell with Physical and Resource Modes ................................................................................................ 44 Figure 3.3 Hazard and Scenario Cells and Event Channels ..................................................................................... 47 Figure 3.4 Activation of Decision Cells .................................................................................................................. 48 Figure 3.5 Components of the Architecture Project............................................................................................... 49 Figure 3.6 Data Sharing in the JIIRP-UBC Project .................................................................................................. 64 Figure 4.1 Assessment in Pipeline Segment P-100 ................................................................................................ 71 Figure 4.2 Methodology to Determine Single and Global Interdependencies ......................................................... 75 Figure 4.3 Water System of Region A ................................................................................................................... 75 Figure 4.4 Water System or Zone A12 .................................................................................................................. 76 Figure 4.5 Critical infrastructure System, Local ..................................................................................................... 77 Figure 4.6 Water system - Power House to Hospital ............................................................................................. 80 Figure 4.7 Functionality Conditions of the Power House ....................................................................................... 83 Figure 4.8 SIF of Water Distribution in Zone A12 .................................................................................................. 84 Figure 4.9 SIF of the Water System at Region A12 ................................................................................................ 85 Figure 4.10 SIF of the Electrical System in Region A12 .......................................................................................... 86 Figure 4.11 Global Interdependency Functionality Concept................................................................................... 87 Figure 4.12 GIF at Zone A12 ................................................................................................................................. 89 Figure 4.13 Overall System Resiliency................................................................................................................... 90 Figure 4.14 Resiliency of Region A and CIs Global Interdependencies .................................................................... 94 Figure 4.15 Region X with Three Assets: PH, H and SS, with Channels and Tokens ................................................. 97 Figure 4.16 Interdependency Matrix .................................................................................................................... 99 Figure 5.1 UBC Location. Source: Google Maps .................................................................................................. 104 Figure 5.2 Power House, Hospital and Main Substation for UBC Test Case .......................................................... 105 Figure 5.3 Region X, and Assets A, B and C ......................................................................................................... 108 Figure 5.4 Structural Damage with Modifiers for II IX ......................................................................................... 112 Figure 5.5 Displacement, Damage for II IX .......................................................................................................... 114 Figure 5.6 Building Occupancy on UBC Campus .................................................................................................. 115 Figure 5.7 Population on Campus at 2 pm .......................................................................................................... 117 Figure 5.8 Casualties for II IX on Campus at 2 pm ............................................................................................... 118 Figure 5.9 Campus Functionality for II IX ............................................................................................................ 119  ix  Figure 5.10 Water Main Pipelines Showing Low Direction. Yellow Lines Refer to Trunk Lines Providing Water from the Reservoir to the Water Station. Blue Lines and Red Lines Main Distribution Lines Providing Water All Across UBC. Source: Google Maps. ............................................................................................................................... 121 Figure 5.11 Trunk Line from the Reservoir to the Water Station. Source: Google Maps ....................................... 122 Figure 5.12 Distributed Damage in the Main Water System, without Interdependencies ..................................... 124 Figure 5.13 Accumulated Damage in the Main Water System, with Interdependencies ....................................... 125 Figure 5.14 Overlaid Damage Assessments, and Main Water System and Buildings Interdependencies ............... 127 Figure 5.15 Functionality Conditions of Roads. Interdependency with Building Functionality ............................... 129 Figure 5.16 Total Casualties ............................................................................................................................... 132 Figure 6.1 Geographic Area ............................................................................................................................... 136 Figure 6.2 Egress Model..................................................................................................................................... 139 Figure 6.3 I2Sim General Model ......................................................................................................................... 143 Figure 6.4 Traffic Zones Considered in the Model ............................................................................................... 144 Figure 6.5 Time for Patient Treatment following Collapse Stage at BC Place ....................................................... 149 Figure 6.6 Critical Time for Patient Treatment, Egress BC Place .......................................................................... 151 Figure 6.7 Roads Considered in the Simulation ................................................................................................... 154 Figure 6.8 Total Transportation Time and Road Conditions after Injury............................................................... 164 Figure 6.9 Assessment and Stabilization Time at ER after Injury ......................................................................... 164 Figure 6.10 Number of Dead Victims due to Delay in Treatment ......................................................................... 165 Figure 6.11 Number of Dead Victims due to Delay in Treatment ......................................................................... 165 Figure 6.12 Transportation Time and Number of Casualties ............................................................................... 166 Figure 6.13 Stabilization Time and Number of Casualties ................................................................................... 166 Figure 6.14 Time for Assessment and Stabilization at 7 Hospitals in Vancouver .................................................. 167 Figure 6.15 Time for Assessment and Stabilization Time, 283 Casualties ............................................................. 167 Figure 6.16 Time for Assessment and Stabilization, 864 Casualties ..................................................................... 168 Figure 6.17 Time for Assessment and Stabilization, 2,591 Casualties .................................................................. 168  x  Acknowledgements I would like to thank my supervisor Dr. Carlos E. Ventura, I deeply appreciate everything he taught me and how he respected my ideas and opinions during our discussions. He was the one who pushed me into vibration topics and laboratory experiences, to play soccer to relieve my stress and to see Vancouver‘s natural scenery as part of my learning experience. Special thanks go to my co-supervisor Dr. José R. Martí. He invited me in to JIIRP and the Olympic projects, and helped me to develop many ideas during the course of the projects. He showed me that interdependencies go well beyond research, and that we are interconnected in ways that we can model with I2Sim ontology. I would also like to thank the members of my committee. Dr. Liam Finn, his experience and advice were key factors in developing my research and the presentation of this manuscript. He taught me to appreciate where I come from, and his warmth heart and home were always opened for Elsa and me. Many thanks to Dr. K.D. Srivastava; I treasured all of the support, advice and opinions that he gave me during these projects. Special thanks to my Alma Mater: Universidad Autónoma Metropolitana, Unidad Azcapotzalco in México, who supported me financially. I really appreciate all the efforts by academics and staff that were there for me whenever I needed support, advice or had decisions to make. I realized that UAM-A does more than we imagine, and this experience in a foreign country has helped me put in perspective many aspects of this great Institution. Special thanks to Dr. Alonso Gómez Bernal who guided me, supported me, and encouraged me to pursue my dreams. He is more a friend than a colleague, but he has taught me many things, like enjoying a Temascal to help turn all vibrations positive. I would also like to thank my friend Dr. Emilio Sordo Zabay, I really appreciate his friendship, support and advice. I would also like to thank Hans Archundia and Eduardo Arellano Méndez. I am grateful of the friendship and support that Dr. Jorge Hollman gave me during the JIIRP project. In JIIRP, and later in the Olympic project I had the opportunity to work with three xi  ladies: Kate Thibert, Ruvini Kaskamange and Shahrzad Rostamirad. I can only say thank you for listening to me and believing in all my ―wondrous‖ stories. I also like to thank my friends at UBC; especially to Alfredo Bohl, Juan Carlos Carvajal Uribe, José Centeno, Freddy Piña, and Ottón Lara; we enjoyed joyful and sad rides along the way with guitars, wine, food and fun, interesting and intelligent sobremesas. Special thanks go to my family that I truly love, especially my wife Elsa whose constant encouragement was the driving hammer in the rock of nothingness. She put aside beautiful sunny summer days, magical travels, and trips of pleasure by being by my side along this research. She was the one who brought my feet back to the ground every night, or who helped me out of the mud I was stuck in. She was there with me, helping me, when she needed me the most. I thank my mom Consuelo and my dad Fernando for loving me and encouraging me to believe in me, for helping me being a better person, and for giving me the gift of life. To my brother Gilberto, for all that he taught me and he always had words of encouragement and support. To my brothers Alejandro and Arturo, and my sister Sadi who visited us in Vancouver; they supported us with messages, songs and words of faith and science. Lorena, Pakita and Ricardo, we know that you kept us in your warm hearts. Many thanks to César and Emma, we were fortunate enough to have you here with us in Vancouver, and I cannot thank you enough for letting Elsa be by my side. To Isabel, Samuel, Nadia and Irasema who visited us in moments when we needed support. Yair, Andrés, Diana and Yesenia who helped us in everything they could, whose eyes and hearts were always opened for us.  xii  Dedication  To my beloved wife Elsa:  …porque traías siempre la luna y los sueños de regreso en mi noche  …because you always brought the moon and dreams back into my night  xiii  1  Introduction  1.1 Problem A major task for Canada, México and other countries is to establish an action plan to guide the identification of risks, the implementation of protective safety measures, and the proper and effective response to disruptions of critical infrastructure systems. The basic problem is how to reduce the vulnerability of critical Infrastructure and how to strengthen its resiliency whenever natural or man-made hazards strike.  1.2 Purpose The purpose of this research is to develop a methodology that will identify the vulnerabilities of Critical Infrastructure Systems and increase their resiliency by identifying interdependencies among them. Seismic Risk Assessment will be used to develop this methodology. A region, city or venue is treated as study space. The study space is disassembled into a finite number of physical layers (critical infrastructures - CIs). Risks to individual systems will be estimated separately, and the effects of their interdependencies will be evaluated.  1.3 Motivation In recent disaster events, inadequate coordination among infrastructures and among hierarchical decision levels has been a major cause of failures in the response of the system and it has been responsible for unnecessary losses in human lives, property, and economic activity. The motivation of this project is clear. Whenever natural or man-made hazards strike, they provoke enormous consequences to different Critical Infrastructure Systems, and above all they endanger human lives, affecting even the basic needs of food and water. If the organizers of mass sporting events, like Olympic Games or World Cups, do not prepare for a surge of spectators to the venues or celebration sites, they may have problems in transportation, shelter, 1  food and water. We are living in a global world, terrorist threats or protests have become part of our world, and hence security measures have been enforced to diminish the effect of antagonistic groups. But organizers have failed to consider the consequences and interdependencies that some of the security measures will cause in the everyday life of a city and the well-being of the normal public and spectators of these mass spectator events. Major threats like hurricanes, earthquakes, floods or landslides impose great affliction, and most of the time, human lives are lost. Therefore it is important to prepare beforehand, to respond during and recover as quickly as possible after hazards have occurred. The way to reduce the consequences and challenges that regions will face is to begin with the following studies:   Multi-hazard. The collection of natural and man-made hazards that a region will face in a given span of time; some regions may face different challenges in different time spans    Vulnerability. This study will help to identify robust and vulnerable infrastructure. It is important to go farther than just the study of building infrastructure    Critical Infrastructure (CI) systems and Objective functions for a region. Light, moderate and great hazards will impose different demands on a region, and set different levels of importance for every CI system. Vulnerability studies and deciding which CI is important may look easy to accomplish, but the challenge is to acquire and process effectively relevant information    Interdependencies. How the CIs are interconnected, and how they depend on each other to sustain liveable conditions in a region is extremely important. Interdependencies will let the manager of the region know in which way an action, or a distress, will cascade or transmit throughout the complex infrastructure network    Resiliency and dynamically changing events in a region. Ways to measure the dynamically changing events once a hazard has been unleashed, can only be tracked by the use of simulators, like i2Sim that was developed by the Critical  2  Infrastructure Interdependent System Group (CIS) at the University of British Columbia (UBC)   Scenarios and regional models. Regional models, test cases and scenario sets need to be developed in order to understand, exercise and test the interdependencies and resiliencies of the region    I2Sim. Simulators are needed to dynamically compute interdependencies and resiliencies  Those steps were discovered during the present research because the motivation was how to prepare, respond and recover from natural and man-made hazards, and I will be addressing all of them in different chapters of this thesis.  1.4 Introduction The Government of Canada recognised that disasters can adversely affect human survival, and they decided to fund research to develop innovative ways to mitigate large disaster situations. This thesis has 7 chapters that will address the interdependency approach of MultiHazard Risk Assessment. These chapters will explain the methodology to evaluate interdependencies and consequences, and it will demonstrate the use of a simulator (i2Sim) to help prepare, respond and recover after light to major hazard events. Modern cities have new and upgraded construction technologies and new infrastructure systems, therefore new connections among Critical Infrastructures Systems are being generated. The methodology proposed in this research takes into account that a region can be exposed to all sorts of natural and man-made hazards, but in the present study it was tested for earthquake hazards. Therefore seismic risk assessment on an extended set of Lifeline Systems was used to estimate damages in CIs after an earthquake has occurred and; the corresponding consequences and cascading effects were evaluated. This research contributes to the enhancement of the resiliency of Critical Infrastructures.  3  The methodology will help all authority levels to respond collectively to risks and target resources to the most vulnerable areas within a region. The main focus of the interdependencies computed in this research is targeted to help all managerial levels, in this sense the focus of this research resides in a human survival objective.  1.4.1 Functionality Definition A regular Google search for definitions would show: ―... the capabilities or behaviors of a system; the total set of its features, or "the things it can do"...‖, among other definitions. In this thesis functionality (or operability) conditions refer to the physical and resource conditions of a Critical Infrastructure System. In this research functionality (or operability) would be the sum of the physical components and the resources that are needed in a system to perform or deliver services or products. Along the discussions, the functionality condition would be interchanged with operability or operational conditions; and the reader should rest assure that it has nothing to do with the definition of function in mathematics, or the associated definition of risk, probabilistic or reliability studies. The degradation of the functionality conditions in a system would be a function of physical components and resources. As an example, 100% functional conditions of a Water Pumping Station would be interpreted as the physical conditions of the Station (structure and non-structural components) being in perfect conditions (100%); and the resources needed to operate the station (electricity – coming from the electrical substation through distribution lines; water – coming from the water station through distribution water pipelines; and personnel – coming from residential areas through transportation and road networks) being in perfect conditions (100%). Another example is the functional conditions of a hospital, that are given by the physical attributes of the facility (structure and non-structural components) and the resources it needs (electricity – coming from the electrical substation through distribution lines; water – coming from the water pumping station through distribution water pipelines; nurses, doctors and technicians – coming from residential areas through transportation and road networks; 4  medicines and other utilities – coming from warehouses through transportation and road networks; etc.). The relationship between the physical components and the resources is investigated empirically with the operators of every system. The reader should be advised that one system might have only two functional conditions: functional (100%) or non-functional (0%).  1.4.2 Objectives The objectives of the research are: 1. To develop a methodology for evaluating risk from natural and man-made hazards that helps expose vulnerabilities in order to strengthen the resiliency of critical infrastructure systems 2. Identify  interdependencies  among  critical  infrastructures  and  obtain  Interdependency Indices for Critical Infrastructure systems 3. Provide documentation on how the risk estimates should be implemented by the UBC-JIIRP project; and 4. Implementation of results, assessments and outcomes for the simulator i2Sim  1.5 Organization of the Thesis This research has proposed a methodology to evaluate interdependencies and therefore understand how to mitigate large disaster situations. The seven chapters of the thesis focus on the interdependency calculations and the way to use this evaluation in the Infrastructure Interdependency Simulator (i2Sim). Chapter 2: Literature review. This chapter will address several methodologies that have been used for Multi-Hazard and Risk Assessment. It will address the beginnings of seismic risk; the reconnaissance efforts after major earthquakes; the Applied Technology Council efforts in earthquake engineering solutions; FEMA and HAZUS -FEMA/NIBS (1997) - research projects and leadership in mitigating risk in the US. This chapter will also describe the evolution 5  of earthquake engineering technology to encompass multi-hazard technology to cope with earthquakes and other natural and man-made hazards, including hurricanes, power outages and terrorist attacks. The latest related research associated with interdependency of critical infrastructure systems is given special attention. In this chapter, a discussion of seismic risk assessment studies all over the world, with emphasis in interdependency themes, is presented and discussed in a chronological fashion. All the past studies acknowledge the lack of comprehensive interdependency among Critical Infrastructure Systems. Chapter 3: The Joint Infrastructure Interdependencies Research Program (JIIRP). JIIRP was part of an effort by the Government of Canada, to fund research to develop innovative ways to mitigate large disaster situations. Six universities across Canada were involved. The title of the project was Decision Coordination for Critical Linkages in a National Network of Infrastructures. The UBC-JIIRP group developed a ―system of systems‖ simulator i2Sim that can model the interactions among infrastructures, and how their failures affect all other infrastructures and the wellness of the system. Part of the project was concerned with people and organizational solutions. Chapter 4 presents the Interdependency approach. In this chapter different ways to measure interdependencies are proposed. It was clear that by revealing the level of Interdependency and the Importance of every component to the whole system, disaster-recovery objective functions are easier to develop. The methodology in this chapter proposes a way to measure Interdependencies and Resiliency in a region or in a system. Events along with decisions to bring a system to a complete recovery can be simulated; several tables and calculations were proposed to reveal interdependencies, interconnections, and consequences of having functional, physical and resource problems. This methodology was implemented in the Interdependency Simulator tool (i2Sim) developed by the CIS group at UBC. Chapter 5 describes the UBC Test Case. The methodology developed during this research was tested in the University of British Columbia (UBC Test Case). On six significant lifeline systems (Critical Infrastructure): Buildings, Water System, Gas System, Steam System, Road System and Electricity System. The geographical location, infrastructure complexity, and 6  the diversity of its population made it an ideal test case to develop, assess and validate the methodology and the simulator (i2Sim). An earthquake scenario was selected as the disaster to be simulated in the test case. Due to the size of the study area and the amount of time and resources available, Risk Assessment was deemed to be the most appropriate method of determining the probable seismic damage. Chapter 6 presents a different Scenario and model, the Downtown Vancouver and BC Place Scenario Case. The methodology proposed here and the simulator i2sim were used to model a scenario incident during the Vancouver 2010 Olympics in the downtown Vancouver area. Casualties result from damage at one of the venues and from crowd surge during evacuation. The results of the study show that depending on the number of victims and on the conditions of the traffic routes, the facilities in nearby SP and VG hospitals are not sufficient and extra measures need to be taken to prevent a large number of fatalities. Chapter 7 presents the Conclusions of the project.  1.6 Contributions to a Multidisciplinary Team In the process of this research, I had to move from an individualistic approach into a multi disciplinary mind set. I had to recognize the value of my contributions as an individual researcher in this project; it is very hard for me to measure the contribution percentage of my work. But I will divide the work into three categories: 1. Conceptualization 2. Simulator (i2Sim) 3. Implementation of methodologies and i2Sim I would rate my participation and contribution to the whole project as 1/3. These contributions can be found in the three categories. My contributions to the project are: 1. Conceptualization: 7  a. In i2Sim framework it was established two operating modes: Physical and Resource. This conceptualization came with the fact that a critical facility needs the structure and NSCs to operate -physical, as well as other utilities arriving to the facility through IT services, pipelines or roads -resource. b. Damage estimation is needed to account for functionality conditions, and hence 5 levels of operability (green – 100%; blue – 75 %; yellow – 50%; orange – 25 %; and red – 0%). These levels will account for the physical and resource modes c. The same level of operability was implemented to channels, as they are also affected by physical and resource modes. d. Egress and Transportation models. The basic ideas of an egress and transportation models, as well as the modifier concept. e. Characterization and definition of the Damage Assessment module f. Methodology for evaluating interdependencies among critical infrastructures g. GIS visualization of interdependencies among Critical Infrastructure 2. Simulator (i2Sim) a. Identification of CIs to be modelled in the simulator b. Events table concept, that has to do with the scenario event c. Visualization and architecture in i2Sim 3. Implementation of methodologies and i2Sim a. Data collection of CIs and damage assessment to be implemented in i2Sim i. Lifeline damage assessment and methodology ii. Building damage assessment, the way to include structure and nonstructural components as part of a functionality condition b. Identification of critical assets c. Scenario developing d. Pre and post process input files and results from i2Sim e. Resiliency concept f. Modelling refinement 8  g. Calculations of casualties due to delay in treatment h. Methodology for evaluating interdependencies among critical infrastructures All of those items will be discussed in the next chapters.  9  2 Literature Review 2.1 Introduction Most of the research on risk has been focused on particular and individual utility systems, on ―stakes‖ taken by insurance companies, or with the aim to allocate monetary resources in catastrophe bonds. Most of the purposes, objectives and efforts of risk assessment studies have been put into minimizing costs. In many studies there are references and even ideas on how to tackle the interaction of several utility systems (e.g. an electrical and a water system). But there are only conceptual ideas, no metrics have been proposed to measure the extension of the consequences when a power outage occurs, and how a sector of a city will survive without power. None of the studies have focused on how to preserve human lives after the disaster has occurred, or how to manage a post disaster situation in a ―war room‖. As seen from experiences in other disasters, ―reaction‖ was the way in which managers responded to the lack of services, and the way they distributed the scarce resources through the affected region; most of the time, the resources and utilities are distributed without coordinated objectives to mitigate the lack of services and utilities in the most vulnerable areas. In this chapter I will address the fact that disasters have imposed serious problems to infrastructure systems. I will talk about different hazards (hurricanes, earthquakes, terrorist attacks, etc.), but the emphasis will be given to earthquakes, but earthquake engineering methodologies can be easily extended to other hazards.  These methodologies can be  implemented to address interdependencies among CIs. Interdependencies have been studied in the past, perhaps using other terminologies, but in the last 10 years interdependency problems have been witnessed after major hazards have occurred. Different countries, Canada and the US, have recognised that this is an important issue that governments should invest in resolving, and therefore they have funded research on this topic. This research and the literature review will focus on the work on interdependency among critical infrastructure in the last few years; this does not imply that in the past these topics were neither important nor emphasized. 10  2.2 Natural and Man-Made Hazards Human lives are lost every year due to hurricanes, monsoons, typhoons, floods and related weather events. Human lives, services and resources are destroyed every year by such events. In Bay of Bengal in 1970, in Bangladesh, 500,000 were killed by the Great Bhola Cyclone; in 2008 cyclone Nargis, in Myanmar, killed 140,000 inhabitants. In 2005 hurricane Katrina killed 1,833 inhabitants and losses were estimated in almost 85 billion dollars; hurricane Andrew in 1992 killed 26 people and the economical loss was estimated in 48 billion dollars, www.wunderground.com. History has proved that natural and man-made hazards can bring a large city to a complete standstill. Operational and functional human errors can lead to cascading effects in other systems, as seen after the Kobe earthquake. There, the decision to re-establish power service in the affected areas, where wooden structures were collapsed, caused the postearthquake fires which devastated the area. Minor storms such as lightning storms or heavy rain, or failure in equipment can cause major events like the North-eastern blackout of US and Canada in 2003. Pandemics also have the potential to cause distress and affection to a whole country, this is the example of the Influenza AH1N1 outbreak in April of 2009 that brought chaos and distress to Mexico. Economical impact was estimated at 1.3 billion dollars due to financial stimuli to related affected business sectors in Mexico, Secretaría de Hacienda y Crédito Público (2009). Finally, natural or man-made events can cause major disruption to our ―normal life‖, and our modern cities and complex systems may become useless as cascading effects will trigger wider consequences in our modern and technological habitat. In the following paragraphs, storms, and man-made disasters will be discussed in a chronological fashion.  2.2.1  Terrorist Attacks, September 11th, 2001 in the USA In the morning of September 11th, 2001, four aircrafts were hijacked, and were crashed  into iconic buildings in New York City and Washington DC in a period of 70 minutes. Two aircrafts crashed into the WTC towers in New York City. A third plane crashed into the 11  Pentagon, in Washington, DC. The fourth plane crashed in a suburban area of Pittsburgh. Around 3,000 were killed in the collapsed buildings, and the four aircrafts. This attack had a great effect on local infrastructure. Local emergency services were seriously affected when hundreds of responders were killed in the collapsing towers. Subway Public Transportation was affected; water supply suffered severe damage as a consequence of the collapsing of both towers. Due to the rupture of main water pipelines, local communication was disrupted; firefighters had to deal with lack of infrastructure and water pressure to deal with fires that followed the collapses; a telecommunication hub and cable system were also damaged and caused the system to stop activities. As a result the financial stock exchange market suffered the consequences of not being able to communicate their exchanges, O‘Rourke et. al. (2003). These events impacted Critical infrastructure (CIs) in two ways. First, CI facilities and operations were directly disrupted by the physical damage. The WTC and the surrounded area had several key businesses that support CIs, including banking and finance, transportation, and communications. Second, the decisions made after those events also impacted CIs: the Federal Aviation Authority (FAA) cancelled all commercial flights; the financial and banking sector temporarily closed key markets as a safety precaution; increased demand for telephone and Internet connections forced carriers to truncate their services to avoid crashing their networks, (Office of Critical Infrastructure Protection and Emergency Preparedness (OCIPEP), 2002). The lessons learned from the Canadian point of view were compiled in OCIPEP (2002), some of them are summarized here:   Communications: Predetermined emergency phone lines should have call priority and immediate service attention. This will assist emergency response in crisis. During a crisis, a deployable emergency information management capacity will assist first responders and victims; it will also help mitigate the affliction on government and health infrastructure following a disaster.    Transportation: The ability of transportation infrastructure to sustain normal functions will be jeopardized if sufficient planning and resources are not dedicated to cope with disasters. The interruption of one transportation channel (air) will result in mass usage and delays of alternative transportation channels 12  (marine, rail, roads, etc) and will affect operations involving critical personnel and sensitive materials.   Energy: Communications and business continuity will collapse after the initial impact of a disaster if backup power generation is not provided with guaranteed access to fuel and maintenance. The rapid restoration of power to critical sites will depend on a list identifying and prioritizing sites which are particularly vulnerable to prolonged outages.    Banking and finance: Advanced planning and communication among CIs will minimize the impact of interdependencies on business continuity plans.    Government: The development of a wide alert system with high levels of security and infrastructure redundancy will improve the government‘s ability to coordinate its response to CI threats. Web portals that post emergency response information will help governments to convey important information to citizens during times of crisis; also emergency communications will assist in dealing with the difficult situations. Rescue services and law enforcement should reach agreements on how to manage the rescue of survivors and restoration of normal city functioning while preserving evidence for criminal investigations.  2.2.2  The North-Eastern Blackout of 2003 in USA and Canada In August 14th, 2003 power outages set up in the Northeastern United States ―a wake-up  call to decision makers,‖ Nozik (2003) stated that existing systems of the US can trigger a cascading series of errors that would leave the population vulnerable. In 1996, western United States lost power because a power line heated up, sagged, and shorted out. In 1998, there were two power failures: ice storms knocked out power systems in eastern Canada and the United States. The Northeastern blackout of 2003 appeared to be from a strike of lightning. A total of 55 million people were affected in the US and Canada; some areas lost water pressure, the northeast railroad corridor service was stopped, passenger screenings were affected at airports, regional airports were shut down, in New York flights were cancelled due to difficulties accessing ―electronic ticket‖ information, communication through cell phones was disrupted, but wired telephone was functional, ELCON (2004).  13  Hurricane Katrina, August 29th, 2003 in New Orleans  2.2.3  On August 23rd, 2005, Hurricane Katrina became a tropical storm somewhere near the coasts of the Bahamas. The tropical storm grew into a catastrophic hurricane in the following seven days, and it made landfall in Florida and then in Louisiana and Alabama. Katrina provoked physical destruction along its path, flooded the historic city of New Orleans, ultimately killed over 1,300 people, and became the most destructive natural disaster in US history. The hurricane hit several towns, cities and states. Individual local and State plans, as well as new plans created by the federal government after the September 11 th, 2001 events failed to account for widespread or simultaneous catastrophes. Hurricane Katrina hit small towns and large cities, thus the largest search and rescue operations in US history was originated. Hurricane Katrina‘s national response included all levels of US government—Federal, State, and local—the private sector, other countries, and individual citizens. People and resources were sent to the region to aid the emergency response. Despite these efforts, the response was not enough and was not predicted in the US federal response plans. Townsend (2006) presented ―The Federal Response to Hurricane Katrina (Lessons learned)‖ and 17 Critical Challenges were addressed in the report: National Preparedness; Integrated Use of Military Capabilities; Communications; Logistics and Evacuations; Search and Rescue; Public Safety and Security; Public Health and Medical Support; Human Services; Mass Care and Housing; Public Communications; Critical Infrastructure and Impact Assessment; Environmental Hazards and Debris Removal; Foreign Assistance; NonGovernmental Aid; Training, Exercises, and Lessons Learned; Homeland Security Professional Development and Education; and Citizen and Community Preparedness It became clear that Hurricane Katrina had exhausted all emergency planned responses in one of the most developed countries in the world. Many of the 17 challenges are unresolved tasks and they are very hard to address. Once again, this natural disaster proved that we are far from planning for effective emergency responses. 14  2.2.4  Beginnings of Seismic Risk Seismic risk assessment has been investigated all over the world. In United States,  particularly in California, seismic risk studies have been conducted for many years. Allin Cornell (1968) in his seminal paper ―Engineering seismic risk analysis‖ introduced probability methods in earthquake problems. Steinbrugge (1982) used the Modified Mercalli Intensity to develop a relationship between damage and ground motion. Kramer (1996) presented a compilation in his book, Geotechnical Earthquake Engineering, and showed how the Seismic Hazard Analysis has been used in recent years. McGuire (2004) published a monograph that presents the state of the art for the methods of seismic hazard and risk analyses.  2.2.5  The 1971 San Fernando Earthquake On February 6th, 1971 an earthquake in San Fernando Valley, California showed that  earthquakes could inflict damages to a set of lifeline systems. Olive View Hospital opened in 1970 suffered damage and collapse of some buildings. It was retrofitted and reinforced and was opened again in 1987. The Sylmar Veterans Hospital also collapsed, and electrical, water, and transportation systems suffered extensive damage. Failure of a critical infrastructure and buildings can expose the weaknesses of emergency response plans of local, state and national authorities, Jennings (1971) and Allen, et. al. (1971).  2.2.6  ATC-13 Earthquake Damage Evaluation Data for California In 1985, the Applied Technology Council released ―ATC 13 - Earthquake Damage  Evaluation Data for California‖, ATC (1985). The document summarizes the four tasks that were studied: 1) identification and characterization of earthquake shaking, in order to estimate damage and losses; 2) development of schemes for facility classifications, that would take into consideration all existing facilities in California; 3) development of earthquake damage and loss estimate relationships in terms of the characterization of shaking intensity (MMI) and the facility class identifications, and; 4) development of inventory data and methodology consistent with the facility classification and the inventory data available at the time in FEMA. In terms of 15  facility classification ATC-13 organized all structures and infrastructures into 78 different facility classes, 40 of which were buildings and 38 were termed as ―other structures‖. The loss of function was related to social function classes; the classification had 35 classes of facilities. For earthquake damage and loss estimates ATC-13 described the expected physical damage due to ground shaking; expected losses due to collateral earthquake hazards, and; the expected percentage of casualties. The expected physical damage by ground shaking sustained by a building is related to Damage Probability Matrices (DPMs). For each facility class, the DPM expresses the probability of being in a certain damage state given the Modified Mercalli Intensity scale (MMI). There are seven damage states, each of which is associated with a range of Damage Factors (DFs) and Central Damage Factors (CDF). These damage factors signify the ratio of dollars lost due to damage to the total replacement value of the structure. Damage to buildings and utility systems were a concern for FEMA. Many limitations were still prevalent at that time: characterization of earthquake shaking; an inventory and classification of prevalent structures in the region; a methodology that relates earthquake damage and loss estimates, and uncertainties.  2.2.7  The 1985 México City Earthquake During the 1985 México City earthquake the functionality of important sectors of the  city was compromised because Critical Infrastructure systems collapsed. Two hospitals collapsed and so emergency agencies could not respond properly during the crisis. School buildings collapsed or had severe damage. Open fields, and stadia were used as morgue facilities; field hospitals and other facilities were used as emergency rooms for injured people. The subway system, in some sectors of the city, was closed due to collapse of neighbouring buildings and damage assessment activities in the subway infrastructure. Public transportation suffered the consequences of such events. Due to ground shaking, collapsed buildings, and permanent ground displacement, main water and sewage pipelines were severed and it took several weeks to re-establish services. Evacuees were forced to leave their homes and neighbourhoods; tents and provisional structures 16  were built in parks, avenues, and open fields of México City as ―temporary‖ shelters. Some of those temporary shelters had evacuees for more than 3 years. Downtown México City was abandoned by residents who moved out to northern metropolitan areas. Due to the lack of services, or limited functionality of some of those sectors of the city, ―normal‖ life was severely affected. Many weeks after the earthquake some of these normal activities were recovered, but some were delayed or permanently lost. After September 19th, 1985 the earthquake engineering society of México began working on methodologies to assess damage in buildings. Iglesias et. al. (1987), and later Iglesias (1989) proposed a methodology to evaluate damage in medium rise concrete buildings. The methodology used techniques and methodologies developed in Japan by JPDA (1977) and Uemura (1980), which were adapted for concrete structures in México City. A simplified method was developed to assess the seismic capacity of medium rise concrete structures. Almost 900 concrete buildings were studied; information to apply the methodology was collected for 300 buildings. Finally, 162 buildings were used to obtain evaluation indices (base shear coefficient corresponding to failure, ―resistance coefficient‖). Only 90 buildings suffered severe damage, and those values were used to elaborate a map of intensities. The map of intensities revealed the interaction between rock and soft soil that led to amplification of the ground motion in some districts of México City. This methodology was adapted to assess the seismic capacity of buildings without damage, so that vulnerable structures were ―discovered‖, in order to take the preventive actions to improve the security of the population in future earthquake events. The intention was to reveal vulnerabilities and set up mitigation strategies. At least 4 seismic sources have been studied, and it is discussed that earthquakes have shown directionality in the Mexico City Basin, Gómez et. al. (1991). In the paper by Noreña et. al. (1989) a building inventory and a three level of increasing precision methodology that would reveal buildings in bad conditions were presented. The main objective for local authorities was to prevent future damage in buildings of high importance, (Group A, according to the Mexico City Building Code); the main objective was to protect the rest of the critical buildings of the city. An emergency building provision was established for Mexico City after the 1985 earthquake, and buildings that were classified in the Group A (critical buildings) were asked to be strengthened and retrofitted to match the new 17  seismic provisions. Mexico City authorities provided this methodology that would include four steps: 1) a building inventory, 2) a simplified screening method to find evident structure problems, 3) the simplified methodology that was used to develop the map of intensities and 4) a detailed structural assessment using linear and non-linear analyses. This methodology was used as a retrofit procedure. The methodology was related to resistance or level of performance in the building. It was closely related to design procedures and overall seismic performance of medium rise concrete buildings. Drawbacks for the methodology were: that it was applicable only to medium rise concrete buildings in Mexico City; it was not a universal methodology, and research needed to be conducted for other construction materials. It did not consider the functionality of the building in different levels of failure; non-structural components or building contents were not considered in the assessment. From 1985 to 1995, the earthquake community in Mexico worked very hard to develop comprehensive seismic provisions, to review proceedings for design and retrofit of old and new structures. It was a great effort and many contributions were developed at this period of time. The main focus was to increase Mexico City‘s building resiliency. In 1995, three moderate earthquakes showed that we were far from prepared in Mexico. Ometepec, Guerrero; Villaflores, Chiapas and Manzanillo, Colima earthquakes occurred in September and October of 1995. Reconnaissance teams were again put together to visit the epicentral areas, and the affected zones of those earthquakes. One of the reconnaissance groups was established by Universidad Autónoma Metropolitana, and it was established to continue with the labor of a Mexican engineer, Jesús Figueroa, who reported damages and isoseismal curves in earthquakes that occurred in the 50‘s, 60‘s and 70‘s; Figueroa (1963) and (1970); Sordo et. al. (1995) and Juárez García et. al. (1997). Figueroa showed that, in some of the isoseismal curves, pockets of ground shaking amplifications were reported, with no further explanation, Figueroa (1963).  18  2.2.8  The 1989 Loma Prieta Earthquake Moderate to large earthquakes proved that we are far from understanding the  consequences that such events impose to infrastructure systems. In October, 17 th, 1989 in Loma Prieta, California an earthquake proved again that we have a lot to learn from earthquakes. Tarakji (1992) stated that: “Not only were the types of vulnerable structures known, but what was needed to be done to improve their seismic strength was known as well. The reasons for this inaction are numerous and semi-legitimate, and the blame for it should be shared by the public who refused to pay the high cost of earthquake preparedness, by government officials who had their priorities in the wrong order, and by the engineering community who had historically opted not to play an active role in government and public policy.” The argument was that technical lessons were numerous and that they were understood and learned. But in terms of management and policy, the lessons from previous earthquakes were ignored. Two important facts were evident after the Loma Prieta earthquake: 1) unreinforced masonry behaves poorly during moderate to large earthquakes; 2) lifeline systems were affected by the damage sustained by the transportation network, and the Marina District, Benuska (1990). Disaster response drawbacks were accounted as: lack of a state-wide emergency communications system; lack of a regional disaster response plan complicated the coordination among agencies; first responders were not equipped nor prepared to deal with the rescue process; and transportation contingencies were not properly pre-planned nor pre-arranged, Tarakji (1992). Tarakji (1992) discussed that actions were needed in earthquake vulnerability mitigation and rescue-response areas in order to reduce risk. Some measures were recommended, and despite the 20th year anniversary of the Loma Prieta earthquake, many of those measures are still 19  current: 1) establish emergency response organizations (federal or local); these organizations should include all levels of stake holders (government agents, first responders, engineers, etc); 2) earthquake engineering research should include as many organizations as possible; the research has been dominated by few large universities. The earthquake engineering research has been focused on the glamorous aspects of it; and ―less glamorous‖ (e.g. unreinforced masonry) areas need attention. 3) Establish and implement practical earthquake disaster prevention measures. Identification of vulnerable lifeline systems (highway structures), and the development of effective mitigation plans for other lifeline systems. 4) Government officials and policy makers should involve engineers and other technical experts in decision making processes; and 5) involve the community in earthquake response measurements. Perhaps some of this criticism was heard by government agencies and managerial levels.  2.2.9  Federal Emergency Management Agency in the US “The Federal Emergency Management Agency coordinates the federal government's role in preparing for, preventing, mitigating the effects of, responding to, and recovering from all domestic disasters, whether natural or man-made, including acts of terror.”, www.fema.gov. Emergency and disaster activities were fragmented for almost two centuries in the US  but in 1979 many of the separated disaster-related responsibilities were merged into the Federal Emergency Management Agency (FEMA). The new agency had unusual challenges that emphasized how complex emergency management can be. The Loma Prieta Earthquake in 1989 and Hurricane Andrew in 1992 focused major national attention on FEMA. In 1993, FEMA had the first agency director with experience as a state emergency manager. Reforms were established to streamline disaster relief and recovery operations, preparedness and mitigation. The end of the Cold War allowed the US government to redirect resources from civil defence into disaster relief, recovery and mitigation programs, FEMA benefited from these policies.  20  Rapid Visual Screening of Buildings for Potential Seismic Hazards was developed in 1988 and updated in 2002, ATC (1988) and FEMA (2002), in order to perform seismic vulnerability assessment from rapid visual screening. Buildings are classified into one of ten prototypes and seismic vulnerability is described in terms of a structural score. This score is determined through the addition of the prototype Basic Structural Score (BSH) and the Score Modifiers (SM). The Basic Structural Score represents the negative log of the probability of collapse of the building given the Most Credible Earthquake (MCE) for the region. Score modifiers are used to account for characteristics that affect the seismic performance of buildings: the height, vertical and horizontal irregularities, the age of the building and the soil type on which it is founded. In 1988 FEMA started research to assess the seismic vulnerability and impact of disruption of lifeline systems in the US. The project‘s objective was to develop a better understanding of the impact of disruption of lifelines from earthquakes and to assist in the identification and prioritization of hazard mitigation measures and policies, ATC (1991). The project was divided in two phases: phase I, provided an overview of lifeline seismic vulnerability and impact of disruption. In the project the following lifelines were included: electric, water, transportation, gas and liquid fuel supply systems, and emergency service facilities. Two drawbacks were emphasized in this report: first, the results were considered as approximate due to the assumptions and methodology utilized, and; second, the available lifeline inventory data lack critical information. Phase II, provided a practical model methodology for the detailed assessment of seismic vulnerability and impact of disruption of water transmission and distribution systems.  2.2.10 The 1994 Northridge Earthquake On January 17th, 1994 an earthquake struck Northridge in the San Fernando Valley, but Los Ángeles was greatly affected. Moderate damage to the built environment was widespread; severe damage included collapsed buildings and highway overpasses. Several commercial buildings collapsed. Shops and offices were closed due to non structural damage. Hospitals were forced to evacuate their patients. The entire Los Ángeles County school system was shut down 21  to allow cleanup and damage repair. Major highway bridges were severely damage or destroyed. Water and gas pipelines broke, flooding streets and starting significant fires. The entire Los Ángeles area lost electric power. Failure in the highway bridges affected the normal functioning of Los Ángeles. The failures in the electric grid had a wide impact, throughout western US and Canada. Water service was interrupted, but wherever water was flowing, a ―water boil advisory‖ was in effect, especially in the epicentral area. Olive View Medical Center evacuated 377 patients, primarily because of water damage and water and power outage. This hospital was a 1980‘s building that replaced the Olive View Hospital that was severely damaged in the 1971 earthquake, Todd et. al. (1994). The 1994 Northridge event demonstrated the vulnerability of lifelines in essential facilities (hospitals, fire stations, emergency operating centres, etc.). Roof top mechanical and electrical equipment did not have adequate seismic design. Breaks, even small ones, in water lines and sprinklers caused flooding in lower floors. Internal lifelines in essential facilities require seismic design. There is also a need to have emergency electric and water supplies at essential facilities, Hall (1995). Reitherman (2000) observed that in California, the hospital earthquake disaster exercises, have not accounted for effects such as elevator and electrical outage, possible structural damage and the need for immediate engineering evaluation, or water leakage. Reitherman concluded that combining hazard reduction techniques (to try to prevent damage) and emergency response planning (to foresee how to control the damage that may occur) is far superior to relying on only one approach.  2.2.11 The 1995 Kobe Earthquake It was considered the worst natural disaster in Japan since the 1923 Kanto earthquake. Normal life in Kobe was completely interrupted after the earthquake; houses and buildings were destroyed, the transportation system was interrupted. Emergency response was carried out under extremely difficult conditions: breakage in the water main pipelines cut off water supply for fire fighters; loss of electricity and 22  communications kept the affected population in the dark. Damages to lifeline systems brought problems for several months to Kobe. Many roads were blocked due to the debris of fallen houses, buildings and damaged highway bridges. School buildings were used as shelters and the education activity was interrupted for a span of time. Utility crews worked effortlessly to restore broken water lines, sewage lines, downed power and telephone poles, and leaking gas lines. Due to lack of pre-disaster planning these activities were performed as part of reaction activities to re-establish services as soon as possible. It is believed that most of the fires ignited after the earthquake was caused by electricity being quickly restored in the most affected sectors of the city. Liquefaction and lateral spreading caused severe damage to man-made reclaimed grounds. One observation from the Kobe earthquake was that disruption of regional lifeline systems can generate a profound effect on the entire country, because of the economic interdependence of lifeline systems and their functions, Chung (1996). Heath (1995) argued that response management in Kobe earthquake was non-optimal. Many response management systems around the world are likely to fail in circumstances similar to the Kobe earthquake. Those systems will not be successful in providing the preparedness and management of resources to support agencies in dealing with multiple impacted sites, with limited access and resources of water, electricity, and gas. Communication and inter-organization interactions are crucial in order to provide optimal integrated management response due to a crisis or a disaster. Heath (1995) suggested that table top exercises or inter-organizational interactions were needed in Kobe. Interaction and integration can help reduce operational and communication barriers between respondents, this can be achieved with combination of proper planning and exercises. Dislocations in coordination and communication will arise in exercises as well as in crisis or disasters. Heath (1995) considered that the greatest lesson from the Kobe earthquake was that response managers have almost no time to respond ―right‖ to a crisis. Time, information, and cost become key elements in the response. The best way to acquire time and information, and to  23  reduce future costs, is to develop plans for response and practices of preparedness in response organizations and their host communities.  2.2.12 HAZUS HazUS (FEMA/NIBS 1997 and 2003) is a loss estimation software package developed by the Federal Emergency Management Agency (FEMA) and the National Institute of Building Science (NIBS) developed in 1997 and updated in 2005. The methodology relates the expected building damage to the spectral acceleration and displacement. Demand spectra are used to describe the input ground motions and the seismic performance of buildings is represented through the use of capacity curves. The level of spectral displacement and acceleration experienced by a building is determined from the intersection of the capacity and demand curves. Five damage states are defined for each building prototype and the probability of being in or exceeding each specified damage state given the spectral displacement is determined from fragility curves. In 2001, a seismic risk assessment study was conducted for hospitals and other essential buildings in Clark County, Nevada, using FEMA/NIBS approaches, Sack et al. (2006). The studies were carried out using two independent deterministic earthquake scenarios that had the same probability of exceedance. The results of both methods indicated that a large number of essential buildings would be severely damaged to the point of failure for the given scenarios. A comparison of the results indicated that the FEMA 154 results were more conservative.  2.3 Interdependency Studies From the characterization of hazards, all the way to the consequences and social studies, there are several researchers focusing in different topics of the Risk Assessment methodology. Several tasks still have to be developed, and new ones are being observed and needed to be addressed, such as the interdependency of critical infrastructure systems. Recent studies have focused in the classification of interdependencies, definitions, topology and testing of the proposed methodologies. Previous studies have focused on small24  sized real networks, and the models are tested using one or two networks. There is a compilation of papers that addressed this topic with different point of views, which reflect the need to investigate in this particular subject. Alexoudi et. al. (2008) proposed a methodology to evaluate the associated losses of interacting lifeline elements. This approach can be directly applied to group decision making problems without loss of information. Complex fragility curves of interdependent elements are produced using the ―inoperability matrix‖ and the fragility curves of independent lifeline components. Strasser et. al. (2008) presented a state-of the art of loss estimation methodology and software as part of a subproject of the Network of Research Infrastructures for European Seismology. This subproject is more specifically concerned with the development of a panEuropean loss estimation tool for rapid post-earthquake response in urban environments. Tsuruta et. al. (2008) stated that there has been insufficient research for interdependency among today's urban infrastructures. In this study, the effect of interdependency among critical infrastructures during earthquake disaster is analyzed qualitatively and quantitatively. First, interdependency among critical infrastructures is investigated from past disasters, sorted out as matrices and analyzed by influence diagrams. Next, a propagation model is developed using interdependency structure matrices. Then, the influence of the interdependency is surveyed quantitatively through a case study of an anticipated earthquake disaster in the Tokyo metropolitan area. The results show that almost all the infrastructures are interdependent, and electric power, telecommunications, and highway systems have a greater influence on other infrastructures. Wang and Au (2008) showed that observations from past earthquakes have clearly demonstrated that crucial water consumers, such as acute care facilities, must remain operational to rein in the losses, and this calls for the need for seismic mitigation of water supply to these crucial water consumers. This research presents a method for identifying critical links of water supply to crucial water consumers under an earthquake based on probabilistic methods.  25  Lee and Graf (2008) showed that detailed models of large lifeline systems are difficult to assess in detail for an exhaustive set of earthquake scenarios. They showed an improved methodology that incorporates the ground motion uncertainty in the process of selecting a reduced set of hazard-consistent scenarios, and demonstrates its application to the Los Angeles Department of Water and Power network systems. Hosseini and Vayeghan (2008) stated that road systems are very important for emergency response activities and reconstruction purposes in the aftermath of major earthquakes. In this paper a risk management model for a transportation system is introduced, as well as parameters affecting the road function in the aftermath of a major earthquake. The proposed risk management model considers the complete ‗road body‘ and its ‗structural components‘. Shi et. al. (2008) described an analytical model for seismic response of buried pipelines to surface wave propagation effects. This model accounts for the mechanism of shear transfer and relative joint pullout movement as a result of soil-structure interaction. Corotis and Hammel (2008) suggested that the role of the professional risk engineer is to supervise the society‘s efforts to mitigate the consequences of earthquakes and other hazards. Furuto et. al. (2008) proposed a method to assist the municipality and people in disaster areas affected by multi-hazards. They examined transit plans for evacuating residents to safer areas. Mitsunari et. al. (2008) repeats that many municipal governments in Japan are now drawing up their action plans. Private sectors are also forced to take a prompt action for establishing their own business continuity plans to protect and maintain their own businesses. Once a severe earthquake occurs, road networks will be the most vital infrastructure to convey food, medicine, people, and all other supplies necessary for victims. They developed a disaster management supporting system for earthquake taking road networks into account, targeted for evacuation and emergency transportation routes.  26  Mujumdar (2008) proposed that a holistic systems level approach should be developed, creating a continuous feedback loop that will provide input on the behavior of the whole system. This systems level approach provides a cost effective and efficient holistic solution to minimize seismic risk to society, as it takes into account interdependency characteristics of critical infrastructure systems. Kuo et. al. (2008) explained that for a hospital, medical functionality (MF) is the most important performance target after an earthquake. The research focuses on the loss analysis of MF due to damages of non-structural components (NSCs). The results of loss analysis in this study can be connected to seismic fragilities of critical NSCs, so that performance of postearthquake medical functionality can be provided to hospitals and concerned authorities. The US President‘s Commission on Critical Infrastructure Protection (PCCIP) in 2001 stated that the economy and national security of the US depend upon 5 critical infrastructure sectors: Banking and Finance; Transportation; Information and Communications; Vital Human Services (Emergency Services, Government Services, and Water Supply Systems) and Energy (Electrical Power and Oil and Natural Gas Production and Storage). These five critical infrastructures are highly interdependent. Potential threats to the normal functioning of these infrastructures are both natural and man-made. PCCIP also stresses that in infrastructure protection, failure to adopt new security technologies means that vulnerabilities in the nation‘s critical infrastructures will persist. To eliminate these vulnerabilities, the government cannot afford to deal only with those firms that are highly motivated to collaborate – it should also engage those private sector owners, operators, providers, and users of critical infrastructure products and services that may not know of, or may not be particularly motivated to adopt, technologies developed through government investment. This PCCIP program, along with private sector efforts, should enhance the security of US critical infrastructures by rapidly identifying, developing, and facilitating the fielding of technological solutions and management tools and techniques to address existing and emerging infrastructure threats and vulnerabilities. PCCIP program provides guidelines, no methodology for implementation was developed for this matter. 27  Rinaldi et. al. (2001) discussed that: “...critical infrastructures are highly interconnected and mutually dependent in complex ways, both physically and through a host of information and communications technologies (so-called “cyberbased systems”), is more than an abstract, theoretical concept”. Later on Rinaldi and others emphasize that the identification, understanding, and assessments of such interdependencies are great challenges. And again, infrastructures affect fundamental systems and services that are critical to the security, economy, and social wellbeing of the study region. Rinaldi and others explain that there are other factors that affect infrastructure operations. In order to describe the infrastructure interdependency, it has to be defined: the infrastructure characteristics; state of operation; types of interdependencies; environment; coupling and response behaviour; and the types of failure. They presented a conceptual framework for addressing infrastructure interdependencies. The framework is used to explore the challenges and complexities of interdependency. They also discussed some of the research challenges involved in developing, applying, and validating modeling and simulation methodologies and tools for infrastructure interdependency analysis. In 2003 the National Infrastructure Advisory Council (NIAC) established a working group to study cross-sector interdependencies and provide risk assessment guidance. The Study Group reviewed previously published studies and recruited participation from all critical infrastructures. The Working Group concluded that cross-sector crisis management coordination is fundamental to the rapid restoration of critical infrastructure(s) and integral to sustain the public‘s confidence in those infrastructures. The working group advised the Department of Homeland Security DHS to adopt the following set of fundamental principles: by defining short-term deliverables, establishing a method to monitor progress of those deliverables, and fostering the commitment of the public and private sectors to partner for progress.  28  The Working Group identified nine issues for the NIAC that if not addressed, could polarize efforts to coordinate across sectors before, during, and after an event: “Issue 1 – Inconsistencies exist in the definitions of the critical infrastructures. Issue 2 – The “sector coordination” role is not broadly understood by industry and therefore is not viewed as a focal point for crisis management coordination within and across the sectors. Further, sector coordinating mechanisms have not been identified for all critical infrastructures. Issue 3 – Crisis management plans do not exist for each sector and are not tested end-to-end across the sectors. Issue 4 – A National Command Center does not exist as a confluence point for the private sectors during times of crisis. Issue 5 – Government-sponsored exercises should actively solicit private industry representation. Issue 6 – There is an underestimation of the dependency of the nation’s critical infrastructures on the Internet. Issue 7 – Coordination in planning and response between public emergency management (federal, state, and local) and private critical infrastructure is inadequate and/or inconsistent. Issue 8 – There is a lack of incentives that would help defray the additional expense burden resulting from strengthening the resiliency of the critical infrastructures. Issue 9 – Sophisticated modeling capabilities exist at the national laboratories and multiple research and development (R&D) studies on cross-sector interdependencies have been completed.” 29  They concluded that cooperation and collaboration are our best defence against risks resulting from cross-sector interdependencies. Critical infrastructures are inextricably linked and they also advised that the infrastructures‘ human counterparts should likewise be linked, McGuinn (2004). In 2007, O‘Rourke discussed critical infrastructure interdependencies, and resilience. O‘Rourke recalled that the concept of a ―lifeline system‖ was developed to evaluate the performance of large, geographically distributed networks during earthquakes, hurricanes, and other hazardous natural events. Lifelines are grouped into six principal systems: electric power, gas and liquid fuels, telecommunications, transportation, waste disposal, and water supply. Lifeline systems are interdependent, because of physical proximity and operational interaction. Damage to one infrastructural component, can rapidly cascade into damage to surrounding components, such as electric and telecommunications cables and gas mains, with system-wide consequences. To complicate matters, much of this critical infrastructure is underground, which obscures the location and condition of components. Lifeline systems all influence each other. This reciprocity can be found among all lifeline systems. O‘Rourke also addresses the resilience concept, which was discussed by Bruneau and Reinhorm in 2007 and by Bruneau et al in 2003; according to these documents both physical and social systems have four infrastructural qualities: robustness; redundancy; resourcefulness and rapidity. It also describes a resiliency graph, like the one in Figure 2.1, which shows the quality of a complete system (functionality) in the first 12 hours after an event. It is also stated by O‘Rourke that resilience can be promoted by awareness, leadership, resource allocation, and planning. But no metrics are proposed in this paper, just a broad idea of how interdependencies can be assessed in a complex system of infrastructures. In 2008, O‘Rourke and EERI published the ―Contributions of Earthquake Engineering to Protecting Communities and Critical Infrastructures from Multihazards‖. The purpose of this document, as stated in the report is: 30  “...to articulate, with examples, the ways earthquake engineering has enhanced public safety and improved the protection of U.S. communities from hazards beyond earthquakes.”  System Resiliency F 110% u 100% n 90% c 80% t 70% i 60% o n 50% a 40% 30% l i 20% t 10% y 0% -1.0  0.0  1.0  2.0  3.0  4.0  5.0  6.0  7.0  8.0  9.0  10.0  11.0  12.0  Time (hours)  Figure 2.1 System Resiliency The development of probabilistic seismic hazard analysis (PSHA) is a good example of innovation stimulated by earthquake engineering. Another example in the report is the rapid procedures for post-earthquake building inspection that were applied in 9/11, which helped in the restoration of the affected area of New York City. Modeling and managing interdependent systems requires simulation capabilities that can accommodate the many geographic and operational interfaces within and among the different networks. Earthquake engineering has contributed with powerful methods of modeling CIs performance. Investigations of earthquake recovery led to improved procedures for post-disaster reconstruction applicable for all natural hazards and human threats. The earthquake community provides leadership for disaster reduction. Its members make notable contributions to address the disaster reduction grand challenges and promote the technical and social advances needed for effective multi-hazard risk mitigation. 31  O‘Rourke (2008) refreshed the equation proposed by Ballantyne in 2003, which stated that Risk for water systems should be calculated as follows:  Another risk equation will be proposed in this research, and it will be addressed later in this thesis. The intention of this EERI report is to lay out the contributions of earthquake engineering. The purpose of the document is education and to help define and encourage leadership. The earthquake community has earned a strong reputation for advances in technology, planning, and policy that reduce seismic risk. These achievements are well regarded by those who work on other natural hazards, engineering and planning personnel responsible for civil infrastructure, and organizations responsible for emergency response and public protection. Leadership in earthquake engineering sets a high standard of performance. The most important leadership challenge is for the earthquake community to define its role in a multi-hazard world. The document presented by O‘Rourke (2008) does not provide a methodology to integrate several lifeline systems that are stressed by a disaster event. They only enlisted a series of contributions in the US that have helped increasing the mitigation and reduction of risk due to earthquakes, and how they have translated these technologies to solve for the risk that other hazards have imposed to the network of critical infrastructure systems of the US. Dueñas-Osorio et. al. (2007a) and Dueñas-Osorio et. al. (2007b) introduced a way to establish interdependencies based upon geographical proximity; and they suggested that we need to understand networks by measuring their properties (global connectivity, local clustering, and overall shape). They also claimed that understanding network failure mechanisms implies further analysis of these systems as a single interacting entity; and also upon the different roles that network elements have to maintain connectivity and facilitate flow transfer among several infrastructures. They proposed a framework to capture the interdependence among infrastructures in a single entity by using conditional probabilities of failure. This research 32  explored the ability of mitigation actions to decrease critical nodes‘ congestion and seismic vulnerability. They concluded that for mitigation, more sophisticated representations of the dynamics of networks need to be implemented. Physics-based models and time-dependent simulations may provide the tools to identify the conditions that lead to cascading failures. Dueñas-Osorio and Vemuru (2007) and (2009) studied the effect of cascading failures in the risk and reliability assessment of complex infrastructure systems. Conventional reliability assessment is limited to connectivity analysis and does not include the effect of increased flow demand or flow capacity of existing network components; the authors claimed that flow dynamics are needed for large-scale cascading failure simulations. These failures are caused by flow redistribution after disruptive events. This research showed that homogeneous network layouts coupled with weak links represent viable strategies to prevent cascading failures. Islanding is a strategy that deliberately disconnects portions of the networks to prevent larger consequences. The cascading analysis method of this study along with topological insights can lead to a pre-identification of network points for topology and load pattern homogenizations, identification of optimal element tolerances, and selection of weak link locations. Chang et. al. (2007) and Chang et. al. (2009) addressed the problem of interdependent failures of critical infrastructures during disasters. They proposed a conceptual framework for investigating infrastructure failure interdependencies (IFIs) from the standpoint of societal impacts. In order to identify empirical patterns, they developed a unique database of IFIs observed in major electric power outage events. The framework and approach are broadly applicable across a range of natural and human-induced hazards. Specific IFIs are identified that frequently occur and cause significant societal impacts. These results provide a basis for considering priorities for risk mitigation. These studies proposed: a conceptual framework for characterizing IFIs on the basis of societal, rather than technical, considerations; they developed a unique database of IFIs observed in a major Canadian disaster, and; they also demonstrated how the framework can be applied to empirically identify IFIs of greatest societal concern. Also, their empirical approach could be used to complement probabilistic, systems-based and simulation models of power outages and their impacts. Further, while this study focuses on IFIs deriving from electric power failure, the framework can be readily extended to assess other types of infrastructure 33  interdependencies and for setting priorities about potential ways to mitigate the likelihood and consequences of their interdependent failures. They have also developed a large database on the societal impacts of infrastructure failure interdependencies (IFIs) in a range of hazard events. Further research should investigate additional types of hazards (e.g., floods, earthquakes) and consider IFIs that result from initial failures in infrastructures other than electric power. The empirical data on IFI impacts can also be used in complement with predictive models of infrastructure failures and their interactions to anticipate potential consequences of future disasters. These can in turn be used to identify priorities for pre disaster mitigation and preparedness. Haimes et. al. (2005) discussed case studies of the inoperability input-output model (IIM) for modeling impacts on interdependent sectors. The IIM is a model for assessing sector vulnerabilities using the inoperability and economic loss impact metrics. Recommendations from this paper included developing cost-benefit-risk balanced solutions for managing and expediting recovery time from potential terrorist attacks. Lian and Haimes (2006) discussed the Dynamic Input–Output Inoperability Model (DIIM), which is an extension to the static Inoperability Input–Output Model (IIM). Both the IIM and the DIIM analyze how the system of interdependent sectors can be adversely affected as a result of initial perturbations to other sectors through attacks or natural disasters. Another limitation of the IIM and DIIM is that the models primarily deal with economic losses and associated risks; however, there are many other aspects of risk that are not directly reflected. These include life losses, personal freedom loss due to political changes, losses due to insurance, injuries, and changes in the international political landscape. In the current model, the capacity loss due to a major disruption is simplified as decreasing output. Therefore, another improvement of the model is to include capacity and resource constraint in major disruptions. Lian et. al. (2007) extended the IIM and the DIIM to encompass probabilistic dimensions of risk analysis. Haimes (2008) addressed the need for developing a body of prescriptive theory and methodology for modelling Systems of Systems (SoS). This will enable analysts to model and understand the SoS evolving behaviour due to the continued forced changes imposed on them.  34  Davidson and Çagnan (2004) developed improved models of post-earthquake restoration processes for electric power and water supply systems. The restoration models used estimates of physical damage to the systems and an understanding of the repair and recovery operations to estimate geographically-disaggregated restoration curves for each system, including uncertainty bounds and explicit representation of the key decision variables guiding the process. Tabucchi et. al. (2009) described a discrete event simulation model of post-earthquake restoration for Water and Power water supply system. It simulated the real-life process in detail, simulating the movement of different types of crews as they inspect, reroute around, isolate, and repair system damage. For any given earthquake, the model provides restoration curves with uncertainty bounds, maps showing the spatial distribution of outages over time, and crew and repair material usage information. Results suggested that the model is capable of accurately estimating the time and spatial sequence of the restoration. It can be useful for loss estimation and resilience assessment, evaluating the effectiveness of hypothetical restoration strategies, and improving understanding of the restoration process and its key determinants. This is the first application of discrete event simulation to post-disaster water supply restoration, and one of the first for any infrastructure system. Brink et. al. (2010) studied three primary strategy plans to use in the event of a significant earthquake in Los Angeles. Casalicchio et. al. (2008) used an agent-based solution to model and to simulate interdependent critical infrastructures. The research used discrete agent-based modeling and simulation and federated simulation (ABMS). They also described the formal model and the methodology that was used for a simulation of a complex system (communication network and a power grid). In all the literature review there are gaps and future work that have been emphasized by the Critical Infrastructure Interdependent research community on interdependency topics. This is a topic that has been explored for the past 10 years, and they have all agreed on things in particular, that need further investigation: 1. The need to evaluate interdependencies among critical infrastructure systems 2. Characterize the information needed to evaluate the interdependencies 35  3. The need to define topologies, ontology and a framework to work with different systems 4. A way to measure resiliency and operability conditions 5. The need of damage assessment of CI systems as a single interacting entity; and connectivity and flow transfer modes of the different network elements 6. The need of more sophisticated representations of the dynamics of networks. Physical models and time-dependent simulations may provide the tools to identify the conditions that lead to cascading failures 7. The need of studies on flow dynamics to simulate large-scale cascading failure simulations. These failures may be caused by flow redistribution after disruptive events 8. The need of implementing different hazards (attacks, storms, earthquakes, hurricanes, flooding, tsunamis, etc.) in the interdependency methodologies that account for disruptive events in a system 9. The need to include life losses, personal freedom loss due to political changes, losses due to insurance, injuries, changes in the international political landscape, and other losses that are not usually accounted for in the methodologies 10. The need to arrive to methodologies or simulations that will produce: time and spatial sequence of the restoration, loss estimation and resilience assessment, evaluating the effectiveness of hypothetical restoration strategies, and improving understanding of the restoration process and its key determinants. In other words ways to prepare before, respond during, and recover after natural and man-made hazards have occurred 11. The need to develop methodologies and tools that simulate disruptive events and their consequences, but also that are suited of following up real-time disaster situations 12. Ways to include the human factor in the interdependencies 13. The need to include reliability and uncertainty assessments 36  3 UBC-JIIRP General Project 3.1 Introduction The system of critical infrastructures (CIs) (power grid, water network, health system, etc.) constitutes the backbone of modern societies. During large disasters (e.g., earthquakes, hurricanes, terrorist attacks, etc.) the situation is very different from normal life in that multiple infrastructures are affected simultaneously, and unless they coordinate each other‘s actions, the overall response process may suffer serious setbacks. Inadequate coordination among infrastructures and among hierarchical decision levels has been a major cause of failures in the response of the system and it has been responsible for unnecessary losses in human lives, property, and economic activity. The need for cooperation and coordination among critical infrastructures is well recognized by the Canadian government which has recently issued the policy blueprint ―Working Towards a National Strategy and Action Plan for Critical Infrastructure‖, Public Safety Canada (2008). There are a number of reasons for the inability of large sectors to cooperate with each other. a) Commercial and competitive issues; b) Security concerns; c) Even if privacy and security were not a concern, it might still not be possible to exchange meaningful information because of incompatibilities among their physical communication devices, misunderstandings in their terminology (semantics), and, lack of knowledge of which information is important to be shared, with whom, and at which time point in the development of the events. The Joint Infrastructure Interdependencies Research Program (JIIRP) JIIRP is part of an effort by the Government of Canada, through the Natural Sciences and Engineering Research Council (NSERC) and former Public Safety and Emergency Preparedness Canada (PSEPC), now Public Safety Canada, to fund research to develop innovative ways to mitigate large disaster situations.  37  Six universities across Canada were involved. The University of British Columbia (UBC) studied decision support for critical linkages in infrastructure networks. The Joint Infrastructure Interdependencies Research Program at the University of British Columbia (UBC-JIIRP) was an effort to assess the impact of physical and temporal interdependencies among multiple infrastructure systems, during the development of large disaster events, since the impact of these interdependencies may be hidden by its temporal consequences, Martí et. al. (2007). The title of the project is Decision Coordination for Critical Linkages in a National Network of Infrastructures. The UBC-JIIRP group, was a team of researchers, graduate students, post doctoral fellows and research engineers from multiple disciplines: Electrical and Computer Engineering, Civil Engineering, Computer Science, Geography, Psychology, and Commerce. The group has been operating under a grant from NSERC, Public Safety Canada, and five industrial partners, under the Infrastructure Interdependencies Research Program (JIIRP). The group has developed a ―system of systems‖ simulator I2Sim that can model the interactions among infrastructures, and how their failures affect all other infrastructures and the wellness of the system. I2Sim incorporates multiple levels of production and response activities, from the physical to the human layers. The work plan in UBC-JIIRP was concerned with developing specific scientific and engineering solutions for the identification of critical linkages and decision coordination tools. But part of the project was concerned with people and organizational solutions. The following seven deliverables were set for the project: 1. Simulator of Interconnected Infrastructure. 2. Inventory of Existing GIS Capabilities. 3. Analysis of Case Study Scenarios. 4. Resource Integration and Data Visualization at System Control Centres. 5. Integration of GIS Resources from Dissimilar Systems. 6. Educational resources in the form of printed materials, presentations and CD‘s for training and education of infrastructure operators and decision makers. 7. Six Workshops and One Mini-Conference for system operators comprising discussions of ―what if‖ scenarios, group behavior and interpersonal dynamics. 38  The I2Sim simulation environment developed during the JIIRP project at the University of British Columbia provides a multi-system representation of all infrastructures involved in disaster response at multiple hierarchical levels (local, municipal, provincial, etc.) of the global system response. The flow of resources between component infrastructures is explicitly represented without revealing their internal details. In this combined environment, each member infrastructure uses its own internal models to determine its possible operating states, while I2Sim combines these operating states into a system-of-systems solution. Damage to the infrastructures during the disaster and greater demand for resources creates a situation where decisions need to be made as to the optimum time and allocation of the available resources. The simulator supports look-ahead and rewind functions to predict the evolution of the system dynamics in order to assess in real time the effect of suggested decisions before they are actually applied to the real system. In addition to its real-time decision support capabilities, the mathematical formulation of I2Sim permits the analysis and discovery of vulnerable points in the system as well as gaps in policies and procedures. The simulation environment is also particularly useful for the training of system responders and managers under realistic dynamically evolving scenarios.  3.2 Integration of Various Methodologies - Interdisciplinary Work This chapter describes a multidisciplinary research project that would integrate advanced analysis, modeling, and simulation techniques to develop new approaches for coordinated decision making in national infrastructures to deal with critical contingencies. The development of a critical infrastructure interdependencies monitor poses several research challenges, such as for example, development of a model of critical interdependencies that can support complexity; ability to gather and distribute appropriate information on the system dynamic states in a way that can be intuitively visualized and is comprehensible to human operators; prediction of the dynamic evolution of system events and anticipation of incipient emergencies, UBC-JIIRP (2004).  39  Due to the complex relationship among infrastructure systems and the many services at stake when a disruptive event occurs; interdisciplinary approaches need to be considered. When a hazard strikes a region, physical systems will be collapsed or broken, and first responders will also be affected. The predicted consequences need to be addressed, so that impacts are well known. The interdisciplinary approach is therefore necessary to fully understand the impacts of a failure in an infrastructure system, and the consequences that will impose to the ―study space‖ that is being serviced. Common language is essential to understand the different techniques and methodologies used by the interdisciplinary team. Exchange of information and understanding the units that are being exchanged by these systems are key elements for the overall assessment of the system. The wellness of the society will be preserved when the consequences of a certain event are accurately predicted and defined. Analyzing complex systems such as those formed by infrastructure networks and decision makers requires a multidisciplinary holistic approach. The field of research in infrastructure interdependencies is fairly new, and lies in the intersection of areas of knowledge such as emergency management, geography, simulation modeling, planning, and safety engineering.  3.3 Motivation Infrastructure systems are fragile, since failure of one infrastructure can bring the failure of others through cascading effects. The underlying susceptibility of these networks to disruptive events is in large part  due to the increasingly complex pattern of  intra/interdependencies, and hierarchies that tie these together. The study of critical infrastructure interdependencies is not only a complex problem in terms of its formalization, but also in terms of the intricacy required to test and validate that formalization. It is, however, only through the study of the interdependencies among infrastructure networks that certain failures or weaknesses in the systems can be discovered. It is challenging to analyze the interconnections 40  between infrastructure systems, and also studying these at moments of stress, when the interdependencies become dynamic. Unexpected outcomes occur. By studying the relationships between interconnected infrastructures and the way they operate at different stress moments, it is possible to come closer to a picture of what could fail or what should be subject to attention given a specific emergency scenario. Motivated by a need for an integrative process, UBC-JIIRP identified aspects of the interconnected infrastructures that need attention during large-scale disasters.  The project  objective was to improve the organization of disaster management through effective knowledge generation, so that the human impact of disaster could be reduced. UBC-JIIRP focused on the human component of disasters and the intersections of disasters and human psychology.  3.4 Ontology for an Infrastructure Interdependency Simulator (I2Sim) Due to their different intrinsic natures and their separate evolution, different infrastructures, for example, the power grid or hospitals may use very different descriptions to model their operation. A fundamental first step in the design of the modelling/simulation tool, named I2Sim by the UBC-JIIRP team, was to define an ontological representation that would allow the large diversity of entities that make up the system of infrastructures to be characterized using the same concepts (Marti, et al, 2008a and 2008b). Another crucial reason to develop this ontology was the large geospatial and temporal extension of the physical systems that need to be represented during major disasters. For example, in order to know how water can be delivered from the water pumping station to the hospital after damage has occurred in the pumping station, it is not necessary to model all the pieces of pipes that connect the station to the hospital. It is sufficient to know how much water can be carried from the pumping station to the hospital and how much spillage there will be. I2Sim‘s ontology defines a study scenario space with the following entities, Figure 3.1: a. Cells (Production Units): A hospital cell requires inputs: electricity, water, doctors, medicine, etc., and produces outputs: patients discharged.  41  b. Channels (Transportation Units): The electricity is carried to the hospital by wires, the water is carried by pipes, and doctors are carried by the transit system. c. Tokens (Exchange Units): Quantities that are the inputs and the outputs of the cells, e.g., water is a token, a doctor is a token, a phone call is a token. d. Controls (Distributor & Aggregator Units). Model the interface between the physical and the decisions making layers. As an example, if electricity supply is limited, how much electricity should be sent to the hospital and how much to the water pumping station (distributor). The total electricity that the hospital receives is the sum (aggregator) of the electricity that comes from the external source (substation) and from the reserve diesel generator. Distributors provide the links between the physical distribution of output resources and the human layer of decision makers. e. Hazards f. Events (real-time and ―what if‖ events) g. Scenarios (Simulation cases) h. Decisions A key aspect of I2Sim modeling is that the internal details needed to relate the inputs and outputs of cells and channels are determined by the owner of the infrastructure. In the case for example of the power grid, the relationship between the input sources of energy (e.g., the high voltage generation and transmission system) and the feeder lines to the loads is determined by very sophisticated software run by the power system control centre, which takes into account a number of technical and operational constraints. From the point of view of I2Sim, all that is needed is for the power utility to provide a table (―Human Readable Table‖, HRT) relating availability of output power for a number of possible operating states of the power system. The Operational Modes take into account the degree of damage that the cell unit may have suffered as a result of the disaster event. HRTs relate multiple inputs to multiple outputs.  42  Distributor Control points PM  RM  X  Cell 1 Aggregator Channel i  +  X  Cell 2 Figure 3.1 Cell, Channels, Aggregators and Distributors  3.4.1  Resources and Human Readable Tables (HRTs) As mentioned previously, HRTs are the means by which the performance of cells and  channels are modeled in I2Sim. It takes into account five states, which represent a percentage in functionality conditions, Table 3.1. Table 3.1 Percentages and Colors for Physical and Resource Modes Colors  Mode 100% 75% 50% 25% 0%  A critical facility (cell) will need the building to be in good conditions (structure); this facility will also depend on the integrity of pipelines, mechanical and electrical equipment (non structural components) inside the building; and it will also need furniture and specialized equipment to be able to operate (building contents). The facility also needs utilities or resources in order to be operable (water, electricity, etc.). The functional condition, in which the structure, the Non Structural Components (NSC) and the building contents are in, is defined as Physical 43  Mode (PM); and the condition in which the facility is operable due to the utilities or resources is called Resource Mode (RM). Consider a Facility which is defined with 3 Physical Modes and 3 Resource Modes, Figure 3.2. These modes depend on the amount of damage that the facility has sustained after a hazard event, and the availability of resources; these modes will be defined by utility managers.  Resource Mode Physical Mode  100 %  100 % 50 % 0%  Cell - Facility  50 %  50 %  0%  0% 0%  Figure 3.2 Cell with Physical and Resource Modes The physical and resource condition of a facility to be able to perform its function was translated into Physical and Resource Modes states for I2Sim, and these modes were defined in the Human Readable Tables (HRTs). Consider that the facility in Figure 3.2 is a Hospital, and that only needs three external resources to perform its function – Resource Mode; and the structure, NSCs and building contents are represented in the Physical Mode, then the HRTs can be defined as shown in Table 3.2.  44  Table 3.2 HRTs for a Hospital Patients/hr PM01 (100%)  output  Water  Doctors  inputs  y  x1  x2  x3  RM01  100%  100%  100%  100%  RM02  50%  60%  50%  40%  RM03  0%  0%  0%  0%  Patients/hr  Electricity  Water  Doctors  PM02 (50%)  3.4.2  Electricity  output  inputs  y  x1  x2  x3  RM02  50%  70%  60%  40%  RM03  0%  0%  0%  0%  Cell Model Cells represent functional units, for example a hospital. Functional units are modeled in  terms of a production model that relates the inputs needed for the unit to produce its outputs. In general the relationship between outputs and inputs of the cell will correspond to a multidimensional nonlinear function, Figure 3.1.  3.4.3  Channel Model Channels represent the means by which tokens (e.g., medicines) are transported from a  source cell (e.g., medical warehouse) to a consumption cell (e.g., the hospital), Figure 3.1. Channels are characterized by a loss coefficient and by a time delay. The loss coefficient accounts for leakage, e.g., in the water pipes, while the time delay accounts for the transportation time, e.g., of the truck delivering the medicines. In a similar way to the cell descriptions, channels are represented by HRT description tables but with the added parameter of a time delay.  45  3.4.4  Scalability The system can be scaled into a local layer of a possibly more complex and extended  disaster scenario. For example, the power to the substation is provided by the ―external source‖. From the point of view of the power utility, this is only one of many substations in its grid. The allocation of power to the various substations during the emergency is a decision that needs to be made at the next higher level in the system, or example at the municipal or provincial level. The same is true for the water supply system, etc. For decisions, for example, at the provincial level, entire cities can now be represented as combined ―larger‖ boxes which are now defined again in terms of total inputs required to produce total outputs. The model, therefore, escalates very well at the multiple hierarchical levels of a large disaster situation.  3.4.5  System Model Cells and channels are discrete abstractions of the physical world. These abstractions and  their input output descriptions allow I2Sim to set up a mathematical description of the interrelationships among interdependent systems. At each operating point along the time line of the developing event, this linearized description, corresponds to a system of discrete time equations and can be represented with a system ―transportation‖ matrix relating the input quantities that ―arrive‖ at the cells with the ―source‖ quantities that are produced at the output of other cells, Martí, et al (2008).  3.4.6  Hazards All the models in I2Sim are representations of the real world. Hazards affect the study  regions, and most of the time all these hazards are well known threats. A region can be affected by earthquakes or hurricanes; but others will be affected by winter storms and man-made hazards; or combinations of hazards to which the managerial level is preparing for. Magnitude or size of the hazard is well known, and the hazard-functionality relationship of cells and channels are also pre-defined and pre-assessed. Time of day, date, and other factors related to the hazard, like recurrence and attenuation can be used to pre-define a set of hazards for the region, Figure 3.3. 46  3.4.7  Events Events trigger changes in the model. They will define physical and operating modes for  cell and channels, and the times in which they occur. Events will be modeled with channels, and be considered as event channels, Figure 3.3. They will be developed for real-time or pre-scripted scenarios. Events can be inserted past-time (simulation), real-time or at a future time. This capability will help I2Sim perform during preparation, response and recovery stages in an emergency situation.  3.4.8  Scenarios (Simulation Cases) A scenario manages a simulation case including the hazards, all the components of the  model (i.g. cells and channels), events to the simulation (including real-time and ―what if‖ events) and simulator states for branching and resuming. The scenarios can be pre-scripted or real-time. It defines time of day, date, hazard-damage state of cells and channels, HRTs and number of events, Figure 3.3. It also activates a decision cell.  +  Hazard 1  +  x  +  +  Scenario Event Channel 1  x  Event Channel 3 Event Channel 2 + + +  Cell 1 +  x x  Cell 2  +  +  Figure 3.3 Hazard and Scenario Cells and Event Channels  47  3.4.9  Decisions The Scenario cell activates the decision cell through a channel. The decision cell defines  repair types for cells and channels, so that physical and operating modes are restored. Time of occurrence of the decision to take place is defined through repair channels. It also defines controls for the distribution of tokens to critical assets and locations and the estimated time in which these repairs and controls should happen. Command channels define the amount of tokens to be sent to the different locations and time of occurrence, whenever physical and operating modes change, Figure 3.4.  + Decision Channel  Scenario  +  x  Decision Repair Channel 1  x  + +  Cell 1  x  + +  Cell 2  Repair Channel 3 Command Channel 3 x  +  Figure 3.4 Activation of Decision Cells  48  3.5 Interdependency among Critical Infrastructures Infrastructures are socio-technical systems with physical and human components. Interactions between humans and objects are necessary elements of such systems. Several classifications have been proposed like Rinaldi (2004) that describes interdependencies between infrastructure networks according to their elements‘ dependency; or Dudenhoeffer et. al. (2006), they described infrastructures as nodes and edges, and they also defined more interdependencies‘ types. UBC-JIIRP aims to model the real time effects of a disaster and identify the interdependencies among critical infrastructure networks. There are at least five components of the projects architecture: the study space with physical and human layers, the hazard (event or scenario), the damage assessment (I2DAM), database (I2DB), and the infrastructure interdependencies simulator (I2Sim), Figure 3.5  Physical  Human  I2DB  Hazard  I2DAM  I2Sim  Figure 3.5 Components of the Architecture Project  49  Critical Infrastructure (CI) is referred to those systems that are needed to support liveable conditions in a region. A CI system might be a collection of: 1) several cells and channels; 2) several cells; 3) one cell; 4) several channels; and, 5) one channel. Along this thesis, all the examples have to do with Critical Infrastructure Systems, for simplicity it will be referred to them as Critical Infrastructure Systems, CI, CIs, cells or channels. But the referring acronym or name should not imply less importance to their critical participation in the whole system that is being represented. Critical infrastructure systems are interconnected with each other through other CIs (channels: pipelines, roads, communication and IT services, etc.); first responders also use these systems to provide aid and health care to people in distress, therefore the way in which infrastructures connect with each other, and with human layers (first responders and local and global authorities) is hard to evaluate and model. Emergency management procedures take into account human layers, this fact increases the difficulties for the modeling and assessment of CIs. Along the UBC-JIIRP project the following 16 steps have been proposed to arrive to a proper model of a study space. These guidelines along with the methodology will help revealing second order and higher interdependencies: 1. Characterize the ―study space‖. The selection of the ―study space‖ depends on the emergency objective of the manager, operator or local authority in charge. In this research a ―study space‖ can be a facility (building or venue), a lifeline system (i.e. a water system), a region, urban sector, a city, a province, etc. 2. Define to which extend the risk will be measured. At least two objectives can be established for the estimation of risk: business continuity and life safety. But the objective is defined by the owner of the study space. 3. Select the critical lifelines. Assets, buildings, networks or lifelines that serve the ―study space‖ are selected according to the emergency objective of the local authority. 4. Acquire general information. General information for all the critical assets is needed; it can be acquired through public domain sources or by asking the local authorities of the ―study space‖.  50  In emergency situations, stakeholders are willing to cooperate and supply the decision makers with data that will contribute to disaster response and preparedness actions. However, stakeholders are worried about revealing secured data that could harm their businesses. In response to terrorism threats, some information about critical infrastructure needs to be secured; because the breaching of those secrets would potentially do more harm than good to disaster response agents. It is a fact that personal information needs to be carefully controlled. During pandemics, the general public needs to be informed about the situation, but at the same time personal information of individual patients or victims should be kept secured. Security and privacy can be achieved through control at the access points of data. The problem in keeping data secure and private is how to assign or revoke access rights to the right personnel and computer systems. The architecture of UBC-JIIRP project assigns the stake holders full control of the access to their data. Sensitive data will never be accessed from non-trusted places. The only exposed part is the interface of this architecture that allows authorized users to query the data. It is more convenient to only expose the querying interface than sharing the entire data set. 5. Disassemble the “Study Space” into Physical and Human Layers. The physical layers represent the physical attributes of each infrastructure network. Each layer contains data such as the geographic locations of network components, physical properties, hierarchy of components, direction of ―flow‖, etc. The ―study space‖ includes human beings; emergency responder, local and global authorities play an important role in disaster planning and response. The human layers include people flow, disaster victim behaviour, first responder actions and decision maker roles. People flow can be viewed on two levels: the micro and macro scale. The micro scale level models evacuation routes in individual buildings: taking into account that some buildings or facilities will take longer to evacuate. The macro scale models where people will go once outside of the building and includes displaced people, emergency shelters and emergency care. Therefore it is important to disassemble the ―study space‖ into physical and human layers. 51  6. Define the Hazard or Events (scenarios or real-time events). For a lifeline system there are basically two operating modes: normal conditions and failure due to a hazard event. The failure of a lifeline system can be divided into a gradient of failure modes, from no damage to totally destruction. The hazard is selected, and all the necessary characteristics are defined, day, time, magnitude of the event, etc. Disasters include natural hazards such as earthquakes, floods, hurricanes and wildfires and man-made disasters. As an example consider a shortage in the water supply to a region, and the interdependencies among CIs, and hence the effects in a Hospital. The shortage will impose restrictions to the treatment protocol in patients of the hospital. Therefore, the functionality condition of the hospital will be jeopardized: imagine that dialysis treatment is on the 6th floor, many lives would be at stake by just a small shortage (it can be a consequence of a natural or a man-made hazard) in the water pressure. 7. Obtain GIS Maps to Represent the Physical and Human Layers. All the information is mapped in GIS, all the physical and human layers are compiled and characterized. 8. Acquire Detailed Information. Data collection is a key part for an interdependency project. Consistency between the scale of the hazard and the study area shall be considered. For example, if an earthquake or a large flood is simulated, the study area should be a city, a municipality, or a large region, and the degree of detail at which data is required is general. The detail of data required is higher when the event is a fire or a localized failure in a building that contains critical infrastructure elements. Granularity is important when defining the finest infrastructure unit/element. This is true for physical assets as well as for human components. The temporal resolution is also a critical factor in the interdependency modeling. Infrastructure systems use different time units to deliver their service units (tokens). In an electrical network, the action of rerouting power will be taken in fractions of a second. In a water system the time interval in which events happen is minutes to hours. In a human layer like emergency operators, decisions are taken in larger time intervals, depending on resources and information availability. This is 52  an additional challenge for modeling multiple infrastructure systems. Granularity reveals one important issue when modeling interdependencies: data quality and accessibility. The size of the region and its model in I2Sim is critical, it defines the quality of the data to model the system to make it operable. The scale will determine the data needs and the cooperation among infrastructure owners. The information will be very detailed when the purpose of the system is to model interdependencies at a local level; but it will be very general when the scale of the model is wider. 9. Perform Damage Assessments. The damage assessment involves the estimation of physical damage to the component, the number of casualties, the amount of economic loss, the loss of function, loss of service and a set of interdependencies as a result of the disaster. Interdependencies of critical lifeline systems are non-linear, time dependant and their behaviour depends on the level and nature of the hazard. In order to be able to identify risks in critical infrastructure (CI), the earthquake hazard and seismic risk assessment methodology was used to test the UBC-JIIRP framework. The Seismic Risk Assessment methodology can be modified and extended to include other natural and man-made hazards. Seismic hazard is often evaluated for given probabilities of occurrence. A general equation for Risk, considering the seismic hazard would be: Seismic Risk = (Seismic Hazard) x (Vulnerability) x (Value) Vulnerability is the amount of damage, induced by a given degree of hazard, and expressed as a fraction of the Value of the damaged item under consideration. As an example, the Monetary Seismic Risk to a building could be evaluated by taking the Seismic Hazard to be the Modified Mercalli Intensity of the appropriate curve for that intensity, and the Value would be the Replacement Cost, Dowrick (2003). The same procedure could be used for other hazards. The consequences by using different earthquake levels, storm and blast loading will show that the hazards impose consequences that range from light to severe. 53  Different hazards will impose different consequences to the ―study space‖, and hence it is important to show a broad variety of hazards in order to explore all consequences, from light to severe. 10. Discuss the Outcomes of the Damage Assessment with the Operators of the System (both at operational and managerial levels). All the results are then mapped in GIS, and the results are discussed in two different levels with the operators of the ―study space‖. Detailed and localized failures of the systems are discussed with operation agents (operational level). Global behavior of the ―study space‖, emergency planning and decision making is discussed with decision agents (managerial level). Results are visualized in maps, tables and detailed information so that operators understand the outcomes of the system after the disaster has been ―played‖. Adjustments to the information and new results are computed after the interviews with the operators.  11. Define Cells, Channels and Tokens. The simulator model is made up of four primary components: tokens, channels, cells and controls. Tokens are defined as the goods and services that are being produced or consumed by the population. Channels are units which transport tokens from one location to another. They are used to model lifeline systems such as roads, water pipelines and electrical wires. In order to get a realistic model of lifeline systems, the amount of tokens that can be transported are limited by the capacity of the lifeline component. A time delay is included to account for real life travel times. Cells are entities which perform a function. Cells required input of certain tokens in order to perform their functions and produce their output tokens. A hospital, for example, requires water, electricity, doctors, nurses and medical supplies in order to provide health services. Controls are needed for supplies and received tokens, (Chapter 3). Damage as the result of a disaster event affects the cells and channels ability to produce and transport tokens respectively. Damage to a cell affects its overall functionality and reduces the output of tokens regardless of the damage to its surroundings. Damage to the channels reduces the number of tokens they are able to transport and increases the time delay. Channel damage also affects the cell 54  functionality by reducing the number of input tokens. Controls will be used as decision points. 12. Define Human Readable Tables HRTs and Operating Modes. The damage assessment provides the functionality condition of critical buildings (cells), losses in pipelines or roads (channels), and the overall functionality condition of lifeline systems after the event has occurred. With the information gathered and developed with the damage assessment, then it is possible to translate the behaviour of the critical lifelines into Human Readable Tables (HRTs) and Operating Mode (OMs) that can be used in the model of region A in the simulator I2Sim. All information produced in the Damage Assessment Module (I2DAM) is stored in I2DB, (Chapter 3). 13. Setting up Databases. The data generated in the human and physical layers are aggregated into a database (I2DB). This database provides a common platform for data storage and is set up to feed the simulator and to receive the output of the simulation. The database updates the system state from this output for visualization and user interaction. 14. Prepare I2Sim. The final step is to prepare I2Sim for simulations by defining cells, channels, tokens, controls, HRTs and operating modes to start running pre-defined or real-time hazard events. 15. Evaluate Evident and Hidden Interdependencies. Using I2Sim, other hidden interdependencies can be revealed, in combination of the interdependencies discovered using the procedures described in Chapter 3. It is required that global and local authorities operate I2Sim with real time hazards or for test case scenarios. 16. Define Strategies to Mitigate Risk; Play Different Decisions Using the Simulator; and Define Action Plans for the “Study Space”. The results of the simulation can be viewed both statically and dynamically. Static visualization provides ―snapshots‖ of the state of the whole study space at certain moments in time. This is accomplished by mapping the data with a geographical information system (GIS) platform. Dynamic visualization allows the monitoring of individual  55  buildings or components in the study space, by plotting their functionality conditions as a function of time. One of the topics that are still under review is related to behaviour and psychological factors affecting disaster victims. These have been divided in: perceived vulnerability, panic, identity and family, grieving and social and antisocial behaviours. Perceived vulnerability evaluates the ―preparedness‖ of a community for disasters. Communities which are more prepared will respond better to a stressful situation and will be less likely to panic. Identity and family takes into account human behaviour regarding loved ones in disaster events. Social behaviours include community outreach and people helping each other while antisocial behaviour takes into account common problems like looting and civil unrest.  3.6 I2Sim Capabilities I2Sim has the capability to simulate the following emergency response situations:  3.6.1  Damage to Stadia and Crowd Egress Events can combine: a) Damage and injuries caused by the disaster directly, and b)  Injuries caused by egress of the crowd. After the victims have been taken out of the stadium, they are triaged and conducted to the care facilities. I2Sim‘s egress model considers both the victims due to the physical event plus the victims due to the crowd egress process. For an earthquake or other hazard event, there are assessment techniques to estimate the number of victims due to structural damage. For the crowd egress, I2Sim‘s new crowd egress model estimates casualties based on crowd density, guidance, rapid response, layout, and demographics. I2Sim‘s model for this situation is based on the concept of a primary factor: crowd density, adjusted by modifiers: guidance, rapid response, layout and demographics. Relationships between victims and density have been proposed in a number of references. I2Sim‘s concept of modifiers permits incorporating the experience of the responders in this type of situation.  56  3.6.2  Traffic and Street Crowds I2Sim can model these events based on the concept of density, similar to the stadia  crowd egress model, adapted to motorized vehicles as well as to pedestrian crowds. In the model, the vehicle‘s speed is associated to the number of vehicles occupying a certain road (bridges or streets) segment (density) as a primary factor and length (distance), adjusted by modifiers: guidance, rapid response, weather conditions, road closures and road damage. An example of guidance is traffic lights not working due to power failure. Another example of guidance is the number of traffic officers redirecting traffic due to closures. An example of rapid response is the time required to clear an accident situation. In the model, vehicles density is readjusted after road closures or openings. Pedestrian crowds behave in a way similar to stadia egress crowds and are affected by density and corresponding modifying factors.  3.6.3  Damage to Physical Infrastructure Evaluating the interdependencies among critical infrastructures (e.g., electricity, water,  gas, health, etc.), and the importance of coordinating the delivery of limited resources and prioritizing the restoration of services is a main capability of I2Sim. Events causing physical damage to the operating infrastructure will have cascading effects in the overall system response. Damage to physical infrastructure can have effects lasting from hours to days. These kinds of events go beyond incident response and require an optimum ranking of essential services. Incidents lasting days instead of hours may exhaust the reserve capacity of major entities. A larger magnitude earthquake would fall into this situation.  3.6.4  Damage to Human Infrastructure Damage to Human Infrastructure corresponds to the non-availability of people to  perform essential infrastructure functions. These include first responders, disaster managers, decision makers, and policy makers on the system response side as well as on the critical infrastructure side (e.g., operators of power system control centers, crews to perform repair jobs, doctors and staff in hospitals, etc.). Some events related to this type include: 57  a. Pandemics, e.g., H1N1 b. Maliciously induced events (e.g., water/food contamination) c. Targeted personal attacks on key personnel In I2Sim‘s current version, the effect of damage to human infrastructure can be modelled as reduced/slow operability of cells and channels. In some cases the degradation in operability can result in the ―closing down‖ of the function, e.g., of a hospital, and in this case, it has the same effect as a building being destroyed by an earthquake.  3.6.5  Egress and Traffic Models for I2Sim Egress and transportation models are based upon the physical characteristics of the roads  and waiting areas. For traffic models, the road characteristics are important: the number of lanes, the number of intersections, traffic lights or traffic signs. For egress modelling it is important the waiting area prior to the exit areas and the demographics of the crowd inside the venue. In an emergency, there are several entities that are affected: 1) the area where the impact has occurred (hazard zone, ground zero or epicentral area); 2) roads, streets, highways or channels that people and vehicles can move through; and 3) a location where evacuees (or other services or products) need to access (shelters, residences, emergency rooms or other zones). In a general sense, according to I2Sim ontology, (Martí et al 2008 and UBC-JIIRP, 2009) during an emergency, a token from cell 1 might be transported through a channel and its destination might be based in cell 2. Cell 1 might be a venue, stadium, a building, a warehouse; the token might be a human being seeking refugee or health services, or spare equipment for a critical building; and cell 2 might be a hospital or an emergency room, or a critical building. And there might be more than one way to go to cell 2. The transportation and egress model developed for I2Sim is unique from other models, because it incorporates the concept of interdependencies of critical infrastructures during disaster events. I2SIM-ETran model consider that other critical systems might be affected by a disaster, and the fact that casualties might be caused by the dynamics of the disaster. In other words I2Sim-ETran considers: 1) evacuation from a building, and the casualties that the egress 58  process might cause, and 2) the transportation of casualties or people moving by any means (vehicles or walking) from one location to another. The egress and transportation models share a lot of common ideas and assumptions; tokens need to be taken from one location to another. Consider a transportation problem in which a repair crew is sent from a workshop (cell 1) to the power house (cell 2) to repair a boiler, and the crew has to travel through a road (channel). The same concept will be applied for an egress situation, in a small scale, consider a stadium in which spectators will begin a normal egress after an event has finished, and they will leave their seats (cell 1), and they will walk through corridors and tunnels (channel) until they arrive to the lobby where the exit areas are located (cell 2). In a larger scale, consider a situation in which a whole community - (cell 1) will be evacuated due to a hurricane threat and they will travel through a highway - (channel) until they arrive to a temporary shelter (cell2). In this last example a transportation and an egress situation are combined, this reflects how related both processes are. In this research has been evident that physical and human factors play important roles. Those factors have a key role in egress and transportation: (at least 5 factors: Layout, Demographics, Rapid Response, Guidance and Electricity) will modify the density (people per m2) in an egress situation; and (at least 5 factors: Guidance, Rapid Response, Road Damage, Road Closures and Weather Conditions) will affect the volume of vehicles in a road for a transportation problem. In both cases the driven factor that will impact the delay of time and the number of casualties is the density. In this thesis, the reader has to realize that ―density‖ refers to: the number of people per square meter for egress process; and in transportation problems is the number of vehicles in the road. It is not the purpose of this thesis to exhaust and thoroughly explain these models. All the information will soon be issued in two papers, that have been prepared along with this thesis, that will explain the concepts, assumptions and metrics developed for i2Sim-ETran models and the egress and the transportation models, Rostamirad et. al. (2011) and Kankanamalage et. al. (2011).  59  3.7 I2Sim Functionality The following types of analysis can be performed using I2Sim‘s functionality: 1. Model a study space and a set of Critical Infrastructure Systems (CIs). 2. Discover factors that have a significant impact on CIs. 3. Stress CI capabilities and limitations, including human and organizational factors where appropriate. 4. Define a set of events that will stress several CIs simultaneously, and investigate how they affect one another. 5. Design multiple scenarios, with several inputs and outputs comprising main scenarios and sub- scenario sets. 6. Organize a set of scenarios that allows the analysis to cover or sample the problem space. 7. Create a base of approved scenarios reflecting the agencies‘ objectives, and thus the required spectrum of responses needed to cope with the events. 8. Design scenarios prior to the study, and revisit the scenario several times to adjust the parameters. 9. Assess the impact of the information and hypotheses of the threats. 10. Assess the impact of CI aspects within the problem definition. 11. Identify and document key scenario assumptions. 12. Judge the applicability and accreditation of the models. 13. Tailor the detail of models and scenarios to the level of capability of all participants.  3.8 Final Remarks on UBC-JIIRP New urban sites, new or evolved technology uses new Critical Infrastructure systems The development of new urban sectors can pose major problems for evaluating Risk. New and upgraded construction technologies and new infrastructure systems are being designed and built. Therefore new developed interdependencies among Critical Infrastructures (CIs) are being generated, these interdependencies are complex and hard to evaluate. 60  Privacy of information and lack of tools to indentify vulnerabilities and risks Organizations are very reluctant to expose the internal details of their operation for fear of losing competitive advantages or exposing vulnerabilities. There is a lack of tools to clearly identify and quantify how risks and vulnerabilities affect other infrastructures. In this project it was possible to use the information available, to properly assess the risk of each individual CI, and; to be able to show and rank the interdependencies among them; and still be able to keep the information private for each specific stakeholder. Information Not all information is important. In some cases it will be important to consider airspace, or cyberspace, in other cases these or other elements might not be relevant. All the planning for the Interdependency of Critical Infrastructures cannot take place in isolation, it requires multi agency planning so that objectives, assessments and outcomes will be properly matched. Exchange of information might become relevant in the future; nowadays the internet shows a lot of information from different systems (CIs). The system owners can be ahead of their competitors in terms of assessing their vulnerabilities, and reduce their risks. Exposing the weaknesses may in fact lead to more resilient systems. Being the weakest link in the chain, might translate in difficult times for business purposes, and unplanned reaction during emergency events might cause more damage or economic loss to the system than revealing information to other competitors. Within the UBC-JIIRP project, and particularly for the UBC Test Case, six lifeline systems were considered. During the information gathering and collection (interviews and interchange of relevant data), interviews with managers, operators and technicians showed the importance of those lifeline systems for the well being of UBC. The information of the different assets is relevant, because it will help reveal interdependencies and interconnections; and it is also important to know which components and systems are being connected and which services or products are being exchanged among them. Details from all assets within a region will help in the overall assessment of interdependencies.  61  Acquiring all the information is a challenging task: first, convince the operators that the information they have is relevant for the well being of an entire region; second, the process of selecting relevant information is tedious and sometimes a slow process, because there are main and secondary assets; third, selecting which connectivity is relevant to show in the different sectors of a region; and, four, define the amount of services and products being consumed within the region. The information collected has inherent difficulties. Sometimes information is not precisely known by the operators, or the way they have the information needs to be translated in a ―common language‖. The process to produce readable data can be quite frustrating. Many times, educated guesses need to be made, and discussed with the operators of the different assets. Once the information is available, it is important to prepare the whole system in order to be able to obtain reasonable outcomes with the acquired information. In the project it was very important to define and understand the vulnerabilities, the Interdependency of Critical Infrastructure, the hazards, and to have proper information. Several interviews and questionnaire formats were developed along the project; as the methodology and the simulator (i2Sim) were being developed, so was the necessity for detailed information. It is not the purpose of the thesis to exhaust the discussion on information, data collection and data enhancement within this research, but to give the reader an idea of the problems, techniques to use for gathering data and information for similar projects. Some of the related issues on information (interviews and questionnaires) can be found in the website (www.i2sim.com) and the related reports of the UBC-JIIRP project. I2Sim A key objective of the authorities of the study space is the protection of the identified Critical Infrastructure from human and natural hazards, and to minimize the consequences on an emergency event. In order to accomplish all that, it is important to understand the vulnerabilities, the Interdependency of Critical Infrastructure, the hazards, and to have proper information.  62  Common language is essential to understand the different techniques and methodologies used by interdisciplinary teams. Exchange of information and understanding of the units that are being exchanged by these systems are key elements for the overall assessment of the system. The wellness of the society will be preserved when the consequences of a certain event are accurately predicted and defined. Analyzing complex systems such as those formed by infrastructure networks and decision makers requires a multidisciplinary holistic approach. On the data management side, the UBC-JIIRP project facilitates centralized data management. We also envisioned the need to manage the data in a distributed way and deploy a data integration system to coordinate the data sources. The interdependency study usually incorporates a large number of data sources, which are owned and maintained by different parties. Usually data is requested by the interdependency study, and hence the data should be updated by the owner; the user of I2Sim just defines data enhancement so that raw data could be transformed and delivered efficiently at request. A distributed system is usually more robust than a centralized system. Consider that in UBC-JIIRP we study large scale natural disasters such as earthquakes which may also destroys information infrastructure, we need to make sure the disaster management system keeps working in a condition that some of the data sources become unavailable or suddenly become unavailable. The development of the I2Sim simulator also impacts the data management. It cannot be taken for granted that the database is just a replacement for input files, and that interfacing with database services could be easily done after testing the simulator with ―dummy‖ input files. This can be misleading, and the design and development of the simulator might be compromised. The simulator would normally be tested with ―dummy‖ input files in its early stages, but it should be tested with the database once a full workflow needs to be run. The consequence of neglecting the database schema is that the simulators miss opportunities to optimize their operations according to appropriate database access patterns. Moreover, the re-interfacing workload becomes heavy and the developers face severe time-effort challenges.  63  I2Sim capabilities range from individual infrastructure systems to regional assessments. This can help in the development of action plans when hazards occur during large public events, like Olympics or World Cups, or during regular public activities. It is possible to observe the performance of individual systems. Stakeholders can predict the result of having vulnerabilities in the system, and how to cope with them once the hazard has developed. With this tool they can enhance the resiliency of the system, and can also share valuable information with other stakeholders Information Sharing Challenges in UBC-JIIRP project When the JIIRP project started, we had aimed at developing a case study focusing on an area larger than the UBC campus. We had envisioned the possibility of modeling the city of Vancouver or maybe even go to larger scale. We soon realized that this task would be too ambitious, not only because of the complexity of the data and the interconnection of infrastructure networks at the city level, but because the data would just not be available to us. We approached the project‘s partners and obtained a different range of responses to our data requests. It was then that the team decided to focus on the UBC as the ―study space‖. Getting access to the UBC infrastructure databases was less difficult, although not free of challenges.  Data sharing in the JIIRP-UBC project JIIRP-developer –Confidentiality agreement (?) –Specification of the use that will be made of the data –Negotiate support from data operator‘s supervisors –Approach at three different organizational levels  Data Owner –Incomplete data –Data not compatible –Multiple versions –Fear of competition, loss of competitive advantage if data is shared –Important data still in owner‘s head –Security concerns (misuse)  Figure 3.6 Data Sharing in the JIIRP-UBC Project  64  Any project that deals with sensitive information will face data sharing issues. This project was no exception. In Figure 3.6 we represent the relationship between the two main participants involved in the project, the data owners and the developers, and the main challenges and concerns associated with them. On the data owner‘s side, the various data issues are presented here with some examples:   Policy issues: Different information sharing policies among the campus planning office, the utilities office, and the IT office. The owners of the data had different levels of concern towards sharing it. Some immediately shared the data, others waited for an approval from higher authorities, a third group feared misuse and were reluctant to share.    Technical issues: Although most of the data on infrastructure networks on campus is maintained in digital format (AutoCAD drawings), we found that it is redundant and inaccurate in many cases. We also found different versions of the same data across departments. Every time a modification is made to the conduits, the AutoCAD maps are updated in the department responsible for the modification. The new version of the file is then shared to other departments. This obviously constitutes a problem when two departments make changes simultaneously. One other issue regarding the infrastructure data is that it is ―flat‖. The CAD drawings do not have a database associated with them. All dimensions, materials, directionality, etc., are stored as annotations on the map, this makes it impossible to use the data for analysis.    Cultural barriers: We also faced some difficulties when obtaining data due to cultural barriers. Communication was difficult with some of the data owners, they saw the project as a merely academic exercise with no particular application to their work. They could not see a direct benefit from sharing the data. Concerns about the security of the UBC community were also raised by some of the data owners if the data was misused.    Non-systematized information: Although most of UBC‘s infrastructure data is systematized in some way, many of the day-to-day operations ‗know how‘ is not documented: the weak points in the networks, the way some of the most common 65  issues are solved every day, the communication issues between departments, the reliance of the quotidian operation on a particular external asset or procedure. All this information is very valuable and is generally not documented. It is in the data owner‘s (the operator‘s) memory.   On the developer‘s side, we also learnt some important lessons when it comes to establishing trust with the data owners. We identified four crucial factors to obtain the data owners‘ cooperation:  1. A confidentiality agreement, or some sort of written arrangement, in which the sensitivity of the data is recognized and in which the limitations of use and publication are specified. 2. A clear understanding on the developer‘s side of the use that will be made of the data. After several interviews with data owners on campus we realized that we could engage them in an agreement to share the data if we had a very clear idea of what we would be doing with their data, and we could communicate that effectively to them. This sounds evident, but as in many projects of the same nature, the data gathering process was done in parallel to the conceptual modeling, and in some cases even prior to the modeling. We knew we would need the data at some point, we just did not know to what extent and when. Not being able to give this reassurance to the data owners probably slowed down the process of accessing the data. 3. Support from higher authority levels. Having written or verbal approval from higher authorities is something that was needed in a couple of cases in order to be able to even sit and talk about data sharing. This, however, was no guarantee that the data sharing would occur. 4. Approaching a department at three different levels. We identified broadly three levels of responsibility profiles that had to be engaged with the project: the higher authority a department director (which in many cases does not know about the data) is the person signing the official agreement; the department manager who needs to be convinced that the agreement will be beneficial for her work and thus should cooperate; and the data operator, who is the person always in contact with the data and the one that knows the flaws and has most of the day to day know-how 66  when it comes to data maintenance. This person needs to be convinced that the flaws will not be publicized and that her/his job is not on the line.  Seismic Risk Assessment (SRA) for Critical Infrastructure Systems Quality of information will determine the refinement of the damage estimation methodology. A deterministic approach was used in this project, because the methodology was focused in producing the damage estimation methodology and a way to rank and evaluate interdependencies of CIs. The number of essential utilities will also define the scope of the SRA methodology to be used. The physical layers were mapped and laid out with all the details of the systems. All the information mapped has to be accompanied by damage estimation techniques. Hazards play an important role to discover interdependencies and cascading effects in CIs. The SRA methodology for lifelines and the implementation of other hazards were used. It was possible to estimate casualties and loss of function for the essential survival utilities that a region needs to survive after a hazard event. Estimating the damage of each layer, it was possible to define internal and external interdependencies for the CIs. These pre-defined interdependencies were used to feed the simulator. Outcome information was then translated in a ―language‖ that can be read by operators and other members of the UBC-JIIRP team. It is important to say that operators do not need earthquake engineering jargon, they need to know the number of casualties, and which level of injuries they have; and the loss of the essential utilities that the critical buildings will face after an event; and the consequences of losing part or total essential utilities in other CIs. Connection and interdependency indices had to be developed. UBC Campus was selected as a test Study Space as described in Chapter 5. To validate the methodology it is important to test it with real cases; to have meetings with the stakeholders to see if results are close to reality; and in other instances to prove the methodology with real 67  past events. It was recognized that having a smaller test case to probe the methodology would help in the development of a better simulator. All these conclusions apply to I2DAM module and Seismic Risk Assessment was used to meet the requirements for Tasks within UBC-JIIRP project; and the deliverables: (1) Simulator of Interconnected Infrastructures, (3) Analysis of Case Study scenarios, (4) Resources Integration and Data Visualization at System Control Centres, and (5) Integration of GIS Resources from Dissimilar systems. Further details and related information for UBC-JIIRP can be found at UBC-JIIRP website: www.i2sim.ca.  68  4 Interdependencies 4.1 Introduction There are at least three phases in a disaster: planning, response and recovery. All measurements of interdependencies, if any, are static measurements that do not take into consideration the dynamic changes of events in time during disasters. For example, a moderate to major earthquake will provoke a main shock that will render some critical infrastructure systems (CIs) to a compromised functional condition. After the main shock, landslides will be triggered a few minutes later, fires might be ignited due to breakage at gas pipelines, a tsunami can also be triggered, and an aftershock will hit the same epicentral area a few minutes later. The study space‘s functional condition will be different during the main shock and the aftershock. Even though some methodologies take into account the fact that a disaster impose dynamic changes in the sates of the CIs, the methodologies only address the preparation and planning processes of a disaster. These methodologies, however, partially consider the real-time dynamically evolving conditions of a disaster, Dueñas-Osorio et. al. (2007a), Chang et. al. (2007), Chang et. al. (2009) and Tabucchi et. al. (2009). The methodology that I propose in this research will bridge this gap by measuring the functionality conditions of the study space at different times during the disaster. It will also take into consideration the interdependencies of critical infrastructure systems; and therefore this methodology is unique and will contribute to help in the three phases of a disaster. In this chapter I will propose different ways to measure interdependencies: Single and Global Interdependencies. I will also define Interdependency Indices and Importance Factors; all these values will help create a Resiliency metric for the system. It is clear that interdependencies need to be evaluated during a disaster, so that recovery processes are in agreement with the objective functions of the managerial levels. Some examples of measuring interdependencies will be shown in this chapter.  69  Interdependency matrices have been proposed in the recent years, Alexoudi et. al. (2008) proposed an ―inoperability matrix‖; Turata et. al. (2008) also proposed interdependency structure matrices; and finally Martí et. al. (2005) proposed an interdependency matrix. In any given hazard, for example after a moderate to large earthquake there is always a large amount of damage on CIs and therefore limited resources in the affected area. There are prevalent power outages, road closures, and some of the utilities are cancelled until damage assessment inspections are finished. There are at least two questions that need to be answered after a moderate to large hazard event has occurred in a region: what will be the overall functionality conditions of a ―study space‖, and where shall the remaining available resources be sent to? This methodology will advance a way to measure Interdependencies and Resiliency in a region or in a system. This methodology will be embedded in the Infrastructure Interdependency Simulator tool (i2Sim) developed by the Critical Infrastructure System (CIS) group of UBC. The simulator will be used to reveal, in a span of time, all the evolving changes in a region whenever a hazard event occurs. Events along with decisions to bring a system to a complete recovery can be simulated with such a tool, several tables and calculations were proposed to reveal interdependencies, interconnections, and consequences of having functional, physical and resource problems. Chapter 4 in this thesis will explain the UBC-JIIRP project, and Chapters 5 and 6 will demonstrate the use of the methodology developed and the simulator i2Sim. In this chapter, I will present a general explanation of the simulator and its ontology.  4.2 Interdependencies 4.2.1  Directionality and Connectivity for Lifeline Systems Directionality issues are complex; for explanation purposes for the connectivity of  pipelines carrying liquids (water, oil, sewage, storm, etc), consider Figure 4.1, and the segment of pipeline P-100. In order to define the loss in volume of water, it is essential to estimate the damage and the corresponding water volume losses due to the hazard being studied (i.e. an earthquake). Once the water volume loss has been obtained for each segment, the total water 70  volume loss throughout the pipeline needs to be evaluated. The directionality of the liquid flowing through the pipelines imposes a complex connectivity and interdependency issue for pipelines. It will be interesting to know the water volume loss at the beginning and at the end of segment P-100, for example. From now on the water volume loss will be referred as loss. In the pipeline shown in Figure 4.1, all the losses have been obtained for the pipeline segments (with different diameters and materials). The four segments considered (blue, orange and red - P-100) have 10%, 5% and 2% calculated losses. The loss in segment P-100 is estimated as 17% of the total volume distributed through these pipelines; at the beginning the accumulated loss is 15% and at the end the loss reaches 17% in pipeline P-100. To account for the losses throughout the pipeline, all the segmented volume losses have to be summed up, and they have to be estimated at the origin, and towards the direction of flow. If the loss of volume has been calculated for a given water pipeline in this way (vloss); and the volume of water served through this pipeline is known (V), it will be a straight forward calculation to account for the loss in water volume (vloss x V).  85 % - 2 % = 83 %  2 % loss  Segment analyzed P-100  100 %-15% = 85 % 5 % loss  10 % loss Figure 4.1 Assessment in Pipeline Segment P-100 71  All the pipelines carrying liquid will be assessed with this procedure. For pipelines carrying gas liquids (steam or gas), a leakage is considered a failure, especially for gas pipelines; therefore this assessment will not be performed for gas or steam pipelines.  4.2.2  Single and Global Interdependencies Along this research it has been clear that Critical Infrastructure systems are like  clockwork machinery that is connected to other CIs. It is therefore evident that interdependencies have different levels. Single Interdependency Functionality is defined as the functionality condition of one critical system, taking into consideration all components that produce or modify tokens or production units. Global Interdependency Functionality is defined as the functionality condition of the study space, taking into consideration all the CI systems needed for the functionality conditions of the region. Figure 4.2 shows a study space (region A) with 3 Critical Infrastructure systems (Water, Electricity and a Hospital) that are needed to maintain normal life conditions for the population in region A, along with one block representing ―other systems‖. The CIs are considered as physical layers, and three of them are emphasized, the fourth one represents other systems such as commerce, educational systems, etcetera. It also shows five components of the Water System, three of them are specialized buildings, and the other two are pipelines. All of them will define the functionality conditions of the water system. Figure 4.2 also emphasizes that damage estimation is different for buildings (cells) and the corresponding pipelines (channels). Nevertheless, once the damage estimation has been obtained, then all 5 components are weighted and they will define the overall functional condition of the water system, which is defined here as Single Interdependency. The Single Interdependency is obtained for all the physical layers of Region A. If Single Interdependency is defined, and then weighted and combined, then the overall functionality condition of region A can be defined as well, this research defines it as Global Interdependency. 72  Actual methodologies will deal with planning, organization and assessment of one critical infrastructure system; it is not the objective of the managers of the CI to include other systems in the evaluation, and therefore the functional conditions of all other systems are supposed to be 100% functional. In normal conditions this might be a good assumption, but in a disaster event this assumption might overlook real problems. As an example, consider the following Water system, Figure 4.3. This system has a Reservoir, one transmission line (trunk water pipeline TL01), one pumping station, three distribution lines that are distributing water throughout the region (distribution line DL A12 is emphasized in this example), and finally the zone that is receiving the water supply (zone A12). This is the water system that is also conceptualized in Figure 4.2. In the following pages Single and Global Interdependencies will be explained, using the water system described before, and other systems for Region A.  4.2.3  Single Interdependency Functionality (SIF) In order to calculate the SIF, every component in region A should be isolated. Consider  that water consumption for Zone A12 is being investigated, after a disaster has happened, Figure 4.2 and Figure 4.3. The water system in Figure 4.4 is comprised of one Reservoir, one transmission trunk line (with three pipeline segments), one pumping station, one distribution pipeline (with two pipeline segments) and Zone A12, in which critical buildings might need water supply (e.g hospital, residences or commerce). But first the components of the system are assessed separately. All of this is a well known procedure for every CI system and also for the damage estimation techniques, whenever the hazard is an earthquake. In order to assess the amount of water received in Zone A12, functionality-hazard relationships need to be developed. The relationships can be as complex as the user of the methodology wants them to be, but it must be kept in mind that granularity plays an important role in the overall solution of region A. The reservoir and the pumping station are assessed taking into account the structure, the equipment, pipelines and contents necessary to distribute the water throughout the zones in the study space. The transmission and distribution pipelines are assessed also according to their characteristics and the soil in which they are buried; liquefaction effects or permanent ground displacement is related to the functionality conditions of the pipelines. Once all of those 73  calculations have been realized then the single interdependency of the system can be formulated. Single Interdependency is in reality the functionality condition of the water system that supplies water to Zone A12. The equation defining the SIF of the water system can be conceptually defined as follows:  Remaining Functionality Condition of the Reservoir, after any given hazard  + Single Interdependency or Functionality Condition of the Water System  Remaining Functionality Condition of the Transmission Pipelines, after any given hazard =  +  Remaining Functionality Condition of the Pumping Station, after any given hazard + Remaining Functionality Condition of the Distribution Pipelines, after any given hazard  74  Study Space  Physical Layer  Critical Components  Damage Estimation  Single Interdependency Functionality  Global Interdependency  (SIF)  (GIF)  Structure  SI of Water System  Reservoir NSCs Water System Electrical System  Region A  Hospitals  Transmissio n Lines  Building Contents  Pumping Station Channel Structure  Distribution Lines  + SI of Electrical System  Single Interdependency taking into account all 5 critical components of the water system  Loss and Delay Other Systems  Functionality  + SI of Hospitals  + SI of Other Systems  Critical Buildings  Figure 4.2 Methodology to Determine Single and Global Interdependencies  Reservoir  “Study Space” Region A  Transmission Lines TL01 Distribution Lines  Hospital Zone A12  Pumping Station  DL A12  Figure 4.3 Water System of Region A 75  Reservoir Structure NSC damage 1  damage 2  damage 3  Pumping Station Structure NSC  damage 4  damage 5  Zone A12  Figure 4.4 Water System or Zone A12 It is up to the manager of the water system to decide the weight or importance of every component in the system. It is clear that the water system is somehow reliable and resilient, that with small losses in the system, it remains functional. It is a common design procedure to help the system distribute water through gravity, and even though pumping stations are non functional, the water is still flowing through the system. Whenever there is decay in the water pressure, it is hard to notice in low-rise buildings. Medium-rise and tall buildings will notice the low pressure. Important critical buildings might need a constant pressure to perform its function, which is the case for hospitals or health care facilities. It is not uncommon to find hospitals which service dialysis patients in upper floors, and those systems need a constant water pressure. That is why the internal functionality factors should be decided by the managers of the water system in conjunction with the health care managers. In a general sense, any system can be conceptualized as shown in Figure 4.4. All critical buildings or structures might be presented as Cells and Channels, according to I2Sim ontology. Critical infrastructure systems might be characterized as parallel or serial systems. As an important conclusion, SIF deals with all the components in a serial Critical Infrastructure system.  76  Interdependencies cannot be evaluated without assessing the damage in the components of the CI. Special attention should be devoted to critical buildings (pumping stations, power houses, etc) and their electrical and mechanical equipment, pipelines and building contents necessary to perform their operations. And transmission lines, such as conduits, pipelines, aqueducts, channels, etc.; the soil where they are buried is crucial for the overall behaviour of these components, as well as their characteristics (diameter, materials, and year of construction), so that performance of the system during a hazard event can be easily characterized. Damage and directionality conditions should be characterized as well. Figure 4.5 shows three cells and two channels, but in fact the system might have more elements. SIF will be assessed by individually evaluating all 5 components, but directionality conditions and average attributes of the components need to be investigated.  Cell 1 Channel 1  Cell 2 Channel 2  Cell 3 Figure 4.5 Critical infrastructure System, Local Damage assessment estimations are performed as usual, all critical buildings are characterized as cells and a building-prototype is given to all of them; damage-hazard relationships are defined for all prototypes. Non-structural components (NSCs) and building contents are evaluated in the same procedure. In this case the functionality of the critical 77  infrastructure (cell) is obtained, and it should be considered as a function of the behaviour of the structure and NSCs, Equation 1: Equation 1. Functionality of cell i  1)  ; these factors are decided by the operator of the CIs.  is the percentage  Where:  in which the structure is a key element for the functionality of the facility; the same is applicable to the other two parameters ( and ), that show the percentage in which the NSCs and the building contents are important for a given facility. Consider a pipeline that delivers water through 1,053 m; those pipelines are usually made of different materials (PVC, asbestos cement, ductile iron, etc) and they can also have various diameters. Therefore, the functionality condition of a distribution or transmission pipeline, or in a general sense – a channel, is comprised of its different segments‘ functionality. Channel‘s functionality condition is a function of the several segments of the transmission or distribution pipelines, Equation 2: Equation 2. Functionality of channel i  ( 2)  There are Critical infrastructure systems that have trunk lines, pipelines or conduits that are buried underground. Depending on the material that is being transported in those lines, there are systems that are:  78    Stable without safety concerns unless pipelines are seriously broken, e.g. electrical conduits    Dangerous systems in which leakage can cause cross contamination or the spillage can be potentially harmful to the environment, e.g. gas, sewage, oil, petroleum pipelines, and    Systems in which leakage does not compromise the functionality of the system or the environment, but the rupture or accumulated loss might render the system non-functional, e.g. water, storm, steam pipelines, or they can cause damage to other pipeline systems.  With these equations, Equation 1 and 2, SIF can be calculated for a given system. An example on how to calculate SIF for a water system is presented below.  4.2.4  Example on SIF of the Water System Figure 4.3 and Figure 4.4 show the general characteristics of the water system that  distributes water to the zone with a hospital building. In this example, one Power House and one Hospital will be considered, Figure 4.6, where pipelines‘ location and other information are shown. These CIs were modelled in the UBC Campus Case - that will be discussed in Chapter 5.  79  Figure 4.6 Water system - Power House to Hospital The first step is to acquire all relevant information on the water system; the second is to collect information on the hazard threat and assess it; the third one is to obtain the associated damage to the system. All of these steps are discussed in Chapter 3 – JIIRP General Methodology. In this example the general characteristics of the distribution water pipeline, providing water from the Power House to the Hospital, are shown in Table 4.1 Table 4.2 and Table 4.3 show the functionality percentages of the Power House for an earthquake hazard (Instrumental Intensities II VI to II XII). The Power House at UBC Campus Case is comprised of two buildings in which pumps and related equipment reside. Results from the functionality conditions are shown in these tables. The earthquake hazard was characterized using the instrumental intensity, Atkinson and Wald (2007) and Atkinson and Kaka (2007). In these tables: S, Structural damage; DS, Drift Sensitive Components; AS, acceleration sensitive components; and BC, building contents.  80  Table 4.1 General Characteristics of the Distribution Pipeline (PH to H) Pipe ID  Segment length (m)  Directionality  Diameter (mm)  Material  P-1571 P-1572 P-1570 P-4181 P-4180 P-4190 P-4220 P-4200 P-35 P-4770 P-4590 P-4490 P-4580 P-53 P-49 P-48 P-4360 P-4340  62 100.9 78.6 75 157.3 100.1 15.7 245.8 94.5 95 4.6 2.7 154.4 55 78.3 31 10.4 100  1 2 3 4 5 6 7 8 9 10 11 12 13 P35-10 P35-11 P35-12 P35-13 P35-14  NA NA 305 NA 337 337 209 337 438 530 530 337 530 199 255 150 305 211  NA NA Ductile Iron NA Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron PVC Ductile Iron Ductile Iron  Table 4.2 Functionality Conditions of Power House Intensities VI to IX Bldg % PH 1 %PH 2  S 95 100  II VI DS AS 50 80 50 80  BC 80 80  S 80 90  II VII DS AS 50 80 50 80  BC 80 80  S 70 90  II VIII DS AS 50 80 50 80  BC 80 80  S 50 80  II IX DS AS 50 80 0 80  BC 80 50  Table 4.3 Functionality Conditions of Power House, II X to XII Bldg  II X  II XI  II XII  S  DS  AS  BC  S  DS  AS  BC  S  DS  AS  BC  %PH 1  0  50  80  80  0  0  80  80  0  0  80  80  %PH 2  60  0  50  50  0  0  50  50  0  0  50  50  Table 4.4 shows the losses in the pipelines carrying water from the Power House (PH) to the Hospital (H) after the damage estimation has been calculated. The Water Station of UBC Campus resides on PH. The first column shows the pipeline identification number as used by the operators; the second column shows the directionality of the water flow (going from 1 to 13), in pipeline P-35 there is a bifurcation and water also flows to segment P-53. The rest of the columns show the water leakage as a fraction of the total volume. 81  It is worth noting that in the loss of water due to leakage or breakage in these pipelines, there are two calculations: (seg), which stands for losses calculated in the segment of the pipeline, e.g pipeline P-4181, at Instrumental Intensity VIII, shows that 3 % (0.03) of the water volume will be lost in that segment; and the (acc) accumulated loss of water volume will be 12 % (0.12), because all segmented losses from pipeline P-1571 to P-4181 are accumulated. Pipeline P-4181 has a segment length of 75 m, and the accumulated length is 316.5 m, Table 4.1. The accumulated loss is due to the directionality conditions of the pipelines. Table 4.4 Losses in the Water Pipelines (Water Station to Hospital Distribution Lines) Pipe ID  Dir  P-1571 P-1572  VI  VII  VIII  IX  X  seg  accum  seg  accum  seg  accum  seg  accum  seg  accum  1  0  0  0.01  0.01  0.03  0.03  0.13  0.13  0.57  0.57  2  0  0  0.01  0.02  0.05  0.08  0.21  0.33  0.92  1.49  P-1570  3  0  0  0  0.02  0.01  0.09  0.05  0.38  0.22  1.7  P-4181  4  0  0.01  0.01  0.03  0.03  0.12  0.15  0.54  0.68  2.39  P-4180  5  0  0.01  0  0.03  0.02  0.14  0.1  0.63  0.43  2.82  P-4190  6  0  0.01  0  0.04  0.01  0.16  0.06  0.69  0.27  3.09  P-4220  7  0  0.01  0  0.04  0  0.16  0.01  0.7  0.04  3.13  P-4200  8  0  0.01  0.01  0.04  0.03  0.19  0.15  0.86  0.67  3.81  P-35  9  0  0.01  0  0.05  0.01  0.21  0.06  0.91  0.26  4.07  P-4770  10  0  0.01  0  0.05  0.01  0.22  0.06  0.97  0.26  4.33  P-4590  11  0  0.01  0  0.05  0  0.22  0  0.97  0.01  4.34  P-4490  12  0  0.01  0  0.05  0  0.22  0  0.98  0.01  4.35  P-4580  13  0  0.01  0  0.05  0.02  0.24  0.1  1.07  0.42  4.77  P-53  P35-10  0  0.01  0  0.05  0.01  0.21  0.03  0.95  0.15  4.22  P-49  P35-11  0  0.01  0  0.05  0.01  0.22  0.05  1  0.21  4.43  P-48  P35-12  0  0.01  0  0.05  0  0.23  0.02  1.01  0.08  4.52  P-4360  P35-13  0  0.01  0  0.05  0  0.23  0.01  1.02  0.03  4.54  P-4340  P35-14  0  0.01  0  0.05  0.01  0.24  0.06  1.08  0.27  4.82  For the evaluation of the output condition of the Power House, Equation 1 will be used, in this case the following NSCs were considered: electrical and mechanical equipment and pipelines. In this case  is considered as 40 %, and  was considered as 60 % (15% for electrical and 30% for mechanical equipments; and 15% for pipelines inside the Power House). In this case, the contributions for building contents were considered as not relevant for the operation of the Water Station. This equation was developed with previous experiences shown in California in FEMA 224 (1991), and meetings with the operators of the Power House. Hence, Equation 1 can be written as: 82  Functionality of the Power House = 15 % (electrical) + 30 % (mechanical) + 15 % (pipelines) + 40 % (structure) Table 4.5 shows the functionality values for the Power House. For Instrumental Intensities XI and above the functionality of the Power House will be considered as zero. Table 4.5 Functionality Conditions at the Power House, after Equation 1 VI  %  VII  VIII  IX  X  XI  F  Tot F  F  Tot F  F  Tot F  F  Tot F  F  Tot F  F  Structure   = 40  0.95  0.38  0.8  0.32  0.7  0.28  0.5  0.2  0  0  0  0  Mech Eq   = 30  0.8  0.24  0.8  0.24  0.8  0.24  0.8  0.24  0.8  0.24  0.8  0.24  Elec Eq   = 15  0.8  0.12  0.8  0.12  0.8  0.12  0.8  0.12  0.8  0.12  0.8  0.12  Pipes   = 15  0.5  0.08  0.5  0.04  0.5  0.02  0  0  0  0  0  0  0.82  0.72  0.66  0.56  0  Tot F  0  100%  Functionality condition s  90% 80% 70% 60% 50% 40%  Structure  30%  Mechanical equipment Electrical equipment  20%  Pipelines 10%  Power House  0% VI  VII  VIII  IX  X  XI  Instrumenta l Intensit y  Figure 4.7 Functionality Conditions of the Power House Figure 4.7 shows the functionality conditions of the structure and non-structural components of the Power House. It is evident that the behaviour of the Power house is related to its structure performance. The mechanical and electrical equipment are stable and reliable, but 83  pipelines might be compromised by the level of shaking and displacement. Nevertheless, the pipeline ruptures inside the Power House can be repaired immediately. Whenever the structure in the Power House suffers heavy damage, all operations will be compromised, therefore the Power House will be non functional for instrumental intensities X and above. The functional conditions are the contributions of all key components in the system. Structural conditions after an earthquake will not reflect the capacity of the system to provide water to a region, several components need to be assessed independently, and later on their interdependencies shall be considered. Once the information has been collected and arranged, and the damage estimation has been performed for the water system it is possible to obtain system Functionality curves. Figure 4.8, shows the functionality conditions of the water distribution in Zone A12, due to the losses in the four components: reservoir, transmission line (TL 01), Pumping Station and distribution line (DL A12), Figure 4.3. It is interesting to note that for this case the losses in the reservoir and the pumping station have greater effects on the overall loss for Zone A. The pumping station is the most critical asset in this system.  100% 90%  80%  Functionality condition s  70%  68%  60% 51%  50% 40% Reservoir TL 01 Power House DL A12 Zone A12  30% 20% 10%  25%  0% VI  VII  VIII  0% IX  0% X  Instrumenta l Intensit y  Figure 4.8 SIF of Water Distribution in Zone A12 In Figure 4.8 the volume of water distribution in zone A12 is obtained through a simple equation: 84  As shown previously, it is considered that a given volume of water is sent to zone A12; the volume sent will be compromised by the damage sustained and the corresponding losses of: the Reservoir (Rloss), the transmission lines (TLloss), the Power House (PHloss) and the Distribution Lines (DLloss). Figure 4.9 shows that transmission and distribution lines are stable for instrumental intensities VI to VIII, their functional conditions will be 100 % to 70 %. There will be no water supply for instrumental intensities IX to XII; and therefore extreme measures have to be exercised in this area, so that water could be restored in the region. The SIF of the water system at zone A12 is different from the functionality conditions of the structure of the Power House; the functionality of the water system is affected by the performance of different components: its structure, non-structural components, and other key components (pressure stations, water treatment plants, and transmission and distribution pipelines).  1 0.9 0.8  Functionality condition s  0.7  68%  0.6 51%  0.5 0.4 0.3  25% 0.2 0.1  Structure Power House Zone A12  0 VI  VII  VIII Instrumenta l Intensit y  0% IX  0% X  Figure 4.9 SIF of the Water System at Region A12 Figure 4.9 shows the SIF of the Water System at region A12; the structure, the behaviour of the structure along with non-structural components (Power House), and the SIF - taking into 85  account the Reservoir the transmission and Distribution pipelines and the Power House. It is evident that the seismic performance of the Power House‘s structure is closely related to the facility‘s equipment. For an Instrumental Intensity VI, the structure has 95% of functionality, but due to non-structural behaviour the functionality conditions may drop to 82%; but if we consider all the elements that are needed to distribute water from the reservoir up to region A12, then the remaining functionality conditions will be close to 70%. Figure 4.9 also shows that water service in the region is possible after instrumental intensity earthquakes VI to VIII, even though ―boiling water advisory‖ might be issued by the managers of this region. For an instrumental intensity IX and above the water service will be compromised unless preventive actions are taken to improve their functionality conditions.  100%  100% 90%  90%  99%  97%  95% 90%  85% 90%  Functionality Conditions  80%  80%  84%  70%  66%  60%  64%  59%  56%  50%  38%  40%  30%  30%  34%  Electricity to H  20%  Substation  10%  Elec Conduit lines  24%  0% VI  VII  VIII IX Instrumental Intensity  X  XI  Figure 4.10 SIF of the Electrical System in Region A12 Figure 4.10 shows the SIF of the Electrical System in Region A12. The functionality condition of the electrical system is closely related to the seismic performance of the SubStation.  86  4.2.5  Global Interdependency Functionality (GIF) Figure 4.11 illustrates the concept of GIF and shows that summation of all SIFs of  Critical Infrastructure Systems in the region, will add up to define the Global Interdependency Functionality. The manager of the region should have good understanding of the CIs that will define the resiliency or functionality conditions of the system.  Other utilities  Personnel  Steam  CELL Gas  Other utilities  Structure Electrical equipment Mechanical equipment  Water  Other equipment Furniture  Electricity  Customers  Water and related utility pipelines etcetera  Figure 4.11 Global Interdependency Functionality Concept GIF can be also obtained for some Critical Infrastructure buildings that will be the end consumers of a collection of services, such as Hospitals or health care systems. These systems often receive critical services, and remain the end user of them; and should be accounted as monitors of the functional condition of the complete system. Water, electricity, gas, steam, sewage, roads and even complex systems have to be operational in order to maintain the hospital functionality. Figure 4.11 shows schematics of a Critical Infrastructure building, (hospital). The functionality of this Cell depends on several resources, and equipment that resides inside the 87  building. In order to define the GIF of this cell, at least eight SIFs need to be investigated, but also the post-seismic condition of the structure and the seismic performance of all the systems emphasized in this figure.  4.2.5.1 Interdependency between the Water Station, Electrical Substation and the Hospital Figure 4.12 shows the interdependencies among the Water Station (residing at the Power House), the distribution pipelines (from the Power House to the Hospital), and the electricity provided by the Electrical Substation. Water distribution is dominated by the Water Station seismic performance through intensities VI to VIII; intensities IX and higher are a challenge for water system operators because a lack of water is likely to occur in this region (UBC Campus Case in Chapter 5) due to extended damage in the main distribution pipelines (IX) and in the Water Station. Water distribution in the hospital is guaranteed, at a certain level, only when the earthquake shaking intensity is VIII or lower. Figure 4.12 also shows the Hospital functionality conditions considering the electricity and the water distribution. The hospital functionality condition is linked to the seismic behaviour of the water and the electrical systems. The functionality conditions decrease at Intensity VIII, and in practical terms the Hospital will remain non functional for intensities IX and higher.  88  100% 90% H interdependency Electricity to H PH service at Zone A12 Hospital  80%  Functionality Conditions  70%  74% 61%  60% 50%  42%  40% 30%  21%  20%  10% 0% VI  VII  VIII  IX  0% X  0% XI  Instrumenta l Intensit y  Figure 4.12 GIF at Zone A12 The Hospital functionality condition is closely related to several services within zone A12. Figure 4.12 shows the impact that all services provided by the Power House have over the Hospital‘s functionality. Even though the structural and the non-structural components performed relatively well under Instrumental Intensities VI to VIII (70 % or more functional condition); the overall functionality condition of the hospital will be 74%, 61% and 42%. This behaviour is closely related to the services that the Power House will be able to provide. Resiliency is often defined as the ability of a system to recover its functionality conditions. This implies that it is known and understood where the damage in the system is located, and what actions shall be implemented to recover the functionality. O‘Rourke (2007) stated that the concept of resiliency is evolving, and is an attribute of a community that reflects preparedness and the ability to respond and recover from a disaster.  89  100% H interdependency  90%  Functionality Conditions  80%  Electricity to H 77%  PH service at Zone A12  70%  Resiliency 65%  60% 50%  44%  40% 30%  26%  20% 11%  10%  8%  0% VI  VII  VIII  IX  X  XI  Instrumental Intensity  Figure 4.13 Overall System Resiliency The Critical Infrastructure System resiliency of the whole region can be calculated using the SIF and GIF concepts. Figure 4.13 shows the average value (in black-dotted line) of WaterSIF, the value of the Electrical-SIF, and the value of the Hospital-GIF. Assume that this average value is enough to measure the whole system resiliency; this means that the system will be losing functionality conditions (77%, 65%, 44%, 26%, 11% and non-functional) for Instrumental Intensities: VI, VII, VIII, IX, X and XI (black dotted line in Figure 4.13). In a sense this is not a resiliency measurement, but in this research I am considering that it is the starting point, from where the resiliency should be measured. It is noticeable how related the resiliency of the system is to the Hospital functional conditions. This fact emphasizes that some key Critical Infrastructure buildings that utilize most of the systems of the region can be assumed as monitors of the system‘s resiliency. During the process of this research, through the interviews with the managers of the utilities, they often observed that Hospital or Health care services noticed low in the pressure of the water system, or the lack of steam or gas. In this region three Critical Infrastructure systems have been considered: Electricity, Hospital and the Power House (water, steam and gas), but other CIs will impact the resiliency of the system. This is the case of the transportation system (road, marine or air), and the first responders (paramedics, firefighters, policemen, etc). 90  The road system and the first responders are very critical during the first minutes and hours of the disaster. The success of the emergency planning process is related to the region‘s ability to provide support through the available road systems, and the availability and commitment of the first responders to help people in distress.  4.2.6  Resources and Human Readable Tables (HRTs) With the calculations of the Single and Global Interdependency, it is then easy to build a  performance table that will help develop the HRTs for a given hazard. This table will have information regarding the inputs and outputs of all stations, and their seismic conditions. Table 4.6 Hospital Performance Table Damage Instrumental Assessment Intensity (Phisical Mode) VI VII VIII IX X XI XII  80% 75% 70% 50% 0% 0% 0%  Water, Steam, Gas Electricity Global Single Single Interdependency Interdependency Interdependency (Resource Mode) (x1) (x2) (y) 68% 90% 74% 50% 84% 61% 25% 64% 42% 0% 56% 21% 0% 34% 0% 0% 24% 0% 0% 0% 0%  Table 4.7 Power House Performance Table Instrumental Intensity  Damage Assessment (Phisical Mode)  VI VII VIII IX X XI XII  82% 72% 66% 56% 0% 0% 0%  Water, Steam, Gas Electricity Global Single Single Interdependency Interdependency Interdependency (Resource Mode) (x1) (x2) (y) 100% 90% 68% 100% 84% 50% 100% 64% 25% 100% 56% 0% 100% 34% 0% 100% 24% 0% 100% 0% 0%  91  Table 4.8 Electrical Substation Performance Table Electricity Global Damage Instrumental Single Interdependency Assessment Intensity Interdependency (Resource Mode) (Phisical Mode) (x1) (y) VI 90% 100% 90% VII 85% 100% 84% VIII 66% 100% 64% IX 59% 100% 56% X 38% 100% 34% XI 30% 100% 24% XII 0% 100% 0%  Table 4.6, Table 4.7 and Table 4.8 show the performance tables for the three CI systems in Region A (UBC Campus), after an earthquake hazard. The two main inputs of the region (water and electricity) coming from the Reservoir and the Main Electrical System are not considered in those tables, as they were considered to be 100 % functional at all times. This assumption does not have an impact in the overall functionality conditions of the system.  4.2.7  Example of Resiliency and GIF of Region A12 A deep/interplate earthquake with magnitude less than 7.5 was considered in this  example. The damage that an earthquake of this type can cause ranges from moderate to severe. The following information was computed in this example, and is summarized below: 1. An earthquake magnitude 7.3 was assumed, causing an Instrumental Intensity VIII for the region 2. Three Critical Infrastructure Systems were considered in Region A: Electrical Substation, Power House (that houses all necessary equipment to operate the water, steam and gas systems) and one Hospital 3. The Power House reported the following functional condition, Figure 4.7: a. The structure is 70 % functional, due to light to moderate structural damage b. Mechanical, Electrical equipment and content items remain 80 % functional c. Pipelines inside the Power House remain 50 % functional, this means that some of the pipelines are broken, there might be water, steam, and gas leakages 92  d. The overall functional condition of the systems in the Power House are 65 % to 70 % functional e. 25 % of the utilities‘ volume served to region A12 - Figure 4.3 - (water, steam and gas) remain functional due to Single Interdependency, Figure 4.9 4. The Electrical System reported the following functional conditions, Figure 4.10: a. The Electrical Substation remains 66 % functional, due to light to moderate damage to structural and non structural components b. Electrical underground conduits remain 97 % functional c. 64 % of the electrical service can be provided by the Electrical Substation 5. The Hospital reported the following functional conditions, Figure 4.12: a. The structure is 90 % functional b. Mechanical and Electrical equipment are 98 % functional c. Pipelines and sprinkler systems are 90 % functional d. Furniture and mobile equipments are 95 % functional e. The Hospital remains 70 % functional, the Hospital‘s Global Interdependency is 42 % 6. The overall system functionality condition is reported as 44 %, due to failures at the Power House, the Electrical Substation and the Hospital due to Global Interdependency, Figure 4.13  93  110% 100%  100%  100%  100%  Functionality Conditions  90% 80% 70%  69%  72%  73%  73%  73%  73%  73%  73%  61%  60% 52%  50%  44%  Resiliency of of Region Region A Resiliency  44%  40% 30%  Water, Steam Steam and Gas [SIF] [Single Water, Interdependency] Electricity [SIF] [Single Interdependency] Electricity  20%  Hospital Hospital  10% 0% t0  t1  t2  t3  t4  t5  t6  t7  t8  t9  t10  t11  t12  t13  t14  time  Figure 4.14 Resiliency of Region A and CIs Global Interdependencies In Figure 4.14, the services provided by the Power House (water, steam and gas) are shown in a solid blue line; the electricity‘s functional conditions are shown in a green solid line; and the hospital Single Interdependency is shown in a red solid line. The Hospital‘s GIF is shown in a red dotted line. The solid black line portrays the Resiliency of Region A, as the average functionality condition of the three systems. It is important that Region A, the earthquake Hazard, all Critical infrastructure systems, and a good model of the region are pre-defined and pre-modelled. This is the key for recovering all systems after a hazard has happened. The following events and decisions are shown in Figure 4.14:   Times t0, t1 and t2 represent the functional conditions of Region A before the earthquake happens    At time t3 a hazard (earthquake) strikes Region A    In the period of time between time t3 and time t4, the instrumental intensity is described as VIII. Damage assessment has been performed and confirmed for the 3 CIs investigated in the Region: Hospital, Power House (water, steam and gas), and Electrical Infrastructure Systems. The functionality conditions of Region A 94  are also described and computed. The system‘s functionality is confirmed to be 44%, due to the following facts: o Water, steam and gas services are 25 % functional o Electricity is 64 % functional o The Hospital is 42 % functional due to Global Interdependencies The knowledge and the coordination of the CIs in Region A will help deciding where to start the repair and recovery efforts. In this region is obvious that the Power House needs a lot of attention, as well as the Hospital. And due to the complexity of the Electrical System, the recovery process of it can be delayed so that search and rescue processes might commence. In this exercise it was assumed that the main objective is to preserve human lives, and therefore the Hospital system needs to be recovered first.   At time t4 it is decided that the debris at the Power House must be cleared; and that all the non-structural components and building contents shall be repaired. Backup systems are turned on to compensate for lack of electricity. In the Hospital the same debris clearing process has commenced    At time t5 leakages in some of the pipelines of the Power House have been recovered and an extended cleaning and repairing process has been conducted in the whole Power House system Distribution and Transmission pipelines, the system has been recovered to 42 %. The functionality of the system is reported as 52 %. Repairs and backup systems start at the Hospital    Times t6, t7 and t8 show different recovery activities that cause the Power House to reach 75 % of its functionality conditions    Times t6, t7, t8, t9 and t10 show different recovery activities that cause the Hospital to reach 100 % of its functionality conditions    Times t6 (52 %), t7 (61 %), t7 (69 %), t8 (72 %) and t9 (73 %) show a recovery of the resiliency of the system    At time t9 it is evident that the recovery process in the Electrical System shall start  95  In this chapter have been shown ways to measure SIF and GIF as well as the Overall Resiliency of the system. These calculations were used as starting point for the measurement of the resiliency of the system throughout the recovery time. It is important to notice that the Resiliency of the whole system will not be recovered unless repair and clearing efforts are conducted in the Electrical System. The resiliency graph also shows that GIF of the Hospital works as a monitor or a sensor of the whole system. It is evident that sensor systems can be established in the Region, and a few key buildings can be also instrumented to measure all services‘ status.  4.2.8  Interdependency Matrix Consider the scheme in Figure 4.15, it shows region X with three assets: a Power House  (PH), a Hospital (H) and an Electrical Substation (SS). A representation of the framework of these critical infrastructures, with interconnected channels, and the tokens being exchanged is shown with 7 degrees of freedom in Figure 4.15-upper and Figure 4.15-lower represents any Region. The external critical infrastructure systems (reservoir, the electrical station, and the water and power grid providing water and electricity to region X) are considered as boundary conditions in the ―Outside world‖ box CIs and the objective functions for the manager of the region must be clearly defined. The CIs then are investigated and defined, and finally all interconnections amongst CIs are then determined. Utility services are usually served by provincial or local sources, such as water or electricity; these important assets are often located outside the boundaries of the Region, and hence they are considered as boundary conditions. These boundary elements (water reservoir and aqueducts or transmission trunk lines) are considered to be 100 % functional; but their functional conditions are also related to directionality and through hazard-damage relationships. The reservoir, the electrical station, and the water and power grid providing water and electricity to region X are then considered to be 100% functional.  96  H  PH  REGION X  electricity  wate r  SS electricity  Outside world  BOUNDARY CONDITIONS  Water  WH  Steam  SH  H Steam 2  Electricity EH  Electricity 4  PH  SS  Water  WPH  Steam  SPH  Electricity ESS  Electricity EPH  Figure 4.15 Region X with Three Assets: PH, H and SS, with Channels and Tokens  97  Figure 4.15-lower shows key elements of the Region; in structural analysis this figure would represent nodes and bars in a structure. In this example, the nodes or connections are key assets which receive utilities like water, steam or electricity. The volume of the services received can be referred as tokens, but also as degrees of freedom (dofs). Every node will have several degrees of freedom, or services that are provided or interchanged. In this way SS node, Figure 4.15-lower, has one degree of freedom (dof) (electricity), and it is considered as output. PH node has three dofs, two outputs and one input (water, steam and electricity). H node has three dofs, three inputs (water, steam and electricity). Consider the nodes PH and H and the channel (Water 1) in Figure 4.15. Now consider that an amount of water is supplied, the following equation applies:  The equation shows that the volume of water received by the Hospital (WvolH), equals the amount of water that the Power House (WvolPH) is able to send minus the amount of water loss (Wlosschannel  Water 1)  in the distribution pipelines. This equation shows the interconnections  between the nodes, and that directionality is an important issue in the process of defining an interdependency matrix. The equation defining the connectivity of all these 3 systems is as follows, in matrix notation can be expressed as: Equation 3. Interdependency Matrix  Where: calculated losses in the channels Interdependency matrix 98  vector of tokens being exchanged (inputs, outputs) If inputs and outputs in PH, H and SS are known, then the losses in the channels will be obtained as shown in Figure 4.16.  1   1 0  s   0 1   e3   0 0    e4   0 0  0 1 0 0 1 0 0 0  0 1 0 0  0 0 0 1  WPH  S  0   PH  E  0   PH  * W  1  H    SH   1  EH     ESS   WPH  0.85  S  1.00   PH  1  0.15  EPH  0.95   s  0.00     if   W  0.70   H   e3  0.05 S  1 . 00  H     E  0.80  e4  0.20  H   ESS  1.00  Figure 4.16 Interdependency Matrix Where: v, accounts for the losses of volume in the water channel s, accounts for the losses of volume in the steam channel e3 and e4, account for the losses in the electrical conduits  99  Whenever an event or a hazard strikes, losses in the channels and also the overall functionality of critical assets can be predicted before hand; hence outputs and losses can be evaluated, as a function of functionality. There are several channels exchanging tokens amongst the three Critical Infrastructure buildings in the region; in order to obtain the Interdependency Matrix, these Critical Buildings and channels were localized. Table 4.9 Interdependency Matrix Channel w1 s2 e3 e4 tw1 te1  Power House Hospital Electrical Sub Station water steam electricity water steam electricity electricity -1 1 -1 1 1 -1 1 -1 1 1  Table 4.9 shows seven channels that serve water, steam and electricity. The transmission channels serving water (tw1) and electricity (te1) are also included, and they serve the Power House and the Electrical Substation. Four distribution channels (w1, s2, e3 and e4) are also shown, they serve utilities from the PH to the Hospital, and from the SS to the PH and H, Figure 4.15. The elements of the Interdependency Matrix are -1 or +1 numbers; +1 represents the input of a utility service (or an exchanged token), and -1 is the output of the utility service. In the Interdependency Matrix of Table 4.9, in row e3, it is shown the connections of the Electrical Substation and the Power House. The Substation is sending electricity (-1) through the channel e3 to the Power House (+1), Figure 4.15-lower. This matrix reflects the interconnections among CI buildings, as well as the channels used. In this way the loss in the channel can be predicted as follows, considering row e3 in Table 4.9: Loss in channel e3 = electricity received by the Power House – electricity sent by the Electrical Substation 100  This last equation can be conceived as the Interdependency equation because it shows the provider, the receiver and the channel that connects these two critical assets. It is also evident that the Interdependency Matrix can be populated with critical assets and transmission channels that are not contained in the study region. For the exercise only those elements inside the region were considered, Figure 4.16. It is clear that whenever two elements of the equation are known, the third one can be calculated; this is obvious, but when underground pipelines are assessed to find leakage or breakages, sometimes the calculation of the amount of losses can become very difficult.  4.2.9  Interdependency Index and Importance Factors Another important issue when dealing with interdependencies and critical infrastructures  is to measure the level or degree of interdependency of every asset, as well as the importance of the asset to the whole region. Table 4.10 Interdependency Indices and Importance Factors Substation (mV) -100% 20% 10% 20% 1 1/8 = 0.125 3 3/6 = 0.500 1 (1/4 = 0.250)  Water station (lt/hr) -100% 20% 30% 2 2/8 = 0.250 2 2/6 = 0.333 2 (2/4 = 0.500)  Steam station (kg/hr)  -100% 50% 2 2/8 = 0.250 1 1/6 = 0.167 2 (2/4 = 0.500)  Hospital (patients/hr)  10 3 3/8 = 0.375 0 0 3 (3/4 = 0.750)  Substation Water station Steam station Hospital Inputs Interdependency Index (inputs) Outputs Importance Factor Interdependency Index (resources)  Table 4.10 shows the amount of service that all the stations in the region provide; it is worth noting that there are 6 output and 8 input channels in the region. A simple way to measure the level of interdependency among these stations is by counting the number of inputs that are needed to perform their functions. It shows the interdependency ranking for this system: substation (1); water station (2); steam station (2); and hospital (3). The same procedure can be used for the Importance Factor, by counting the outputs of every asset: the substation (3); the water station (2); the steam station (1); and the hospital (0). In terms of resources, there are 4 resources that the system needs: water, electricity, steam and discharged patients, the 101  interdependency can also be evaluated through the number of resources that any station or asset needs to perform its function. From Table 4.10 it is clear that the most interdependent asset is the Hospital (3/8), it needs 3 out of 8 inputs in the region; this information is relevant for the manager of the region, because it emphasizes the number of channels needed for the hospital. But in terms of resources the Hospital is also the most interdependent asset (3/4), and the least interdependent is the Electrical Substation (1/4). The most important asset in this case will be the Electrical Substation, its Importance Factor is 0.500, and the Hospital is the less important. But this only has to do with the number of output channels in the region.  102  5 UBC Test Case 5.1 Problem Definition: UBC Test Case 5.1.1  Description The main campus of the University of British Columbia (UBC), located on the Point  Grey Peninsula at the Western edge of the City of Vancouver in British Columbia, Canada, Figure 5.1, was used as a demonstration site for the i2Sim simulation framework. The main UBC campus can be considered as a municipality independent of the City of Vancouver. It covers an area of almost 2,000 acres and, during the academic session, it has a population of approximately 60,000 occupants during daytime and 10,000 full-time residents. The geographical location, infrastructure complexity, and diversity of its population made it an ideal test case to develop and validate the i2Sim methodology. There are four main roads at UBC: South West Marine Drive, 16 th Avenue, University Boulevard and 4th Avenue, Figure 5.1. The majority of lifelines that supply the campus follow these four routes. The university manages the distribution of water, gas, steam and electricity to all of its buildings through two key buildings, the power house and the electrical substation. These buildings receive water and power respectively and distribute the utilities across campus. The campus also has its own hospital complex as well as fire, police and ambulance stations.  103  Figure 5.1 UBC Location. Source: Google Maps  104  5.2 Methodology: UBC Test Case The UBC Test Case was disassembled into a finite number of physical layers (critical infrastructures - CIs), that contain both the Building infrastructure and the Lifeline systems. Risks will be estimated separately, and interdependencies will also be evaluated. Figure 5.2 shows the schematics and ontology of the Power House (blue), the Hospital (red) and the Main Electrical Substation (yellow). In the Power House reside two important cells: the water and the steam station. Four channels are also shown, the main substation distributes electricity to the Hospital and the Power House; the water station distributes water and the steam station distributes steam to the Hospital. Some inputs and outputs are not considered in Figure 5.2 in order to simplify the explanations.  Power House x x ~  Hospital  +  +  ~ water  + + electricity 2 X  Sub-Station  Figure 5.2 Power House, Hospital and Main Substation for UBC Test Case  105  5.3 Damage Assessment: UBC Test Case The UBC campus was disassembled into a finite number of physical and human layers. These layers and associated infrastructures were classified according to i2Sim‘s ontology into cells, channels, tokens, and control points. Examples of UBC cells are: hospital, water pumping station, electrical substations, telecommunications substations, administrative services, fire hall, ambulance service, RCMP police station, classrooms, research labs, student residences, etc. Once the cells and channels were identified, their input/output transfer functions were determined in the form of human readable tables (HRTs) from the owners of each of the infrastructures. For example, the HRT for the hospital was constructed through technical specifications for water reserve, and electricity reserve, and from operational experience by the hospital administrators regarding the needs for resources to be able to maintain a number of long-term, mid-term, and emergency beds. The HRT tables are constructed for a set of expected operational states. A main factor influencing these operational states was the expected degree of damage to building and internal lifelines (electrical wires, water pipes, etc.) suffered during the disaster (e.g., an earthquake in this case study). The evaluation of earthquake damage to Critical Infrastructure at UBC Campus using SRA involved the development of databases, the assessment of the expected level of damage to lifeline systems (buildings - structural and nonstructural components -, water, roads, gas and electricity systems), and the estimation of monetary, human and functionality losses. In general, damage functionality conditions, loss of service and interdependencies were obtained and evaluated. The following methodologies and procedures were used for damage assessment computation: Bell (1998); Blanquera (1999); Cook (1999); Onur (2001); Onur, Ventura and Finn (2005); Thibert (2008); ATC (1985); and Hazus FEMA/NIBS (1997) and (2003). For buildings, direct losses were the result of earthquake damage and include the loss estimation of human (casualties), monetary and functionality conditions. The number of casualties is determined based on the level of structural damage suffered by a building and the 106  number of occupants at the time of the earthquake event. Three times of day were selected for the casualty estimation: 2am, 2pm and 5pm. Lifeline systems, such as Water, Gas, Electricity and Road systems were also assessed. The methodology used for lifeline systems is a traditional SRA methodology. Permanent ground displacements were taken into consideration for the damage estimation of underground pipelines, and slope failure was also considered for the functionality conditions of roads. The consequences observed in this study included the total number of casualties, the direct and indirect economic losses and the loss of function. The consequences determine the level of risk associated with a particular seismic event. This risk level should be evaluated by policy makers and government officials to determine if the level is acceptable.  5.3.1  Identification of Critical Lifeline Systems Global or local authorities often have a ranking of important CIs. For UBC Test Case ten  Lifeline Systems have been identified as critical assets; and in the JIIRP project survival tokens have been compared to Critical Sectors, as shown in Table 5.1: Table 5.1 Comparison between Survival Tokens vs Critical Sectors, Source: JIIRP Presentation (2008) SURVIVAL TOKENS Water (suitable for drinking) Food (adequate for emergency situations) Body Shelter (breathable air, clothing, temperature, housing) Panic Control (hope, political and religious leaders, psychologists, media) Personal Communication (whereabouts of loved ones) Individual Preparedness (education) Sanitation (waste disposal, washing) Medical Care (medicines, physicians, nurses) Civil Order (fire fighters, police, army)  CRITICAL SECTORS (PSC, CANADA) WATER FOOD ENERGY FINANCIAL COMMUNICATIONS TRANSPORT HEALTH SAFETY, ORDER GOVERNEMENT, DEFENCE MANUFACTURING  For UBC Test Case, six lifeline systems have been considered: Buildings, Water system, Electrical system, Steam system, Gas system and Roads. These six systems comprised all critical sectors according to Public Safety Canada, and the survival tokens that need to be 107  available during emergency events. During the UBC Test Case information preparation, interviews with UBC stakeholders showed the importance of those lifeline systems for the well being of UBC.  5.3.2  Information (Required vs Outcomes) Imagine asset A provides water to a region X in a city, Figure 5.3, and that is important  to investigate the impacts in region X when asset A has problems.  A  REGION X C B  Figure 5.3 Region X, and Assets A, B and C An example of the type of information needed for the project is shown in Table 5.2 and Table 5.3. A key aspect of the damage estimation is to predict the final outcomes, (outputs or tokens) within region X. Table 5.2 and Table 5.3 shows a list of information required and important outcomes obtained from the damage estimation:  108  Table 5.2 General information used for Earthquake Damage Estimation Lifeline  General  Information -  Comments  Soil type Environment Earthquake hazard Earthquake parameters (a, v, d) GIS Maintenance Layout information Site response analysis For pipeline assessment: Soil characteristics Materials Geometry characteristics (diameter, length, segments) Directionality  Used for all the systems  Table 5.3 Information Used and Outcomes for Earthquake Damage Estimation Lifeline  Information  Outcomes  - Total of Buildings - Geometry of buildings (area, storeys) - Population - Occupation characteristics Buildings  - Materials  Structure Non structural components Casualties  - Structural information  Functionality  - Foundation  Monetary loss  - Preload history - Operational and functional components - Components of the water system: - Reservoir - Trunk lines - Level stations Water System  - Water station - Main pipelines - Pressure stations  Leaks / km  - Materials  Breaks / km  - Service consumption  Loss / km  Components of the system: Gas System  ▪ Trunk lines ▪ Station (Gas or Steam) ▪ Main pipelines  Steam System  ▪ Materials ▪ Service consumption - Main roads - Secondary roads  Road System  - Materials  Structure  - Soil characteristics - Directionality - Transmission lines Electrical System  - Main substation - Secondary substations  Loss  - Line conduits  109  5.3.3  Hazards  5.3.3.1 Earthquake Based on the risk matrix of the British Columbia Provincial Emergency Program (PEP), a ranking of critical events for the UBC campus was developed. Due to the susceptibility of the area to earthquakes, this disaster scenario was selected for the test case.  5.3.4  Results  5.3.4.1 Earthquake Damage Assessment Hazard Seismic Risk Assessment (SRA) was conducted on UBC Test Case for seven levels of Instrumental Intensity, VI through XII, in order to cover all possible scenarios for the use of i2Sim. The results for intensity IX are presented herein; other intensities can be found in the JIIRP final report (2009) and Thibert (2008). Buildings and Lifeline Acquired Information Two primary sources of information were used in this study: the UBC Planning Department and the Records Office. The UBC Planning Department provided the results of a similar assessment conducted by Delcan in 1995. This study contained an existing database with the information for approximately 200 buildings. The UBC Records Office was instrumental in gathering data for the remaining buildings, for updating the database from the Delcan study and to create a database for all the lifeline systems. Lack of information was solved with sidewalk surveys along UBC campus. All the information collected was used for the damage assessment, see Table 5.2 and Table 5.3.  110  Buildings 364 buildings of the 600 on UBC campus were assessed. In this research single family homes located outside of the university campus, storage sheds, barns and other buildings that do not play a significant role to the university community were excluded from the study. Buildings in the database were classified into prototypes and the expected seismic damage sustained by a building is related to ground shaking intensity through Damage Probability Matrices (DPMs). Casualties were estimated for three times of day 2 am; 2 pm and 5 pm. 2pm was determined to be the critical time of day because campus population is the greatest at this time. Monetary loss and functionality trends were examine with respect to earthquake intensity and it was revealed that for moderate intensity earthquakes, the losses depend primarily on nonstructural damage, while structural damage plays the most important role for higher intensities. Building Classification. The buildings were classified into one of the BC 31 building prototypes. The majority of buildings on campus are constructed of wood or concrete. Structural Damage. Figure 5.4 presents the results of the structural damage for II IX for each building in the study area. The majority (73%) of buildings is still in the moderate damage state; however there is a significant increase in the number of buildings in the light (13%), heavy (11%) and major (2%) damage states.  111  Figure 5.4 Structural Damage with Modifiers for II IX  112  Non Structural Damage. It has been observed in recent earthquakes that casualties are attributed to structural damage, but economic losses and loss of function are dominated by the damage to Non Structural Components (NSC), at least in important critical service buildings. ATC 13 was the first study to address the damage to NSCs and the methodology is similar to that for buildings; the expected damage is evaluated through the use of damage probability matrices. NSCs are grouped into six facility classes: residential equipment, office equipment and furniture, electrical equipment, mechanical equipment, high technology equipment and laboratory equipment. Recent studies FEMA/NIBS (1997) and (2003) separate NSCs into two categories: displacement sensitive components (include partition walls, exterior wall panels, architectural finishes, piping, cladding and penthouses); and acceleration sensitive components (consist of electrical and mechanical equipment, piping, cantilever elements, parapets and racks). Building contents such as shelving and furniture are also deemed to be acceleration sensitive. Damage to NSCs was calculated in a similar way as the one for buildings. The resulting NSC damage map for displacement sensitive components is presented in Figure 5.5. Casualty Assessment. The number of earthquake casualties is estimated based on the number of occupants in the building at the time of the event and the structural damage caused by the shaking. The number of occupants depends also on the use of the building. Six use classes were defined for the UBC Test Case: residential, educational, industrial, commercial, hotels and healthcare. Figure 5.6 presents the distribution of building and their locations. The majority of buildings on campus are educational and residential, making up 55% and 36% respectively. Most of the residential buildings lie on the eastern edge and around the perimeter separated from the educational core of the University.  113  Figure 5.5 Displacement, Damage for II IX  114  Figure 5.6 Building Occupancy on UBC Campus  115  The total number of people estimated to be on campus at 2am, 2pm and 5pm is 7,000, 40,000 and 21,000 respectively. Figure 5.7 presents the number of people in each type of building at 2 pm and their locations. At 2am the majority (85%) of the population is residential. At 2pm and 5pm most of the population is comprised of students, faculty and staff, although there is significant portion in residential buildings at 5pm. The total number of casualties estimated for UBC campus for an intensity IX earthquake occurring at 2am is 30; at 2pm is 361; and at 5pm the total number of casualties due to an intensity IX earthquake would be 204. Figure 5.8 shows casualties at 2 pm. In terms of casualties, the worst case scenario for UBC campus would be an earthquake occurring at 2pm, when the population is the highest. This is a special characteristic of a university campus. Functionality Assessment. The functionality assessment was performed according to the methodology described in Thibert (2008). The buildings were divided into functionality categories according to the damage estimations: structural, displacement sensitive components, acceleration sensitive components and building contents. The worst case category was chosen as the overall functionality. Figure 5.9 shows the functionality category for II IX for each building.  116  Figure 5.7 Population on Campus at 2 pm  117  Figure 5.8 Casualties for II IX on Campus at 2 pm  118  Figure 5.9 Campus Functionality for II IX  119  Lifelines The extent to which a lifeline system is affected depends on the extent of damage to the main components, distribution components, and service components of the facility. In some cases these components are buildings, specific facilities, pipelines, pumping stations, control rooms, etc. The methodology used is a traditional SRA methodology; the water system will be described in this thesis. Water Supply System. The water supply system is comprised of several elements: Reservoirs and Transmission Aqueducts; Pumping or pressure Stations; Storage Reservoirs; Treatment Plants; Terminal Reservoirs/Tanks; and Trunk Lines. The water distribution system of UBC is comprised of four main water pipeline systems that distribute water across campus and a Water Pumping Station. The water system of UBC will be considered with these four main components, Figure 5.10: 1. General Reservoir (GR) 2. Secondary Reservoir (Pressure leveler) (SR) 3. Water station (WS) 4. Trunk lines (PL) UBC Test Case Water Supply System. Only two systems were considered for the evaluation of the Water System, the Water Station and the Trunk Lines (pipelines going from the reservoir to the water station and pipelines going from water station to UBC campus), those systems were evaluated as independent components. And the results were then used to evaluate the overall functionality and interdependencies. Interdependencies were obtained based on functionality conditions. Pipeline Damage Assessment. The computation of the damage factors for the pipelines is a straight forward procedure that is not discussed in this report. The methodology used is described as a level two analysis, FEMA/NIBS (1997) and (2003), that includes probability 120  estimates of (1) component functionality and (2) damage, expressed in terms of the component‘s damage ratio (repair cost to replacement cost). It also includes the expected number of leaks and breaks and a simplified evaluation of the potable water system network performance. For pipelines, two damage states are considered: leaks and breaks. There is another set of secondary distribution water pipelines providing water all across UBC; this set is ignored because the main system is considered as the critical part, Figure 5.11.  Figure 5.10 Water Main Pipelines Showing Low Direction. Yellow Lines Refer to Trunk Lines Providing Water from the Reservoir to the Water Station. Blue Lines and Red Lines Main Distribution Lines Providing Water All Across UBC. Source: Google Maps  121  Figure 5.11 Trunk Line from the Reservoir to the Water Station. Source: Google Maps Figure 5.11 shows a portion of the pipeline providing water to the water station and Table 5.4 shows information used for the damage assessment, this information is distributed. Pipelines are often made from different materials and diameters, and hence the damage estimation is computed by segments as shown in Table 5.5. Table 5.4 also shows directionality, which is represented with numbers (1 to 5); thus water flow goes from pipe P-1000 through pipe P-1052. It should be noted that damage to pipelines is calculated with breakages and repair rates. In this research I used the breakage estimation to account for volume losses. Table 5.4 Part I, Portion North-East Pipe ID  Length (m)  Directionality  Diameter (mm)  Material  P-1000 P-1010 P-1050 P-1051 P-1052  381.30 441.11 709.40 100.00 70.60 1702.41  1 2 3 4 5  641 591 591 618 591  Ductile Iron Ductile Iron Ductile Iron Ductile Iron Ductile Iron  122  Table 5.5 Repair Rates of Part I, Portion North-East, See Table 5.1 for Details Pipe ID P-1000 P-1010 P-1050 P-1051 P-1052  VI 0.00 0.00 0.00 0.00 0.00 0.01  Instrumental Intensity VII VIII IX 0.01 0.05 0.23 0.01 0.06 0.27 0.02 0.1 0.44 0.01 0.05 0.21 0.01 0.03 0.14 0.07 0.29 1.29  X 1.04 1.21 1.94 0.91 0.64 5.75  Loss estimation in pipelines. Table 5.6 shows the final losses in the water pipeline, the segmented and the accumulated losses, due to the damages throughout the pipeline. Intensities IX, X, XI and XII imposed 100% loss of water, which is why the tables do not show intensity XI and intensity XII values. Table 5.6 Accumulated and Segmented Loss in the Pipeline Pipe ID P-1000 P-1010 P-1050 P-1051 P-1052  VI Seg 0.00 0.00 0.00 0.00 0.00 0.01  VII Acc 0.00 0.01 0.01 0.01 0.01 0.01  Seg 0.01 0.01 0.02 0.01 0.01 0.07  Acc 0.01 0.03 0.05 0.06 0.07 0.07  Instrumental Intensity VIII Seg Acc 0.05 0.05 0.06 0.11 0.1 0.21 0.05 0.26 0.03 0.29 0.29 0.29  IX Seg 0.23 0.27 0.44 0.21 0.14 1.29  X Acc 0.23 0.51 0.94 1.15 1.29 1.29  Seg 1.04 1.21 1.94 0.91 0.64 5.75  Acc 1.04 2.25 4.19 5.11 5.75 5.75  Summary of Pipeline Damage Assessment. A summary of the necessary steps for estimating losses in pipelines carrying liquids could be described as follows: 1. Obtain the information of the pipelines by segments (diameter, material, length) 2. Information of the soil in which the different segments are buried 3. Compute the damage estimation by segments (repair rates and losses) 4. Compute the accumulated losses (throughout the length of the pipeline) 5. Pull out the functionality conditions of the Water Station (including the performance of non structural components) 6. Evaluate the interdependency of the different components of the water system 7. Evaluate the interdependency of the water system to other systems within the ―study space‖ 123  Localized Damage in Pipelines at UBC Test Case. Figure 5.12 shows a map with the distributed losses in the main water pipeline system. This information is very important because it shows the damages by segment, it should be noted that different information elements were produced: maps, databases, tables, etc. For intensity IX, pipelines with major leakage problems are shown in orange to red colours. The operators (operational level emergency agents) of the water system now have a map of possible leakages in the main water system.  Figure 5.12 Distributed Damage in the Main Water System, without Interdependencies This kind of information is relevant for the manager of the water station, because maintenance crews can be sent to repair the critical segments. An internal interdependency evaluation will reveal the overall functionality condition of the system, if the evaluation is performed with the connectivity and interdependency methodology discussed before. Losses in Pipelines at UBC Test Case (interdependency). Figure 5.13 shows the losses in the water system, this loss is based on the functionality conditions of all the key 124  components of the system: the structure of the water station; equipment inside the water station (the electric generators, pumps, control rooms, equipment, internal pipelines, etc); and the main pipeline system. It is revealed that regardless of the damage condition of the pipelines, the water service depends on the functionality conditions of the water station. For this case, the water station is non functional due to the extensive damage suffered by the structure and all equipment related to water services for an Intensity IX earthquake. This internal interdependency condition can be used to assess the interdependency impact to other systems.  Figure 5.13 Accumulated Damage in the Main Water System, with Interdependencies The impact on other systems is relevant to global operators (Emergency Operators, EOs), because then they can establish action plans for the systems that will be affected. Figure 5.14 shows two maps, the first map shows the building and the water pipeline damage assessment, no interdependencies have been evaluated. This map is useful for local operators (operational agents), because they can locate damages and proceed for the repair activities, but  125  this same information is useless to global operators (EOs) because the impact to other systems has not been yet revealed. Interdependency between Systems (water supply and building systems). The second map of Figure 5.14 shows the interdependency of the water system and the building system. It shows that a shortage of water is imminent in UBC Campus after an Intensity IX earthquake. Global operators now clearly see the implications of one system failing, and they can prepare an action plan for such a situation. Health and emergency facilities can now evaluate their efforts in a situation where water is scarce. But again, for local operators such a map can be useless because damages are not localized.  126  Figure 5.14 Overlaid Damage Assessments, and Main Water System and Buildings Interdependencies  127  Interdependency between Road and Building Systems. The same type of analysis was performed for all other 5 lifeline systems. Figure 5.15 shows the damage to the road system. The damage to the local roads (roads with low traffic conditions) turns to be 6.2% to 7.3 %; and to highways (roads with heavy traffic conditions) was computed as 5.4% to 6.1 %. Minor cracks and bumps might be observed in the local roads, and light to no damage in highways. This map would be very relevant for local operators, who might want to verify the serviceability of the roads, which for this amount of damage, would be functional for traffic. But when interdependencies are computed with other CIs, all of a sudden the functionality conditions of local roads and highways declined to 0%. Figure 5.15 shows the interdependency between the road and the building systems. Buildings were evaluated through their functionality conditions. Two components were evaluated: the structure of the building; and all the utilities, contents and equipment inside the buildings often referred as non structural components (NSCs). In the map orange buildings are 0% functional due to failure of NSCs; while red buildings are 0% functional due to partial or total collapse. Interdependencies show that most of the local roads and highways are non functional due to debris, collapses, and population trying to reach emergency units. Even though the road functionality is close to 100%, the associated damage to other systems will cause the roads to be closed for some undetermined time. When other conditions of the building system are being analyzed (casualties or structural damage) with the road system, different road closures will be revealed. Then Global operators will be able to define the best action plan for the whole region, and where to place the necessary closures in the system. The mapping of the closures in roads will define emergency routes for the emergency units, right after the earthquake has happened.  128  Figure 5.15 Functionality Conditions of Roads. Interdependency with Building Functionality  5.3.5  Human Functionality In order to address the surge of physical and psychological injuries during a disaster, the  hospital functionality will be used to measure Human Functionality for the UBC Test Case.  5.3.5.1 Hospital Functionality An arithmetic model is proposed to help ascertain the extent to which a hospital‘s operational efficiency would be compromised during a major earthquake (Instrumental Intensity IX). It is recognized that this is a simplified model because it translates behavioral phenomena and hospital practices into numbers but this simplified model can provide a heuristic for rapid decision-making.  129  5.3.5.2 Fundamental Assumptions of the Model One basic way to conceptualize cell efficiency is to compare capacity to load or demand. When Capacity > Load, the cell (the hospital) can cope. This would represent normal operating conditions. It could be argued that ―normal‖ for a hospital sometimes means high load, it is assumed that the hospital cell is equipped, staffed, and designed so that it can manage day-today demands. The converse, Load > Capacity, threatens the efficiency of the hospital. This might be manifest as poorer health outcomes, staff burnout, complications and/or longer hospital stays. By definition, every disaster features a load > capacity situation. For the purposes of the UBC Test Case, we are assuming that the disaster occurs at 2 pm and the population, casualties and level of injuries are presented in Table 5.7. Table 5.7 Population, Casualties and Level of Injuries after an Intensity IX Earthquake Population  Casualties  Injury 1  Injury 2  Injury 3  Injury 4  39,210  361  264  65  10  22  The level of injury refers to: injury 1 (green), also called ―walking wounded‖; injury 2 (yellow), the level of injuries are serious and need immediate attention; injury 3 (red), the level of injuries are life threatening and they need immediate medical attention; and, injury 4 (black), this are considered deceased or expectant. In a triage, the green and black coded injuries are separated; yellow and red codes are sent to hospital for immediate care, being the red ones the priority for hospitals and emergency rooms. Regarding time duration, the model is working with estimates covering the first 90 minutes after an earthquake.  5.3.5.3 The Arithmetic Model For a hospital responding to a disaster-related surge in patient numbers, the main sources of load are as follows: 130    Number of patients    Severity and nature of patients‘ physical condition    Severity and nature of patients‘ psychological condition    Number of persons searching for missing loved ones  Generally, patients‘ physical injuries are colour-coded in four levels, in increasing order of severity: green, yellow, red, or black. In an arithmetic model, physical casualties can be crudely calculated as severity x number of patients.  The damage assessment module has  produced the numbers shown in Table 5.7. Psychological casualties are more prevalent than physical casualties in the event of an earthquake. Benchmarks for the number of psychological casualties presenting to a hospital in the event of a disaster are set at 5000 per 1 million population, Shultz et. al. (2006). These casualties are generally given a severity rating equivalent to Green, or 1. Thus, the patient severity calculation for psychological casualties in a population of 40,000 people would be 200 x 1 = 200. Many of these casualties would be triaged to other facilities away from the hospital. Nonetheless, they would be treated at the hospital first. It is also difficult to estimate the number of people who would come to a hospital searching for missing loved ones. We do know from past research, however, that this number is usually high enough to have significant impact on the hospital‘s efficiency. For the purposes of our model, we are assuming a 1-to-1 ratio of ―searchers‖ to physical casualties. If these people present with sufficient psychological distress, they might be moved into the psychological casualty category. If not, this issue should be addressed, hopefully by an agency outside the hospital in a waiting area with information links to the hospitals. The effectiveness of this strategy relies on other critical infrastructural elements, such as the communication grid. The severity rating for these people would be less than the Green (1) level. For the purposes of the model, we have assigned a severity rating of 0.5 for the ―searchers.‖ This would mean that a population of 40,000 would yield 361 x 0.5 = 181 for a severity score in this category. 131  In total, the load calculation for a hospital serving 40,000 people experiencing a common surge is shown in Table 5.8 Table 5.8 Load Calculation of Patients and People to be Served by the Hospital after an Intensity IX Earthquake Casualties and level of injury Hospital Waiting Area  1 2 (Green) (Yellow)  4 (Black)  TOTALS  10  22  361  Earthquake casualties  264  Psychological casualties  196  196  181  181  People searching for loved ones  65  3 (Red)  181  People searching for loved ones with psychological distress Total  181  641  65  10  22  738  Table 5.8 shows that the psychological casualties are larger than the physical ones; they double the number of overall casualties in the hospital, from 361 to up to 738 physical casualties; and 181 people searching for loved ones. Load calculations will reach 919 people exhausting hospital services. In the first 24 hours after a large event, the critical casualties (2 and 3, or yellow and red) will arrive later than 90 minutes after the earthquake. Around 100 casualties will require help to be pulled out of damage buildings, and therefore their treatment  Number of casualties or people in waiting room  might be at stake due to number of people reaching health care services. 1300 1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0  Total casualties Total green Psychological Waiting room EQ casualties Green Yellow Red  VI  VII  VIII Instrumental Intensities  IX  X  Figure 5.16 Total Casualties 132  Figure 5.16 shows: casualties (green, yellow and red) in dotted lines after different levels of earthquake instrumental intensity (II); the total earthquake-related number of casualties in dotted black line, (at (II) VIII, 200 casualties); the number of people that will reach the waiting areas of the hospital (around 70 people for (II) VIII, in a pink solid line with blue dots); psychological casualties in a pink solid line with black triangles (around 260 for (II) VIII); the total number of green casualties (earthquake related and psychological casualties) in solid green line (380 for (II) VIII); and, finally in solid black line the number of total casualties, for (II) VIII, there are 400 casualties. It is clear that after the first hours of the earthquake the hospital will be exhausted by people with psychological distress, people seeking support, and people searching for missing loved ones, but the key issue is for yellow and red casualties which barely reach 200 casualties for (II) X. Those yellow and red casualties are probably waiting for assistance in damaged buildings, and for first responders to arrive to rescue them. Red casualties are beyond 50; yellow ones remain below 150; but the task for the health care systems is to deal with people that will overflow the hospital‘s capacity and to implement emergency plans for mass casualties.  133  6 Downtown Vancouver Scenario Vulnerabilities and Mitigation Studies for Downtown Vancouver during the 2010 Olympic Games Case Study: Transportation and Treatment of BC Place Victims  6.1 Introduction To demonstrate the usefulness of the methodology and I2Sim for a closer coordination of CI service availability with emergency responder actions, it was decided to simulate a disaster scenario that is of particular interest to the City of Vancouver Emergency Management team within the context of 2010 Olympic Games. For this purpose, the capabilities of I2Sim were extended under the sponsorship of Defence Research and Development Canada (DRDC) to incorporate the following models: a) traffic model; and b) crowd egress model. Ten sample scenarios have been suggested as being of particular interest to the City of Vancouver (CoV) during the Olympic Games. With the addition of the new I2Sim models, these types of scenarios can be combined into multiple simultaneous events integrated with critical infrastructure damage.  6.2 Sample Complex Scenario The following sample scenario combines two main aspects of disaster response during complex situations: 1) Simultaneous major disaster incidents. A second major incident occurs within the time frame of the response to the first incident, thereby requiring a reallocation of infrastructure and response resources to optimize the overall system-wide objective of saving human lives; and 2) Coordinated versus distributed decision making in optimizing the delivery of insufficient resources and in setting resource restoration priorities by the utilities. 1. Context: the scenario takes place during the Vancouver Winter Olympic games of 2010. The games take place between Friday, February 12 th, 2010 to Sunday, 134  February 28th, 2010 (for Olympics), and Friday, March 12th, 2010 to Sunday, March 21st, 2010 (for Paralympics). The games impose safety, transit and transportation constraints. 2. Participants:   Vancouver Olympic Committee (VANOC) and related agencies    City of Vancouver    Provincial and federal authorities    Venue owners and Venue managers    Critical Infrastructure managers    Emergency Communications (E-Comm) and Ambulances    Translink or Transportation stakeholders    Police and traffic force    Decision makers and other participants  3. Environment: the average temperature range in February is 1 oC to 7oC and there is a fair amount of precipitation (more than half the days will see some rain). Sunset: 5:07 pm to 5:51 pm. March‘s temperature range is 3oC to 10oC. The weather in March continues to warm up. March, like so many months in Vancouver, has plenty of precipitation. Sunset 5:53 pm to 7:24 pm. Two ―vignettes‖ or small scenarios will be used for the Sample Scenario: transportation and egress vignettes. The following elements have been considered:   Time frame of the analysis: hours. This is related to the intensity of the triggering event. Different intensity of events will be considered in order to review the time frame that those different events will impose to the geographic area.    Geographic area: Figure 6.1    Meteorological environment: is described in Environment (3) above.    The scenario will be played during the Olympics in Vancouver, under safety, transportation and traffic constrains. Vancouver and Provincial Emergency Operation Centres (EOCs) will be activated during the games. 135    The triggering event can be equipment failure and structural damage due to accidents, an earthquake, or a terrorist group with the intention of causing damage to critical infrastructures and casualties among spectators of the games.  The objectives of the scenario are: a. To determine ―best‖ decisions regarding priorities in CI‘s restoration and responder resources allocation. b. To reveal vulnerabilities of the included CIs in this setup and scenario c. To reveal factors with a significant impact on the outcome. To test the egress and the traffic models d. To find thresholds in the transportation of injured people and the hospital rate of discharged patients per hour e. To propose mitigation strategies and resource management for emergency response, or recovery activities  Figure 6.1 Geographic Area  136  6.2.1  Scenario Events Major events: Double transformer failure due to an explosion in the downtown Murrin  Substation. Substantial electrical loss of service in the downtown core is reported. An uncontrolled multitude forms around Granville Sky Train Station. At the time these incidents occur, an Olympic event is taking place at BC Place. Within the time frame of the response to the first events, a stage with structural problems in BC Place fails, causing a stage and related supporting infrastructure to collapse on top of spectators. Critical Infrastructures Modelled: The following critical buildings, pipelines, conduits and roads were modelled: 1. BC Place (BCP), and its egress model 2. Two hospitals: Vancouver General Hospital (VGH) and St Paul‘s Hospital (SPH) 3. E-comm and the ambulance dispatch unit 4. Two alternate routes, that will access the hospitals and where ambulances and emergency vehicles will loop to transport injured people a. BC Place to VGH and back: two routes will be considered b. BC Place to SPH and back: two routes will be considered Evolution of events in time: In this simulation, the following six events were considered:   [Time : 00 minutes]  An explosion occurs in downtown Murrin Electrical  Substation.   [Time : 00 minutes]  BC Place Emergency Power generator malfunctions.  Resulting lack of internal pressure control renders Emergency Exits inoperable.   [Time : 05 minutes] A stage collapses on top of spectators at BC Place.    [Time : 06 minutes] A multitude starts an uncontrolled egress from BC Place.    [Time : 10 minutes] Trapped people and casualties are reported in BC Place after the collapse of the stage. 137    [Time : 19 minutes]  Casualties are reported in BC Place due to egress of  spectators. In order to explore different responder strategies, a sensitivity assessment with several alternatives can be considered. It was not the purpose of the exercise presented here to exhaust these alternatives, but to emphasize that other options and alternatives might be explored with the current model: a. Different available resources allocation and sequence of service restoration of electricity and water resources from Murrin and Del Grauer Substations and from the water pumping station, according to distributed or coordinated responses by the electrical and water utilities. Murrin Substation serves GM Place, VGH, water pumping station, and Del Grauer substation. Del Grauer Substation serves Saint Paul‘s Hospital and BC Place. b. Different allocation of first responders‘ resources (e.g., treating Granville with higher priority than Murrin or BC place). c. Different utilization of resources (e.g. regulated versus non-regulated evacuation, police in charge versus pedestrians in charge). Activation of Muster Zones and Disaster Response Routes (e.g. usage of Burrard versus Granville bridges). d. Allocation of victims to SPH and VGH hospital according to resources available at these facilities and traffic congestion. e. Evacuation of residential areas in the vicinity of Murrin Substation f. Downtown public transit buses are diverted to Granville Sky Train Station and BC Place to help evacuation g. Ambulance service performance is affected due to street congestion h. Evacuation of other venues This is a general description of a complex scenario. This sample base case scenario can be modified to analyze other possible simultaneous events affecting two or more critical infrastructures. In the process of producing the model of I2Sim some aspects of this scenario have changed in order to produce significant results. In the following pages, a description of the models and the scenario will be described. 138  6.2.2  I2Sim-ETran Model The Egress model incorporated in I2Sim, will be called I2Sim-ETran, and it is shown in  Figure 6.2. The Infrastructure interdependency simulator, with egress and traffic model capabilities is shown in Figure 6.3. Initial conditions are reflected in BC Place, GM Place, Granville Sky Train Station; both hospitals (VGH and SPH) and E-Comm dispatch centre. In these conditions there is no physical damage to those infrastructures, but they have resource problems. It also shows the electrical system, the four substations, feeders, and the decision points (distributors) of the power system. It portrays BC Place, GM Place and the Granville Sky Train Station and their egress models; E-Comm and Ambulance Servcies models; the Hospitals‘ models Vancouver General Hospital (VGH) and St. Paul‘s Hospital (SPH). In this scenario, three venues and their egress models were modelled. The main objective of the egress model is to estimate the evacuation time and the number of casualties due to the egress process. Even though the egress models of all three venues have similar structures, their different physical attributes impose different metrics in their egress models.  6.2.2.1 Egress Description: The egress model considers the following model sections [blocks]:  Standing Area  Divider (Casualties of Structural Damage)  (1) (2)  Casualties Collection Queue  First Responders ’ Route  (7)  (8)  Corridor (3)  Waiting Area  Evacuees  (4) Actual Density  Modifiers  Type 1 Casualties  Divider (Injured/Healthy evacuees) (6)  Type 2 Casualties Healthy Evacuees  Modifier Block (5)  Figure 6.2 Egress Model 139  1. Standing Area: represents the location where spectators sit inside the stadium, in the egress process they will evacuate this area. The output of this block is the instant number of people that are evacuating from the Standing Area and entering the next common area. 2. Divider, separates the spectators that were injured due to structural damage (Type 1 casualties) from the rest of the evacuees. Since the casualties are yellow and red coded-casualties, they have to wait for the first responders to evacuate them via designated routes. 3. Corridor, this block is simply represented by an average delay time to run, walk or crawl through the corridors (depending on the physical conditions of the corridors, and the level of injuries of the spectators) from the standing area to the waiting area. 4. Waiting Area or Exit Area, this block is the immediate area before the exit doors where people await for their turn to egress. All egress casualties often occur in these areas. This block requires the following metrics according to its physical attributes: 1) Total area (m2), 2) egress rate (evacuees per time) and 3) maximum evacuee density (people per m2). The Waiting Area block determines the actual evacuee density and the number of people exiting at each time interval. 5. The Modifier Block calculates the evacuation time and the number of casualties based on the effective density of evacuees at the waiting area. The density of evacuees waiting by the exits is adjusted by the Modifier Block, by considering different factors: Rapid Response, Layout, Guidance, Demographic and Electricity. Using this calculated effective density, the Modifier Block determines the ratio of casualties due to egress and the additional delay of the egress process. 6. The Divider block (Injured/healthy evacuees) shows the number of injured (Type 2 casualties) and non-injured evacuees due to egress that are exiting the building at each time interval. 7. The Casualties Collection Queue represents the area where the type 1 casualties (casualties due to structural damage) are waiting to be rescued by first responders. The output of this block is the number of casualties that could be moved by first responders at each time interval.  140  8. The First Responders‘ Routes represents the time it takes to rescue from the collapsed area all type 1 casualties. It is important to note that the modifiers are a function of the evacuation conditions of the venue (physical attributes or layout, and security personnel). For example, if there are sufficient guidance (personnel or instructions) and signs to lead evacuees outside the venue, the guidance modifier shall be set to 1. However, if there are not enough guidance nor signs and the evacuees have to find their way to the exit areas, the guidance modifier shall be set to values larger than 1. If there is a power outage and only emergency lights are on, the electricity modifier has to be a value larger than 1, because evacuation will take longer, hence the effective density will increase and therefore more spectators will get injured. The transportation time of casualties is critical in serving the objective of preserving human lives. In I2Sim, the traffic model determines the transportation time delay; this is taken as a function of speed and distance, and speed changes with various conditions: guidance, rapid response, weather conditions, road closures and road damage. I2Sim traffic model has the capability to simulate from normal road conditions to activated Disaster Response Route (DRR) network, with the possibility to include crowds in the streets. Details for this model are discussed in the traffic model‘s manual. In the model from the scenario, three traffic zones were considered, Figure 6.4: Zone 1: Venue perimeter, in this closed environment the traffic restrictions and the number of pedestrians will create difficult road conditions. Zone 2: Downtown core area, road restrictions and pedestrian streets will create different set of conditions. Zone 3:  Area outside the Downtown core of Vancouver. Some of the roads will  experience no-parking and non-stopping restrictions; for this model, some of the roads and transportation trajectories cross Zone 3 to access VGH.  141  6.3 Scenario Outcomes The following outcomes can be obtained with I2Sim. Even though the results presented in this chapter do not address all the issues shown below, it is worth noting that depending on the objectives and conditions, the I2Sim user can revisit the model and use other options. Outcomes available:   Effectiveness of coordinated versus distributed decision making    Identification of key vulnerabilities    Assessment of system resiliency    Identification of communication channels among key players    Identification of hidden interdependencies    Optimization of allocation of limited resources (e.g., spare parts for electrical substation repairs)    Sensitivity analysis to data availability  142  Figure 6.3 I2Sim General Model  143  Figure 6.4 Traffic Zones Considered in the Model  6.4 Effectiveness of Response The following metrics will be used as measures of effectiveness of response in the scenario outcomes:   Minimization of human injuries and casualties    Speed of recovery (ensuring continuity of the games)    Best utilization of resources  144  6.5 Simulation Results The model of downtown Vancouver developed in I2Sim can reveal interdependencies and cascading effects during a disaster. The model can be used to perform sensitivity analysis and results to answer objective function questions. For the example presented in this report, three different assessment simulations will be discussed. Table 6.1 shows the parameters and results that were defined and obtained in the scenario. The following objective function was defined for these simulations: Objective: Evaluation of the time to assess and stabilize casualties at ERs in different Hospitals of Metro Vancouver, and an estimation of the fatalities due to the delay in treatment of those casualties. In order to address this objective, three conditions were independently varied in the model of Downtown Vancouver: 1. Road conditions: Two hospitals were considered (VGH and SPH); 30 casualties were defined (20 yellow-coded and 10 red-coded casualties) that were divided equally between both hospitals. The following road conditions were modeled with the transportation model developed in I2Sim: a. Normal road conditions during the 2010 Winter Olympics, including delays caused by large numbers of pedestrians in Zone 2 b. Normal road conditions and no pedestrians during the 2010 Winter Olympics. Here, emergency responders manage to contain pedestrians and prevent them from accessing the emergency routes. c. A ―Best Case‖ scenario where DRRs are already activated when the event occurs. 2. Number of casualties: Two hospitals were considered (VGH and SPH); Casualties were divided equally between both hospitals. In all of these cases, it was assumed  145  that DRRs were available for transportation to the hospitals.  Three levels of  casualties were considered: a. 30 casualties b. 50 casualties c. 100 casualties 3. Number of casualties – Severe Case: A large number of casualties were assumed in order to exhaust Metro Vancouver Health Services. Seven hospitals were modelled. In all of these cases, it was assumed that DRRs were. Three levels of casualties were considered: a. 280 casualties b. 860 casualties c. 2,500 casualties In order to make clear estimates of the variables involved in the model, Figure 6.1 shows input and output parameters necessary to carry on the simulations with the three conditions that were discussed before. It is a good idea to arrange all parameters in a user‘s critical pre-ranking order. It is also recommended that time variables must be grouped together, and that all output values be also clearly defined. This table was used to observe input parameters, and to define which outputs might be relevant for the objective of the scenario. Figure 6.1 has five colour schemes, which were used to defined groups of parameters sharing similar characteristics: number of casualties; number of hospitals; time variables; transportation variables and results from I2Sim. This arrangement can lead the user to think clearly about the input and output parameters to be used in the model; this constitutes a pre-process activity for the parameters for the model. Some of these parameters may lead to further refinement; this is the case for the time variables: initial assessment, rescue process, egress and triage. Figure 6.1 was used to present the results of the different simulations, this arrangement of input and output parameters can give the first responders a way to look at results in a wink of the eye.  146  Table 6.1 Variables for Object Function Simulations Time variables No of No of casualties Hospitals  Initial assessment  Rescue process  Egress  Transportation variables Triage  Routes  Number of ambulances  Patients in ambulances  Road traffic conditions  Number of casualties No of Hospitals Initial assessment Rescue process Egress Triage Routes No of ambulances Patients in ambulances Road traffic conditions Transportation time Arriving time at ER (red-coded casualties) Time in which red-coded casualties are assessed and stabilized in ER Time in which all patients have been assessed and stabilized in ER  RESULTS FROM I2SIM  147  Results from road conditions Time frame for the simulations The following time frames for transportation of casualties were assumed: Transportation of casualties (type 1 casualties) due to a collapsed stage at BC Place, Figure 6.5: t 1. Initial assessment. Once the structure has collapsed, trapped people will require help to escape. Security personnel will assess the security conditions and the extension of the damage due to the wreckage prior to providing aid. Figure 6.5 shows this time as 2 minutes; it was assumed that safety personnel at BC Place will have enough knowledge to quickly assess the situation and start helping and clearing the collapsed scene. t 2. Rescue of trapped people including triage. Security personnel and emergency first responders will try to pull out trapped people from the collapsed structure. In this process the severity of injuries will be decided (triage). Figure 6.5 shows this time in two categories: rescue (3 minutes) and triage (5 minutes). It is assumed that there are paramedics inside BC Place, and that these first responders will quickly start the triage process. t 3. Transportation of casualties. It is worth noting that it is assumed that 2 ambulances load patients and leave the venue area at a time (first responders can handle two casualties and two ambulances every minute). The transportation time is the time that it takes for an ambulance to travel from BC Place to the designated hospital. Figure 6.5 shows transportation time as 21 minutes; but there is also an additional time for transportation of casualties (6 minutes). This is assumed to account for the time that ambulances take to line up, to prepare the casualty inside the ambulance, and the handing off the casualty once they have arrived to the ER.  148  Critical time for patient tretament BC Place to VGH (Road conditions: Normal Roads) 5 red-coded casualties due to a collapsed structure at BC Place  Assessment and stabilization of casualties …  24  Transportation of casualties  6  1st patient to arrive (transportation time)  21  Triage  5  Rescue  3  Initial assessment  2  0  5  10  15  20  25  30  35  40  45  50  55  60  65  Time (minutes)  Figure 6.5 Time for Patient Treatment following Collapse Stage at BC Place t 4. Assessment and Stabilization of Casualties at ER. Once the casualties have arrived to the ER, it is assumed that both Vancouver General and St. Paul‘s hospitals have the capacity to handle 10 Code Red casualties per hour, with an average time to stabilize a given patient of 30 minutes. It was assumed that the human and physical resources to maintain the ER‘s rate will be focused on assessing and stabilizing the red-coded casualties. The 30+ minutes to stabilize the patients was assumed to account for aggressive ER treatment such as CPR or other medical procedures necessary for the casualties to undergo comprehensive treatment, such as trauma surgery, in which case the casualties are transferred to other hospital locations such as traumas, intensive care, OR units, etc. Figure 6.5 shows that all red-coded casualties will be assessed and stabilized 61 minutes from when the events started unfolding. The first patient will arrive to the ER after 31 minutes, and all red-coded patients will arrive after 37 minutes. Two important issues should be mentioned: triage of mass casualty disasters differs in important ways from traditional triage processes: victims can be distributed in a wide area (this is true for earthquakes); medical resources are limited; the length of time before a patient can receive standard care is unpredictable; and early evacuation might not be possible, Schultz et. al. (1996). CPR is not performed, unless sufficient resources are available without jeopardizing the lives of other victims. Dynamic triage techniques have developed to overcome these situations in mass casualty disasters, Benson et. al. (1997).  149  In terms of triage for medical casualties, first responders and EMS staff are generally trained in a triage system known as START (simple triage and rapid treatment). Practitioners are trained to classify disaster survivors into four categories based on a colour-code scheme: Green (minor); Yellow (delayed); Red (critical); Black (diseased/expectant). The triage process assumed in this simulation takes into consideration the fact that a dynamic triage is in place in BC Place. Hospital triages may use the same color-coded system. In the triage process it is assumed that black-coded are separated from the rest of the casualties; green-coded are released at the scene (walking wounded); yellow-coded and red-coded are transported to ERs. The transportation priority is red-coded casualties, those casualties will be transported first and then yellow-coded ones. Time frame for the transportation of casualties due to egress in BC Place, Figure 6.6: t 1. Egress Time from BC Place. I2Sim egress model is able to include the time it takes to rescue trampled and crushed people as well as time for triage process. Security personnel, first responders and spectators will try to pull out trapped people during egress, the severity of injuries (triage) will be assessed, and readied for transportation by ambulance. t 2. Transportation Time of Casualties. Figure 6.6 shows the transportation time that the ambulance will take to drive from BCP to VGH (21 minutes). It also shows the additional transportation time for all casualties to arrive at ER in VGH (5 minutes). t 3. Awaiting Treatment. In the emergency process of the current simulation, it is assumed that the first casualties to be treated at ER are the red-coded victims; after that time has passed, then the yellow-coded victims will be treated. t 4. Assessment and Stabilization of all Casualties. All 10 yellow-coded casualties will be treated in 60 minutes. It‘s worth noting that CIS group has decided to break up the treatment time as mentioned before (first all Code Reds, then all Code Yellows), but it can be divided in different activities or in a different order, according to the first responders‘ activities.  150  Egress and Transportation Assumptions The numbers were estimated by taking measurements at the venues and timing the evacuation of spectators after a football game. The numbers represent the evacuation of the venues during normal conditions; the time was adjusted for an emergency evacuation case. For I2Sim Egress model the following assumptions were made:  BC Place Egress Assumptions Total number of spectators at the time of the event  60,000 people  Rate of people evacuating the standing area  19,440 people per minute  Rate of people evacuating the venue through turnstiles  5,760 people per minute  Rate of people evacuating the through emergency exits  2,880 people per minute  Waiting area by exits  1,000 m2  Total delay time of casualty evacuation by first responder  5 minutes  Critical time for patient tretament BC Place to VGH (Road conditions: Normal Roads) 10 yellow-coded casualties due to egress process at BC Place Assessment and stabilization of all …  60  Awaiting treatment  24  Transportation of all casualties  5  Transportation time  21  Egress  16 0  5  10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125  Time (minutes)  Figure 6.6 Critical Time for Patient Treatment, Egress BC Place The numbers for GM Place were estimated using the maps available on General Motor Place website and taking measurements outside the venue.  151  GM Place Egress Assumptions Total number of spectators at the time of the event  14,000 people  Rate of people evacuating the standing area  15,600 people per minute  Rate of people evacuating the venue through turnstiles  5,400 people per minute  Rate of people evacuating the through emergency exits  10,800 people per minute  Waiting area by exits  26,309 m2  Total delay time of casualty evacuation by first responder  5 minutes  Additional egress assumptions:   The triage process takes about 5 minutes for yellow and red-coded victims. The assumption is that for mass casualties ―…triage may involve providing some basic life saving measures, but it is not meant to be the time at which aggressive or definitive care is provided. It should be remembered that triage in a disaster situation is different from triage in a hospital setting in that the time to definitive care is unknown in a disaster situation…‖, Delaney et. al. (2002).    The number of casualties from the egress process is calculated based on the density of the spectators at the waiting area near the exits. It is assumed that when the density reaches 4 people/m2, a percentage of evacuees would get injured (the majority will be green-coded casualties or minor injuries).    Even though the model considers five modifying factors: Demographics, Guidance, Rapid Response, Layout and Electricity to estimate the number of casualties and the egress time. The lack of electricity is the only one that has a repercussion in the outcomes of this scenario. The external source of electricity is disconnected (Murrin Substation is down) and only the emergency lights are on during the evacuation process. In this case, it is assumed that the lack of electricity increases the delay time and hence the number of casualties.    Two different types of exit gates are used in BC Place for egress purposes. The physical setting of this stadium and its inflatable dome makes it necessary to maintain a higher air pressure inside the stadium so that the dome structure 152  remains in place. Therefore normal exit doors were defined as turnstile style and a two door system for wheel chairs and other vehicles. The other type of exit doors are meant for emergency purposes, a computer system will equalize the stadium‘s and the outside‘s air pressures, so that the emergency exit doors can be opened completely and the egress process can be faster than a normal one. In this example it was assumed that the computer system that equalize the air pressure inside the stadium fails, and therefore only turnstiles and the two-door systems can be used for egress. Assumptions for traffic model (speed considered for emergency vehicles):   Zone 1: Considered as the related parking areas and access streets of BC and GM Places, a constant time delay is considered, Figure 6.4.    Zone 2: Downtown core area with pedestrians leaving BC Place, Figure 6.4. Maximum speed considered: 30km/hr.    Zone 3: Area out of the downtown core area, with normal traffic conditions, Figure 6.4. Maximum speed is 60 km/hr.  If vehicles travel in regular road network they will travel through all the three zones and will experience different speeds due to different road conditions. When DRR is activated the designated emergency vehicles can go through DRRs, traveling through different zones and without time delays. The maximum speed considered is 90 km/hr. Figure 6.7 shows the roads that were considered in the simulation, it was assumed that those routes will be cleared for emergency transportation purposes.  153  Normal Roads BCP-SPH  DRR BCP-SPH  DRR BCP-VGH Normal Roads BCP-VGH  Figure 6.7 Roads Considered in the Simulation  6.5.1  Road Condition Results The objective function for the simulation is the span of time to treat casualties at ERs in  two hospitals (VGH and SPH). Table 6.2 shows important input parameters and results obtained from the simulation. Thirty casualties were defined for the simulations: 5 red-coded casualties were assumed during the collapse of the stage inside BC Place. The collapsing stage, as well as the blackout produced an uncontrolled egress from BC Place provoking a stampede, and 25 casualties were assumed as well (5 red-coded and 20 yellow-coded casualties). Three road conditions were investigated and timeframes for casualty treatment were also defined. These road conditions were set up in order to verify their impact in the delay of time for treatment. Columns in Table 6.2 show the three road conditions, the 30 casualties are equally distributed in the hospitals; routes, distances and road conditions are defined. ―Normal roads‖ refers to streets and avenues during the 2010 Winter Olympics in Vancouver; three zones were 154  assumed with different speed values in the model, Figure 6.4. ―DRR‖ refers to Disaster Response Routes, which are roads meant to be exclusively for emergency response vehicles, it is assumed that these roads will have clear access; hence a constant speed was assumed. ―Normal roads and no pedestrians‖ refers to roads during the 2010 Winter Olympics in Vancouver, but somehow emergency responders manage to restrain people from accessing the roads and streets where emergency vehicles are circulating. This is a condition that might be observed during the Olympics, as there are dedicated pedestrian streets in downtown Vancouver. The results from I2Sim are emphasized. The total transportation time in the table is the summation of the initial assessment, rescue, egress, triage and transportation times. It is worth noticing that in the simulation the egress time also accounts for the triage, as it is unknown when the casualties start accumulating, when they are detected and rescued during the egress process. The I2Sim user has to realize that the process for transporting victims from the place where the disaster occurred to the ER in a nearby hospital, involves several activities. If ―normal roads‖ are considered then the transporting time will reach 41 minutes due to traffic conditions; but if emergency routes are defined ―DRRs‖ this time will be 31 minutes. Ten minutes might be the difference between being dead or alive. Figure 6.8 shows the relationship between road conditions and the transportation time. The bars account for the type of casualties and the two hospitals consider in the simulation: BCP-VGH and solid lines, account for the transportation from BC Place to Vancouver General Hospital; BCP-SPH and dashed lines, account for the transportation from BC Place to St Paul‘s Hospital; the black bars account for the transportation of yellow-coded casualties, and the red bars account for the transportation of the red-coded casualties. It is clear that DRR conditions have a significant impact in the transportation time, it reduces dramatically with DRRs. This is also evident in Figure 6.9, where the relationship between road conditions and assessment and stabilization time is shown. Red-coded casualties are dealt within 60 minutes; yellow-coded casualties are dealt within 120 minutes. DRRs have a significant impact on red-coded casualties, but no difference is observed for yellow-coded casualties.  155  Figure 6.10 shows the relationship between elapsed time after injury and mortality rates, these numbers were observed during World War I by Marquis Moulinier (1918), and it was subsequently used by Cowley in his ―Golden Hour‖ concept. Figure 6.11 is an extension of the concept, considering the 20 yellow-coded patients from the simulation. It shows the relationship between delay in treatment and number of fatalities. Below 180 minutes there is a chance that 2 fatalities might be observed among the 20 yellow-coded casualties. Table 6.3shows a way that results can be presented. This table makes it easy for first responders to observe which option is most suitable for their objectives. This table shows What, How, Where and When to handle the situations. Options 3 and 4 show that, using DRRs, will reduce the transportation time to 31 minutes. This will give the victims almost 30 minutes to be assessed and stabilize at ERs.  6.5.2  Number of Casualties Table 6.4 shows six columns with different casualties (30, 50 and 100) and other  important parameters. It shows 5, 10 and 15 red-coded casualties with the rest assumed to be yellow-coded casualties. For these simulations DRRs were assumed for the road conditions. In this case it‘s worth noting that DRRs are supposed to be activated immediately. Figure 6.12 shows the relationship between the transportation time and the number of casualties. 100 casualties will be transported around 46 minutes; and red-coded casualties are transported in less than 30 minutes. Figure 6.13 shows the relationship between stabilization time at ERs and the number of casualties. 15 Red-coded casualties will be assessed and stabilize within 2 hours, this is a limiting factor for ER. Decision makers might consider sending the rest of the casualties to different hospitals; it will probably make a difference in yellow-coded victims. This is a decision that might be worth considering for mass casualties: red-coded casualties should be sent to nearby hospitals (VGH and SPH) and yellow-coded victims should be re-directed to further hospitals and thus their chances for survival will increase. 156  6.5.3  High Casualty Case  The following assumptions were made in the high casualty case:   There are 30 ambulances available for response. The time to send an ambulance to a venue is assumed to 5 minutes, on average.    All casualties come from BC Place    There are police stationed at each venue, along with two ambulances and EMTs capable of performing triage.    DRRs are up and traffic is clear to all hospitals    The disaster is large enough that VGH and SPH increase their capacities to 20 and 15 patients per hour, respectively (typically assumed to be 10 patients/hour)    The decision is made immediately to send casualties to any available hospital, not just SPH and VGH    The percentage distribution between hospitals has been modified to ensure that all waiting rooms empty at around the same time - that way all hospitals are used for the entire time period    Only Code Yellow Casualties are tracked in this example (in such a large disaster, Code Reds will have to be treated on site or by other means, and Code Greens will be considered as walking-wounded)    Once an ambulance is assigned to a given venue, it will cycle back to that venue from the hospitals until the driver is instructed otherwise.  Table 6.5, Table 6.6 and Table 6.7 show important results for the simulations. Figure 6.14 shows that regardless of the transportation efforts, something different needs to be done in order to reduce the transportation time, and the assessment and stabilization of victims. Field hospitals and other means of transportation should be implemented. Perhaps a different rate for casualties at ER should be considered. Figure 6.15, Figure 6.16 and Figure 6.17 reveal in detail at every hospital the amount of hours to transport and treat patients. It is likely that 283 victims could be handled at Vancouver region, but more than a 1,000 victims might compromise the response and the chance of survival unless emergency decisions are made to handle mass casualties. 157  Table 6.2 I2Sim Results for Road Conditions Road conditions BCP - VGH (1)  Road conditions BCP - SPH (1)  Road conditions BCP - VGH (2)  Road conditions BCP - SPH (2)  Road conditions BCP - VGH (3)  Road conditions BCP - SPH (3) 15  No of casualties  15  15  15  15  15  Red-coded casualties  5  5  5  5  5  5  Yellow-coded casualties  10  10  10  10  10  10  Initial assessment time (min)  2  2  2  2  2  2  Rescue time (min)  3  3  3  3  3  3  Egress time (min)  16  16  16  16  16  16  Triage time (min)  Route (streets and avenues)  5  5  5  5  5  5  Pacific Blvd  Smithe St  Pacific Blvd  Pacific Blvd  Pacific Blvd  Smithe St  Quebec St  Burrard St  Cambie St  Nelson St  Quebec St  Burrard St  2nd Av  12th Av  2nd Av  Laurel St  Laurel St  Distance of route (km)  4.4  2.2  3.8  2.1  4.4  2.2  No of ambulances  15  15  15  15  15  15  Patients per ambulance  2  2  2  2  Road conditions  Normal roads  Normal roads  DRR  DRR  Transportation time (one patient) Total transportation time (all patients) Time for red-coded patients to arrive at ER Time for red-coded patients to be assessed and stabilized in ER Time after all patients have been assessed and stabilized in ER  21  9  9  7  2 Normal roads No pedestrians 15  2 Normal roads No pedestrians 7  42  34  31  31  36  30  37  21  24  22  28  21  61  49  50  48  56  48  121  109  110  108  116  108  158  Table 6.3 Results from Road Conditions When are the casualties being handled? What happened?  a) An explosion occurs in downtown Murrin Electrical Substation. Traffic lights are out of service in major streets: Quebec, Main, Station, Union, Pacific Blvd, Keeler, Columbia, Abbot, Dunsmuir, Georgia, Smithe, Granville, Seymour, Burrard, Robson, etc.  How is the situation being handled?  opion 1 15 casualties to VGH 15 casualties to SPH  b) A multitude starts an uncontrolled egress from BC Place. c)  opion 2  d) Trapped people and casualties are reported in BC Place after the collapse of the stage.  opion 3 30 ambulances  e) Casualties are reported from the egress process at BC Place.  1 patient per ambulance opion 4  Result: 30 casualties are reported at BC Place, 20 yellow-coded, 10 red-coded opion 5  opion 6  Initial assessment  Rescue  Triage  Egress  VGH Normal roads Pacific Blvd Quebec St 2nd Av Laurel St SPH Normal roads  34 min  Burrard St VGH DRR Pacific Blvd Cambie St 12th Av SPH DRR Pacific Blvd Nelson St VGH Normal roads No pedestrians Pacific Blvd Quebec St 2nd Av Laurel St VGH Normal roads No pedestrians Smithe St Burrard St  Total Transportation Time  42 min  Smithe St [Distribution: 10 yellowcoded and 5 red-coded]  A stage collapses on top of spectators at BC Place.  Where are the ambulances traveling to?  31 min  2 min  8 min  5 min  16 min 31 min  36 min  30 min  159  Table 6.4 Results from I2Sim Simulation with 30, 50 and 100 Casualties 30 Casualties BCP - VGH  30 casualties BCP - SPH  50 Casualties BCP - VGH  50 casualties BCP - SPH  100 Casualties BCP - VGH  100 casualties BCP - SPH  No of casualties  15  15  25  25  50  50  Red-coded casualties  5  5  10  10  15  15  Yellow-coded casualties  10  10  15  15  35  35  Initial assessment time (min)  2  2  2  2  2  2  Rescue time (min)  3  3  3  3  3  3  Egress time (min)  16  16  16  16  16  16  Triage time (min) Route (streets and avenues)  5  5  5  5  5  5  Pacific Blvd  Pacific Blvd  Pacific Blvd  Pacific Blvd  Pacific Blvd  Pacific Blvd  Cambie  Nelson St  Cambie  Nelson St  Cambie  Nelson St  12th Av  Burrard  12th Av  Burrard  12th Av  Burrard  Distance of route (km)  3.8 Km  2.1 Km  3.8 Km  2.1 Km  3.8 Km  2.1 Km  No of ambulances  15  15  15  15  15  15  Patients per ambulance  2  2  2  2  2  2  Road conditions  DRR  DRR  DRR  DRR  DRR  DRR  Transportation time (one patient)  9  7  9  7  9  7  Total transportation time (all patients) Time for red-coded patients to arrive at ER (min) Time for red-coded patients to be assessed and stabilized in ER (min) Time after all patients have been assessed and stabilized in ER (min)  30  29  38  36  48  46  24  22  25  23  26  24  50  48  79  77  109  107  110  108  169  167  319  317  160  Table 6.5 283 Yellow-Coded Casualties and 7 Hospitals Hospital  VGH  SPH  Women's and Children's  Mount St. Joseph's  UBC  Burnaby General  Richmond General  No of casualties Initial assessment time (min) Rescue time (min)  108  79  28  25  14  14  14  2  2  2  2  2  2  2  3  3  3  3  3  3  3  Egress time (min)  16  16  16  16  16  16  16  Triage time (min)  5  5  5  5  5  5  5 Pacific Blvd  Pacific Blvd Pacific Blvd Route (streets and avenues)  Pacific Blvd  Cambie  Pacific Blvd  Pacific Blvd  Cambie  Main St  Broadway  Cambie  Pacific Blvd  Cambie  Cambie  Broadway  Hastings St  Granville  12th Av  Nelson St  King Edward  Broadway  University  HWY1  Marine Dr  Oak St.  Kingsway  Wesbrook  Boundary  Russ Baker Way  Kincaid St.  Gilbert Rd Westminster Hwy.  Distance of route (km)  3.8 Km  2.9 Km  5.7 Km  3.0 Km  13.0 Km  9.3 Km  15.3 Km  No of ambulances  10  7  3  2  2  2  4  Patients per ambulance  2  2  2  2  2  2  2  Route conditions Transportation time (one patient - minutes) Total transportation time (all patients - hours) Time after all patients have been assessed and stabilized in ER (hours)  DRR  DRR  DRR  DRR  DRR  DRR  DRR  5  3  7  4  13  11  15  2  1  2  1  3  2  2  6  5  6  5  6  6  6  161  Table 6.6 864 Yellow-Coded Casualties and 7 Hospitals Hospital  VGH  SPH  Women's and Children's  Mount St. Joseph's  UBC  Burnaby General  Richmond General  No of casualties Initial assessment time (min) Rescue time (min) Egress time (min) Triage time (min)  328  242  86  78  43  43  43  2  2  2  2  2  2  2  3 16 5  3 16 5  3 16 5  3 16 5  3 16 5  3 16 5  Pacific Blvd Cambie 12th Av  Pacific Blvd Nelson St  Pacific Blvd Cambie King Edward Oak St.  Pacific Blvd Cambie Broadway Kingsway  Pacific Blvd Cambie Broadway University Wesbrook  Pacific Blvd Main St Hastings St HWY1 Boundary Kincaid St.  3.8 Km 10 2 DRR  2.9 Km 7 2 DRR  5.7 Km 3 2 DRR  3.0 Km 2 2 DRR  13.0 Km 2 2 DRR  9.3 Km 2 2 DRR  3 16 5 Pacific Blvd Cambie Broadway Granville Marine Dr Russ Baker Way Gilbert Rd Westminster Hwy. 15.3 Km 4 2 DRR  5  3  7  4  13  11  15  4  3  5  4  7  6  4  16  15  18  15  17  16  15  Route (streets and avenues)  Distance of route (km) No of ambulances Patients per ambulance Route conditions Transportation time (one patient - minutes) Total transportation time (all patients - hours) Time after all patients have been assessed and stabilized in ER (hours)  162  Table 6.7 2,591 Yellow Coded Casualties and 7 Hospitals Hospital  VGH  SPH  Women's and Children's  Mount St. Joseph's  UBC  Burnaby General  Richmond General  No of casualties Initial assessment time (min) Rescue time (min) Egress time (min) Triage time (min)  985  725  259  233  130  130  130  2  2  2  2  2  2  2  3 16 5  3 16 5  3 16 5  3 16 5  3 16 5  3 16 5  Pacific Blvd Cambie 12th Av  Pacific Blvd Nelson St  Pacific Blvd Cambie King Edward Oak St.  Pacific Blvd Cambie Broadway Kingsway  Pacific Blvd Cambie Broadway University Wesbrook  Pacific Blvd Main St Hastings St HWY1 Boundary Kincaid St.  3.8 Km 10 2 DRR  2.9 Km 6 2 DRR  5.7 Km 3 2 DRR  3.0 Km 2 2 DRR  13.0 Km 3 2 DRR  9.3 Km 2 2 DRR  3 16 5 Pacific Blvd Cambie Broadway Granville Marine Dr Russ Baker Way Gilbert Rd Westminster Hwy. 15.3 Km 4 2 DRR  5  3  7  4  13  11  15  11  10  14  10  13  17  13  46  44  48  42  42  46  47  Route (streets and avenues)  Distance of route (km) No of ambulances Patients per ambulance Route conditions Transportation time (one patient - minutes) Total transportation time (all patients - hours) Time after all patients have been assessed and stabilized in ER (hours)  163  Total transportation time and road conditions after injury 30 casualties: 20 yellow-coded and 10 red-coded  Road conditions  NORMAL ROADS  NORMAL ROADS, NO PEDESTRIANS  BCP-VGH all casualties  BCP-SPH all casualties  DRR  BCP - VGH red coded casualties BCP- SPH red-coded casualties 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44  Transportation time (minutes)  Figure 6.8 Total Transportation Time and Road Conditions after Injury  Assessment and stabilization time at ER after injury 30 casualties: 20 yellow-coded and 10 red-coded  Road conditions  NORMAL ROADS  NORMAL ROADS, NO PEDESTRIANS  BCP-VGH all casualties  BCP-SPH all casualties BCP-VGH red-coded casualties BCP-SPH red-coded casualties  DRR  40  45  50  55  60  65  70  75  80  85  90  95 100 105 110 115 120 125  Assessment and stabilization time (minutes)  Figure 6.9 Assessment and Stabilization Time at ER after Injury 164  Mortality due to delay time from injury 80% 70% 60%  Mortality  50% 40% 30% 20% 10% 0% 60  120  180  240  300  360  420  480  540  600  Time ellapsed after injury (minutes)  Figure 6.10 Number of Dead Victims due to Delay in Treatment  Number of dead victims due to delay in treatment 20 yellow-coded casualties considered Number of dead victims due to delay of treatment  15 14 13 12 11 10  9 8 7 6 5 4 3  2 1 0  0  60  120  180  240  300  360  420  480  540  600  660  Time after injury  Figure 6.11 Number of Dead Victims due to Delay in Treatment 165  Transportation time after injury 50 All casualties BCP-VGH  45  All casualties BCP-SPH Red-coded casualties BCP-VGH  40 Number of casualties  Red coded casualties BCP-SPH 35  30 25 20 15 10 5 20  25  30  35  40  45  50  Transportation time (minutes)  Figure 6.12 Transportation Time and Number of Casualties  Stabilization time at ER after injury 50 Red-coded staibilzation time ER BCP-VGH 45  Red coded stabilization time ER BCP-SPH All casualties stabilized at ER BCP-VGH  Number of casualties  40  All casualties stabilized at ER BCP-SPH  35 30 25 20 15 10 5 40  80  120  160  200  240  280  320  Stabilization time (minutes)  Figure 6.13 Stabilization Time and Number of Casualties  166  Time for assessment and satibilzation at ER Hospitals at Metro Vancouver Yellow-coded casualties 283 Casualties  3  5  TRANSPORTATION DURATION ASSESSMENT AND STABILIZATION  864 Casualties  5  12  2591 Casualties  14  0  3  34  6  9  12  15  18  21  24  27  30  33  36  39  42  45  48  time (hours)  Figure 6.14 Time for Assessment and Stabilization at 7 Hospitals in Vancouver  Time for assessment and satibilzation at ER Hospitals at Metro Vancouver 283 yellow-coded casualties Richmond General  2  Burnaby General  2  UBC Hospital  3  3  Mount St. Joseph's  3  1  Children's Hospital St. Paul's  4  3 2  4  1  Vancouver General  4 2  0 TRANSPORTATION DURATION ASSESSMENT AND STABILIZATION  4 1  2  3  4  5  6  time (hours)  Figure 6.15 Time for Assessment and Stabilization Time, 283 Casualties  167  Time for assessment and satibilzation at ER Hospitals at Metro Vancouver 864 yellow-coded casualties Richmond General  4  10  Burnaby General  6  10  UBC Hospital  7  Mount St. Joseph's  10  4  11  Children's Hospital  5  St. Paul's  12  3  Vancouver General  11 4  12  0 1 2 TRANSPORTATION DURATION ASSESSMENT AND STABILIZATION  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  time (hours)  Figure 6.16 Time for Assessment and Stabilization, 864 Casualties  Time for assessment and satibilzation at ER Hospitals at Metro Vancouver 2,591 yellow-coded casualties Richmond General  13  Burnaby General  33 17  UBC Hospital  29  13  Mount St. Joseph's  29  10  32  Children's Hospital  14  St. Paul's  34  10  Vancouver General  34  11 0  3  6  Transportation duration assessment and stabilization  34 9  12  15  18  21  24  27  30  33  36  39  42  45  48  time (hours)  Figure 6.17 Time for Assessment and Stabilization, 2,591 Casualties  168  6.6 Implications of the Study As a planning tool, I2Sim provides emergency managers with the ability to uncover hidden interdependencies between critical infrastructures and to determine key vulnerabilities that arise from them. This enables the development and refinement of mitigation strategies that address these vulnerabilities before they become a problem. Within a given operational period, the use of I2Sim at the provincial operations level allows managers and responders to keep track of the intricate web of interconnected infrastructures in order to make more informed tactical decisions regarding the deployment of limited resources. The model of Downtown Vancouver developed in I2Sim can reveal interdependencies and cascading effects as a consequence of a disaster. The model can be used to perform sensitivity analysis and provide answers to specific operational questions. Table 6.1 provides a template to relate input and output variables of interest to the study. This arrangement is aimed at helping the user consider the input and output parameters to be used in the model; this constitutes a pre-processing activity for the model. Some of these parameters may lead to further refinements, as for example, in the case of the time variables: initial assessment, rescue process, egress and triage. The I2Sim user needs to make a number of ―reasonable‖ assumptions related to the evolution of the scenarios. Triage of mass casualty during disasters differs in important ways from traditional triage processes. CPR is not performed, unless sufficient resources are available without jeopardizing the lives of other victims. In terms of triage for medical casualties, first responders and EMS staff are generally trained in a triage system known as START (simple triage and rapid treatment). The triage process assumed in this simulation takes into consideration the fact that a dynamic triage is in place at a time in BC Place. In the triage process it is assumed that blackcoded victims are separated from the rest of the casualties; green-coded victims are released at the scene (walking wounded); yellow-coded and red-coded victims are transported to the ERs. The transportation priority is red-coded casualties, those casualties will be transported first and then yellow-coded ones. 169  Three road conditions were investigated and timeframes for casualty treatment were also defined. These road conditions were set up in order to verify their impact in the delay of time for treatment. ―Normal roads and no pedestrians‖ refers to roads during the 2010 Winter Olympics in Vancouver. It is assumed that emergency responders manage to restrain people from accessing the roads and streets where emergency vehicles are circulating. This is a condition that might be enforced during the Olympics, since there will be dedicated pedestrian streets in downtown Vancouver. The I2Sim user has to realize that the process for transporting victims from the place where the disaster occurs to the ER in a nearby hospital, involves several activities. Table 6.3 is designed to make it easy for first responders to observe which option is suitable for their objectives. This table shows the What, How, Where, and When of the situations. This is part of the post-processing activities; it is worth noting that the activities of pre and post processing are very important. They provide the first responders to assess the resources needed for the preparation and recovery activities.  170  7 Conclusions The I2Sim simulator was developed at UBC, under federal government funding, to assess the consequences of a natural or man-made disaster; to facilitate pre-event planning, mitigation, the optimization of post-event emergency response and recovery measures. This thesis makes two major contributions to the capability of the simulator, to handle seismic events and events that affect dense concentrations of people, such as a crowd in a football stadium. The distinguishing characteristic of an earthquake event can affect the city and all the surrounding regions, causing damage to all lifeline systems, such as water, power, transportation and medical services. In its original form, I2Sim could model the damage and impact of each system on its own, but was unable to account for the effects of all other systems. The interdependency between systems is a crucial element for determining the impact of an earthquake and the time for recovery. The thesis describes the development of a procedure to cope with the interdependencies between the lifeline systems. The core of this procedure is an interaction matrix which quantitatively accounts for the interaction of one lifeline system and another. UBC Campus. An example of how important this interaction is provided by the simulation of the seismic consequences for the UBC Campus, the Test Case used to validate the procedure. The functionality of the water system – considering the Structure and the Non Structural components - was reduced to 55 % for an earthquake with an intensity of II = IX. When the impact of other systems on the water system – Reservoir, Transmission and Distribution water pipelines - were taking into account, the post-earthquake functionality of the water system was reduced to 0% at the same intensity of shaking. It is especially crucial to take interdependency into account when evaluated the functionality of hospitals, which are critical elements in any recovery process. The functionality of the hospital at UBC, based only on structural and non-structural damage is 50%, but when the impact of the functionality to other critical systems, such as water, power, steam and transportation of casualties, medical personnel and medical supplies are taken into account the overall functionality of the hospital drops to 21 % for a shaking intensity of II = IX.  171  Therefore I2Sim enhanced by the interaction matrix can produce very realistic pictures of the state of the events, given the relevant data. The damage estimates discussed so far are based on data from past earthquakes as interpreted by the experts in each particular system. Once an event has occurred and real data become available, I2Sim can produce updated estimates of the functionality of each system. This dynamically changing knowledge is the basis for optimizing response measures, and the deployment of resources to ensure the best wellness outcome to the entire system. GIS mapping was used, and the following maps were created:   Building maps (with related attributes: area, year of construction, etc.)    Critical Infrastructure maps (for water, electrical, roads and other CI)    Damage estimation maps (showing relevant information on risk calculations)    Distributed damage information maps (Single Infrastructure maps - every CI system will have a damage estimation map that do not include interdependencies)    Interdependency maps (showing all damage in CI systems, or subsystems)  Single Infrastructure maps are useful for local operators (operational agents). With them they can locate damages and proceed with repair activities. Interdependency maps will show that the impact to other systems is relevant; and they will help managing the disaster for global operators, because they can establish action plans for the affected systems. With this information, managers and stakeholders will be able to allocate resources and re-established critical assets according to the objective function of UBC Campus. Downtown Vancouver Scenario. The second significant enhancement of I2Sim are the incorporation of an Egress Model to simulate the consequences of panic in a large crowd caused by a major threatening event, such as a terrorist attack; and a traffic model, with the capability to simulate: normal roads; Disaster Response Routes (DRR), which are free of general traffic and pedestrians; and streets and avenues full of crowds and pedestrians. DRR conditions have a significant impact in the transportation time. These particular models were tested by examining a realistic event, postulated for testing responses to an event happening during the Vancouver Olympic Games. This scenario event considered an explosion wrecking a power substation, resulting in a power failure in BC Place 172  Stadium during a rock concert with 70,000 people. In the initial panic it was postulated that a stage would collapse and injured spectators. Thereafter casualties would result for the frantic efforts of the crowd to exit the stadium. I2Sim modeled all the significant elements subsequent to the disaster: recovery of injured from collapsed stage, removal of those injured during egress, triage and classification of casualties for prioritization for treatment, and subsequent transportation to hospital under conditions that serious casualties had to be stabilized within one hour. Longer stabilization times than this resulted in rapidly increasing mortality rates. I2Sim optimize the routes to the hospital taking into account postulated conditions along the routes. There were some dedicated emergency routes, along which no parking was allowed during the Olympics, these routes gave the shortest time for patient delivery. I2Sim assessed the capability of the nearby hospitals to handled serious casualties, within the desired time limit, and determined when other hospitals in the area had to be called into service. Finally, I2Sim established the casualty limit when emergency medical treatment service would be necessary for treatment of serious casualties within time limits to keep mortality low.  7.1 Future Work 7.1.1  Disaster Response Network There is a need to link together a network of expert centres, geographically dispersed,  into an integrated virtual community with the capability to prepare and respond effectively to large disaster emergencies. Disasters can happen anywhere in the world. However, the expertise to analyze the vast and complex amounts of information related to disaster events may reside in specialized research and analysis centres that may be located far away. The diversity and complexity of the disaster events, e.g., earthquakes, tsunamis, hurricanes, ice storms, terrorist attacks, floods, forest fires, and others requires specialized and diverse knowledge which is normally not available at a single location. There are two aspects to the capability of effectively responding to large disasters, one the actions of the onsite managers and responders that are directly in contact with the particulars of the situation, and two the capability of the expert support community to assess the effects of the interactions among multiple infrastructure systems (electricity, water, 173  transportation, etc.) that support the preparation, response and recovery actions. To bring the local responders and dispersed knowledge support communities together requires to be able to share in real time large amounts of geographical and infrastructure data. Large bandwidths of secure and reliable data communication are required to make sharing the disaster data possible. Large bandwidths of data communication are also required to create realistic ―virtual disaster response room‖ environments where experts can simultaneously manipulate the data and discuss possible best actions for training and real-time response situations. The poor response to disasters, like the Southeast Asia tsunami and hurricane Katrina in New Orleans, illustrates the need for integration and coordination of the reconnaissance of the ground situation with the knowledge and expertise to assess the global situation and make best decisions. There is a need to bring together the capability of forming an international network of disaster experts and responders that can cooperate across continents for the purpose of best response and saving human lives. Technically, the following challenges are addressed: a) Large amount of data that need to be shared across large distances in real time; b) Expertises that need to be coordinated across a number of centres globally, c) Capability of recognition, enrichment, and best use of large volumes of data for multiple disaster events, and d) Interfaces, processing, and storage capabilities that allow the creation of virtual disaster room facilities for the interaction of experts, decision makers, and responders across continental distances. The Complex Interdependent System Group at UBC has already defined a Disaster Response Network called DR-NEP (Disaster Response Network Enabled Platform), whose capabilities were defined at two levels: Dynamic Training Environment: The primary functions of DR-NEP are dynamic in nature. A major function is for ―live‖ table top exercises involving virtual participants at multiple nodes in the platform. During these exercises, tools and expertise from the multiple teams around the world can be brought together to develop realistic and complex scenarios which will then form part of a knowledge database. In another application, the scenarios will become training exercises for disaster managers and responders. These training exercises will be augmented through real-time remote interaction between the technical experts and the trainees. In addition, the DR-NEP system will allow increased communication between these centres of 174  expertise allowing them to share results and enhance the study of past disaster events, as well as the improvement of future models. Real-Time Disaster Response: The DR-NEP system will be able to provide live support during ongoing disasters. This will be an advisory role to support managers and responders in the area. This application will be most useful after the initial response to the disaster, which is the responsibility of the local responders. After the initial hours of a global or major event, strategies need to be devised as how to best restore the system. This recovery period can last days, weeks, or months, depending on the disaster. The real-time capacity of DRNEP will allow the expertise and database cases in the DR-NEP data repository to be used as starting points of discussion in the recovery efforts. Continued real-time updating of the ground situation will allow the DR-NEP support system to use its simulation tools and response strategies to suggest best modes of recovery to the local authorities.  7.1.2  Design of Sustainable and Resilient Communities The objective of Smart Energy Communities is to optimize the use of locally available  energy resources to form self-sufficient resilient communities that preserve the sustainability of the earth resources and the livability of the environment. The simultaneous interactions among multiple energy infrastructure systems can be modeled with I2Sim. A main concern of any community is to have a sustainable and resilient region. To address this concern, the Complex Interdependent System (CIS) group of the Faculty of Applied Science of UBC has addressed two main objectives: Objective 1. Economical Analysis and Optimization of the Carbon Footprint in any region. To minimize the carbon footprint, energy systems and critical buildings can be modeled with I2Sim; alternative scenarios will be considered to maximize the reduction in the carbon footprint within acceptable cost differentials. Objective 2. Self-Sufficiency and Resiliency design of the region’s Infrastructures. This analysis will involve detailed models of buildings and infrastructures of the region with the purpose of assessing survivability and resiliency of critical systems. 175  Both Objectives 1 and 2 will require the integration of I2Sim with the possible multiple smart energy sub-systems provided by external vendors. This integration will make possible the global optimization of multiple interdependent sub-systems.  7.1.3  Extended Objective Functions for an Infrastructure Interdependent Simulator (I2Sim) Ultimately, the capabilities of the methodology and the simulator shall be extended to  design better communities, that will be sustainable and resilient, and the methodology should be extended to incorporate the following objective functions: 1. Life safety   Health care    Shelter    Fulfillment of basic needs (food, water, security, etc.)  2. Business continuity   Service    Economical income    Reliability of services and Critical Infrastructure systems  3. Sustainability and environmental   Smart energy communities    Reduction of carbon footprint    Use of available local energy sources and materials  4. Design of Critical Infrastructure, Social or Financial networks  176  7.2 Final Remarks I have learned about the progress made in multi-hazard and risk assessment through earthquakes studies; the enhancement of our mathematical tools to assess the seismic performance that our structures will withstand; the importance of uncertainties and probability studies. We have bridged the gap between design spectra to time history analysis; we have evolved from elastic models to inelastic ones. We have made tremendous progress in monitoring the overall behaviour of structures; we have arrived at 1 to 1 real testing models in spectacular shake table facilities. We have arrived at complex damage indices, and alternate ways to measure the behaviour of our structures; we have produced machinery to reduce shock and vibration, and that allegedly will reduce the seismic forces in specialized buildings. But we are still short in our efforts; we still stand paralyzed when extended damage has occurred through utilities that were not resilient enough. And most of all, human lives during catastrophic events will be lost. The scary part is to realize that our world will suffer the consequences of losing precious utilities in a disaster. Technology is at stake when a failure occurs, and even a lack of basic survival tokens might render the whole society to a complete stand still. Disasters are inevitable, and damages will affect our everyday lives. We have to recognize these events as facts, and be prepared. Risk calculations and damage estimations need to evolve to Interdependency estimations. Multi-Hazard Interdependency Assessment is the next step; this methodology will help in three phases of the disaster: planning, response and recovery. Single and Global Interdependencies will show the level of affliction of one utility service to an entire community and other utility services, and also the impact on the community. Interdependencies can also be shown through Interdependency mapping. GIS tools have been used extensively in the last few years to show information, but as mentioned before, distributed damage estimation maps are not enough for emergency planners; we need to evolve to dynamic, distributed and coordinated Interdependency maps. Single and Global Interdependencies will help to arrive to Resiliency Graphs. 177  I2Sim is a promising tool, and it was enhanced with interdependency estimations. It is now possible to uncover interdependencies, and to have a real-time follow up of the recovery phase.  178  References Alexoudi, M. N.; Kakderi, K. G. and Pitilakis, K. D. (2008). Seismic Risk of Interdependent Lifeline System using Fuzzy Reasoning. Proceedings of the 14 th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Allen, C. R.; Bolt, B. A.; Hales, A. L.; Hamilton, R. M.; Handin, J. W.; Housner, G. W.; Hudson, D. E.; Kisslinger, C.; Oliver, J. E.; Steinbrugge, K. V.; Berg, J. W.; and Bove, A. N. (1971). The Joint Panel on the San Fernando Earthquake. The San Fernando Earthquake of February 9, 1971. Lessons from a Moderate Earthquake on the Fringe of a Densely Populated Region. Nationall Academy of Sciences. National Academy of Engineering. Washington, D.C. ATC (1985) ―Earthquake Damage Evaluation for California‖, ATC-13, Applied Technology Council, Redwood City, CA, 492 p. ATC (1988). Rapid Visual Screening of Buildings for Potential Seismic Hazards: A Handbook. Applied Technology Council, (ATC-21). Redwood City, California, 1988. ATC (1991). Seismic Vulnerability and Impact of Disruption of Lifelines in the Conterminous United States. Applied Technology Council, (ATC-25 or FEMA 224). Federal Emergency Management Agency, 1991. Atkinson, G. M. and Kaka, S. I. (2007). Relationship between felt intensity and instrumental ground motion in the central United States and California, Bull. Seismol. Soc. Am. 97, no. 2, 497–510. Atkinson, G. M. and Wald, D. J. (2007). ―Did You Feel It?‖ intensity data: A surprisingly good measure of earthquake ground motion. Seismological Research Letters 78, 362–368. Auf der Heide, E. (2004). Common Misconceptions about Disasters: Panic ―the Disaster Syndrome and Looting‖. Available at: http://atsdr1.atsdr.cdc.gov/emergency_response/common_misconceptions.pdf. 179  Ballantyne, D. (2003). Multi-Hazard Risk Assessments of Water Systems, Elements in Common with Seismic, Security, and Other Risk Studies, Proceedings of the Third US-Japan Workshop on Water Systems Seismic Practices, organized by the Los Angeles Department of Water and Power, supported by the AwwaRF, August 6-8. Barbera, J. and McIntyre, A. (2004). Medical Surge Capacity and Capability: A Management System for Integrating Medical and Health Resources During Large-Scale Emergencies. CNA Corporation Web site. http://www.cna.org/documents/mscc_aug2004.pdf. Bell, L. (1998). ―Building prototypes, seismic damage probability matrices for buildings located in Southwestern BC‖. Report prepared for The Department of Civil Engineering University of British Columbia, BC, Canada. Benson, M.; Koenig, K.L. and Schultz, C.H. (1997). ―Disaster triage: START, then SAVE: a new method of dynamic triage for victims of a catastrophic earthquake‖. Prehospital Disaster Med; Vol 11 No 2: pp 117-24. Benuska, L. (1990). Loma Prieta Earthquake Reconaissance Report. Earthquake Engineering Research Institute. Earthquake Spectra, Special Supplement to volume 6, May 1990. Volume 6, Issue S1, 448 p. Blanquera, A. (1999). ―Evaluation of structural earthquake damage to buildings in Southwestern BC.‖ M.Sc. Thesis Department of Civil Engineering, University of British Columbia, Vancouver, BC. Brink, S. A.; Davidson, R. A. and Tabucchi, T. H. P. (2010). Strategies to reducedurations of post-earthquake water service interruptions in Los Angeles. Structure and Infrastructure Engineering. Available from: http://dx.doi.org/10.1080/15732470903517975. Browning, C. R. and Cagney, K. A. (2003). Moving Beyond Poverty: Neighborhood Structure, Social Processes and Health. Journal of Health and Social Behaviour, 44 (4), 552-571.  180  Bruneau, M. and Reihorn, A. (2007). Exploring the concept of seismic resilience for acute care facilities. Earthquake Engineering Research Institute. Earthquake Spectra, volume 23, Number 1, pp. 41-62. Bruneau, M. S.; Chang, R.; Eguchi, G.; Lee, T.; O‘Rourke, A.; Reinhorn, M.; Shinozuka, K.; Tierney, E.; Wallace, W. and von Winterfelt, D. (2003). A framework to quantitatively assess and enhance the seismic resilience of communities. Earthquake Engineering Research Institute. Earthquake Spectra, volume 19, Number 4, pp. 733-752. Casalicchio, E.; Galli, E. And Tucci, S. (2008). Modeling and simulation of Complex Interdependent Systems: a federated agent-based approach. CRITIS, Rome, Italy. Chang, S. E.; McDaniels, T. L. and Beaubien, C. (2009). Societal Impacts of Infrastructure Failure Interdependencies: Building an Empirical Knowledge Base. Lifeline Earthquake Engineering in a Multihazard Environment. Proceedings of the 2009 ASCE Technical Council on Lifeline Earthquake Engineering Conference Chang, S. E.; McDaniels, T. L.; Mikawoz, J. and Peterson, K. (2007). Infrastructure failure interdependencies in extreme events: power outage consequences in the 1998 Ice Storm. Natural Disasters 41(2007): 337-358. Chung, R. (1996). The January 17, 1995 Hyogoken-Nanbu (Kobe) Earthquake. Performance of Structures, Lifelines, and Fire Protection Systems. NIST Special Publication 901. Buildings and Fire Research Laboratory. National Institute of Standards and Technology. Clague, J. and Turner, B. (2003). Vancouver, City on the Edge; Living with a Dynamic Geological Landscape. Tricouni Press, Vancouver, Canada. Cook, S. E. (1999). ―Evaluation of non-structural earthquake damage to buildings in Southwestern BC.‖ M Sc. Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, BC.  181  Cornell, C. A. (1968). Engineering seismic risk analysis. Bulletin of the Seismological Society of America; October 1968; v. 58; no. 5; p. 1583-1606 Corotis, R. B. and Hammel, E. (2008). The Role of Risk Perception in Tradeoffs for Earthquakes and Other Hazards. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Davidson, R. and Cagnan, Z. (2004). Restoration modeling of lifeline systems. In Research Progress and Accomplishments, 2003–2004. Buffalo, NY: MCEER. Delaney, J.S.; Drummond, R.; Ducharme, J. and Marx, R.G. (2002). ―Mass casualties and triage at a sporting event‖. British Journal of Sports Medicine; Vol 36: pp85–88 DELCAN (1994). Screening Study. With respect to Behaviour under Seismic Loading of University Controlled Campus Buildings. Campus Planning Department, University of British Columbia. Dowrick, D. J. (2003). Earthquake Risk Reduction. John Wiley & Sons Ltd., West Sussex, England. Dudenhoeffer, D. D.; Permann, M. R. and Manic, M. (2006). ―CIMS: A Framework for Infrastructure Interdependency Modeling and Analysis‖, Proceedings of the 2006 Winter Simulation Conference, IEEE, December 2006, pp 478– 485. Dueñas-Osorio, L. and Vemuru, S.M. (2007). Risk Assessment of Complex Infrastructures. Special Workshop on Risk Acceptance and Risk Communication, Stanford University Dueñas-Osorio, L. and Vemuru, S.M. (2009). Cascading failures in complex infrastructure systems. Structural Safety, 31 (2): 157-167 Dueñas-Osorio, L.; Craig, J. I.; Goodno, B. J. and Bostrom, A. (2007a). Interdependent response of networked systems. Journal of Infrastructure Systems 13 (3), 185–194.  182  Dueñas-Osorio, L.; Craig, J.I. and Goodno, B.J. (2007b). Seismic response of critical interdependent networks. Earthquake Eng Struct Dyn 2007;36(2): 285–306. ELCON (2004). ―The Economic Impacts of the August 2003 Blackout,‖ Electric Consumer Research Council (ELCON), February 2, 2004. http://www.elcon.org/Documents/EconomicImpactsOfAugust2003Blackout.pdf FEMA (2002). Rapid Visual Screening of Buildings for Potential Seismic Hazards: A Handbook. Second Edition. Federal Emergency Management Agency. FEMA/NIBS (1997). ―Earthquake Loss Estimation Methodology – HAZUS‖, Technical Manual, Vol. 1, Federal Emergency Management Agency (National Institute of Building Sciences) Document Number 5201. FEMA/NIBS (2003). HAZUS-MH MRI Earthquake Model Technical Manual, FEMA Washington DC. (2003). Figueroa, J. (1963). "Isosistas de macrosismos mexicanos", Report Nº 67, Instituto de Ingeniería, UNAM (in spanish). Figueroa, J. (1970). "Catálogo de sismos ocurridos en la República Mexicana", Instituto de Ingeniería, UNAM (in spanish). Fundación ICA A. C. (1988). Catálogo de Temblores que han afectado al Valle de México del siglo XIV al XX. Fundación ICA and Editorial Limusa, 258 p. (in spanish). Furuto, T.; Yamada, H. and Kakumoto, S. (2008). Development of Evacuation Support System in the Areas Affected by Multi-Hazards. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Gómez, A.; Juárez, H. and Iglesias, J. (1991), ―Intensidades y demandas de ductilidad de sismos recientes en la ciudad de México‖, Revista de Ingeniería Sísmica, SMIS, No. 43, pp. 318, (in Spanish). 183  Haimes, Y. Y.; Horowitz, B. M.; Lambert, J. H.; Santos, J. R.; Lian, C. and Crowther, K. G. (2005). Inoperability input-output model for interdependent infrastructure sectors. I: Theory and methodology, J. Infrastruct. Syst., 11(2), 67-79. Haimes, Y.Y. (2008). Models for risk management of systems of systems. International Journal of System of Systems Engineering. Volume 1 - Issue 1/2. Hall, J. F. (1995). Northridge Earthquake of January 17, 1994: Reconnaissance Report. Earthquake Engineering research Institute. Supplement C to Volume 11, Earthquake Spectra, April 1995. Volume 11, Issue S2, 521 p. Heath, R. (1995). The Kobe earthquake: some realities of strategic management of crises and disasters. Disaster Prevention and Management, Volume 4, Number 5, pp. 11–24. Hirshberg, A.; Holcomb, J.B. and Mattox, K.L. (2001). Hospital Trauma Care in Multiple Casualty Incidents: a Critical View. Ann Emerg Med 2001; 37: 647–52. Hollman, J.A.; Martí, J.R.; Jatskevich, J. and Srivastava, K.D. (2007). Dynamic Islanding of Critical Infrastructures, a Suitable Strategy to Survive and Mitigate Critical Events. International Journal of Emergency Management, Vol. 4, No. 1, pp. 45-58. http://www.ece.ubc.ca/~jiirp/JIIRP_Open_Publications/jiirp_i2c_038.pdf Hosseini, M. and Vayeghan, F. Y. (2008). A Risk Management Model for Inter-City Road Systems. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. http://www.trauma.org/archive/history/resuscitation.html Iglesias, J. (1989). The Mexico Earthquake of September 19, 1985- Seismic Zoning of Mexico City after the 1985 Earthquake. Earthquake Spectra, Vol 5, Issue 1, pp. 257-271.  184  Iglesias, J.; Manuel, J.; Aguilar, J.; Mota, O.; Terán, A.; Gómez Bernal, A.; Ortega, R.; González, E.; Guerrero, J. J.; Juárez García, H.; Paniagua, J. P.; and Sordo, E. (1987). Estudio de las intensidades del sismo del 19 de septiembre en la Ciudad de México. México; México. Universidad Autónoma Metropolitana. Unidad Azcapotzalco; 69 p (in spanish). Ixtlixochitl, F. de A. (1975). Obras Históricas. México, U.N.A.M. (serie Historiadores y Cronistas de Indias), I, 264-265 (in spanish). Japan Building Disaster Prevention Association (1977). Standard for Evaluation of Seismic Capacity of Existing Reinforced Concrete Buildings. Jennings, P. C. (1971). Engineering Features of the San Fernando Earthquake of February 9, 1971. Technical Report: CaltechEERL:1971.EERL-71-02. California Institute of Technology. Juárez García, H.; Whitney, R. A.; Gama, A.; Vera, R.; and Hurtado, F. (1997). The October 9, 1995 Manzanillo, México Earthquake, Seismological Research Letters, V 68, n 3 pp. 413 – 425. Kankanamalage, R.; Rostamirad, S.; Juárez García, H.; Martí, J.R. and Srivastava, K.D. (2011). ―Transportation Model for an Infrastructure Interdependent Simulator (i2Sim)‖, paper in progress. Kramer, S. L. (1996). Geotechnical Earthquake Engineering. Prentice-Hall Civil Engineering and Engineering Mechanics Series, Upper Saddle River, N.J. Prentice Hall. Kuo, K.C.; Banba, M. and Suzuki, Y. (2008). Loss Analysis of Medical Functionality Due to Hospital‘s Earthquake-Induced Damages. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008.  185  Lee, Y. and Graf, W. P. (2008). Development of Earthquake Scenarios for Use in Earthquake Risk Analysis for Lifeline Systems. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Lian, C. and Haimes, Y. Y. (2006). Managing the risk of terrorism to interdependent systems through the dynamic inoperability input-output model, Syst Eng, 9(3), 241–258. Lian, C.; Santos, J. and Haimes, Y. Y. (2007). Extreme risk analysis of interdependent economic and infrastructure sectors. Risk Analysis, 27, 1053–1064. Love, R.J.; Ballantyne, D.; Basoz, N.; Eadie, Ch.; Goltz, J.; Greene, M.; Meyer, J.; Tierney, K.; Tubbesing, S.; Wyllie, L.; and Youd, T. L. (2004). Learning from earthquakes. The EERI Learning from Earthquakes Program: a Brief Synopsis of Major Contributions. Earthquake Engineering Research Institute, California, 31 p. Mallonee, S.; Shariat, S.; Stennies, G.; Waxweiler, R.; Hogan, D. and Jordan, F. (1996). Physical Injuries and Fatalities Resulting from the Oklahoma City Bombing. JAMA, 276(5), 382-387. Marquis Moulinier, S. P. (1918). ―Da Shock Tramatique dans les blessures de Guerre, Analysis d'observations‖. Bull. Med. Soc. Chir., 1918, 44: 205. Martí, J.R.; Hollman, J.A.; Ventura, C. and Jatskevich, J. (2005). Transportation Matrix Model for Infrastructures Disaster Coordination. JIIRP Seminar presentation, Vancouver, BC, Canada. Martí, J.R.; Hollman, J.A.; Ventura, C.E. and Jatskevich, J. (2008a). Dynamic recovery of critical infrastructures: real-time temporal coordination. International Journal of Critical Infrastructures (IJCI), Vol. 4, Nos. 1/2, pp. 17-31. http://www.ece.ubc.ca/~jiirp/JIIRP_Open_Publications/jiirp_i2c_043.pdf  186  Martí, J.R.; Ventura, C.E.; Hollman, J.A.; Srivastava, K.D. and Juárez Garcia, H. (2008b). I2Sim Modelling and Simulation Framework for Scenario Development, Training, and Real-Time Decision Support of Multiple Interdependent Critical Infrastructures during Large Emergencies, NATO, RTA/MSG Conference on ―How is Modelling and Simulation Meeting the Defence Challenges Out to 2015?‖ – Vancouver, Canada. McGuinn, M.G. (2004). Cross Sector Interdependencies and Risk Assessment Guidance. Final Report and Recommendations by the Council. The National Infrastructure Advisory Council (NIAC), Mellon Financial Corporation. McGuire, R. K. (2004). Seismic Hazard and Risk Analysis, Earthquake Engineering Research Institute, California, USA. 2004, MNO-10, xvii, 221p Mexico City Building Code: NTCS-04 (2004) ―Normas Técnicas Complementarias para Diseño por Sismo‖, Gaceta Oficial del Distrito Federal, Tomo II, No. 103-BIS, October, pp. 5577, (in spanish). Mitsunari, K.; Ebi, T.; Takahashi, Y.; Amano, R.; Hirayama, Y.; Akoshima, M. and Takahashi, T. (2008). A Development of Disaster Management Supporting System for Earthquake Taking Road Networks into Account. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Molyneaux, G. (2002). British Columbia: An Illustrated History. Raincoast Books, Vancouver, British Columbia, Canada; 154 p. Mujumdar, V. (2008). Effective Seismic Risk Reduction Through Integrated Interdisciplinary Research. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008.  187  Mustard, P. S.; Clague, J. J.; Woodsworth, G. J.; Hickson, C. J.; Jackson, L. E., Jr; Luternauer, J. L.; Monger, J. W. H.; Ricketts, B. D.; Turner, R. J. W.; Hunter, J. A. and Monahan, P. A. (1998). Geology and geological hazards of the Greater Vancouver area; in, Urban geology of Canadian cities. Karrow, P. F. (ed.); White, O. L. (ed.). Geological Association of Canada, Special Paper 42, 1998; pages 39-70(GSC Cont.# 18894). NATO (2002), ―Code of Best Practice for Command and Control Assessment‖, NATO Research and Technology Organisation (RTO), Command and Control Research Program (CCRP). Noreña, F.; Castañeda, C. and Iglesias, J. (1989). The Mexico Earthquake of September 19, 1985-Evaluation of the Seismic Capacity of Buildings in Mexico City. Earthquake Spectra, Vol 5, Issue 1, pp. 19-24. Nozik, K. (2003). Eastern Power Outage Unfortunate but Entirely Predictable. Press release, Aug 14, 2003. Rocky Mountain Institute. O‘Hanley, J. A. (1993). Earthquake Risk Mitigation of Hospital Facilities: a Case Study of Vancouver General Hospital. University of British Columbia. O‘Rourke, T. D. and the Steering Committee of the EERI Workshop on Contributions of Earthquake Engineering, Seismology, and Social Science (2008). Contributions of Earthquake Engineering to Protecting Communities and Critical Infrastructures from Multihazards. Earthquake Engineering Research Institute. 74 p. O‘Rourke, T.D. (2007). Critical Infrastructure, Interdependencies, and Resilience. Bridge, National Academy of Engineering, 37(1):22–29. O‘Rourke, T.D.; Lembo,A.J. and Nozick L.K. (2003). Lessons Learned from the World Trade Center Disaster about Critical Utility Systems. Pp. 269–290 in Beyond September 11th: An Account of Post-Disaster Research. Special Publication 39. Boulder, Colo.: Natural Hazards Research and Applications Information Center.  188  Office of Critical Infrastructure Protection and Emergency Preparedness (OCIPEP) (2002). The September 11, 2001 Terrorist Attacks – Critical Infrastructure Protection Lessons Learned. Incident Analysis, Number: IA02-001, September 27, 2002. Government of Canada, 23 p. Onur, T. (2001), ―Seismic Risk Assessment in Southwestern British Columbia‖, PhD. Thesis, The Department of Civil Engineering, University of British Columbia, Vancouver, BC. Onur, T.; Ventura, C.E. and Finn, W.D.L.: (2005) ―Regional Seismic Risk Assessment in British Columbia damage and loss distribution for Victoria and Vancouver‖. Canadian Journal of Civil Engineering, Vol. 32, p. 361-371, NRC Canada. Orozco y Berra, M. (1888). Historia antigua y de la conquista de México. México, Tipográfica de Gonzalo A. Esteva, III, 169 p. (in spanish). PCCIP (1997). President‘s Commission on Critical Infrastructure Protection, Critical Foundations: Protecting America‘s Infrastructures. PCCIP (2001). President‘s Commission on Critical Infrastructure Protection, Report on the Federal Agenda in Critical Infrastructure Protection Research and Development. Preliminary Seismic Microzonation Assessment for British Columbia (1994). The Province of British Columbia, Published by the Resources Inventory Committee, Version 1, Canada. Public Safety Canada (2008). Working Towards a National Strategy and Action Plan for Critical Infrastructure. http://www.ps-sp.gc.ca/prg/em/cip/strat-part1-eng.aspx Quarantelli, E. L. (1987) 'Disaster Studies: An Analysis of the Social Historical Factors Affecting the Development of Research in the Area', International Journal of Mass Emergencies and Disasters 5: 285-310.  189  Reitherman, R. (2000). Application of Earthquake Engineering Information in Hospital Emergency Response and Recovery. Consortium of Universities for Research in Earthquake Engineering. Rinaldi, S. M. (2004). Modeling and Simulating Critical Infrastructures and Their Interdependencies. Proceedings of the 37th Hawaii International Conference on System Sciences. Rinaldi, S.; Peerenboom, J. and Kelly, T. (2001). Identifying, Understanding, and Analyzing Critical Infrastructure Interdependencies. IEEE Control Systems Magazine, IEEE, December 2001, pp. 11-25. Rojas, T.; Pérez, J. M. and García, V. (1987). Y volvió a temblar, cronología de los sismos en México (de 1 pedernal a 1821). México, CIESAS. Cuaderno de Casa Chata, 135, 53 p. (in spanish). Rostamirad, S.; Kankanamalage, R.; Juárez García, H.; Martí, J.R. and Srivastava, K.D. (2011). ―Egress Model for an Infrastructure Interdependent Simulator (i2Sim)‖, paper in progress. Sack, R.; Luke, B.; Ebrahimpour, A. and Keller, J. R. (2006). Seismic risk assessment of essential buildings in Clark County, Nevada; Proc. 8th U.S. Nat. Conf. on Earthquake Eng.; Earthquake Eng. Res. Inst. Oakland, CA; CD-ROM 10 p. Scawthron, C.; Khater, M. and Rojahn, C. (1991), Seismic Vulnerability and Impact of Disruptions of Lifelines in the Coterminous United States. ATC-25 report prepared for FEMA and ATC, FEMA 224, Washington, DC, U.S. Government Printing Office. Schultz C.H.; Koenig, K.L. and Noji E.K. (1996). ―A Medical disaster response to reduce immediate mortality after an earthquake‖. Massachussets Medical Society; The New England Journal of Medicine, Vol 334, No. 7: pp 438-444  190  Secretaría de Hacienda y Crédito Público (2009). Press release 023/2009: ―El Gobierno Federal anuncia apoyos económicos emergentes para enfrentar el brote de influenza‖. May 5, 2009, Secretaría de Hacienda y Crédito Público, Mexican Government, www.hacienda.gob.mx, (in spanish). Shaw, C. and McKay, H. (1942/69). Juvenile Delinquency and Urban Areas. Rev. edn 1969. Chicago, IL: University of Chicago Press. Shi, P.; O‘Rourke, T.D.; Wang, Y. and Fan, K. (2008). Seismic Response of Buried Pipelines to Surface Wave Propagation Effects. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Shultz, J. M.; Espinel, Z.; Hick, J. L.; Galea, S.; Shaw, J. A. and Miller, G.T. (2006). SURGE SORT SUPPORT: Disaster Behavioral Health for Health Care Professionals. Miami FL: Center for Disaster & Extreme Event Preparedness, 83 p. Sordo, E.; Gómez Bernal, A.; Juárez García, H. and others (1995). The September 14, 1995, Ometepec Mexico Earthquake. Earthquake Engineering Research Institute Newsletter, December, vol 29, no 12. Stein, M. and Hirshberg, A. (1999). Medical Consequences of Terrorism: the Conventional Weapon Threat. Surg Clin North Am 1999; 79:1537-1552. Steinbrugge, K. V. (1982). Earthquakes, Volcanoes, and Tsunamis: an Anatomy of Hazards. New York: Skandia America Group, 1982, xv,392p Strasser, F.O.; Stafford, P.J.; Bommer, J.J. and Erdik, M. (2008). State-of-the-Art of European Earthquake Loss Estimation Software. Proceedings of the 14 th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Tabucchi, T.; Davidson, R. and Brink, S. (2009). Simulation of post-earthquake water supply system restoration. Civil Engineering and Environmental Systems. Available from: http://dx.doi.org/10.1080/10286600902862615. 191  Tarakji, G. (1992). Lessons Not Learned from 1989 Loma Prieta Earthquake. Journal of Professional Issues in Engineering Education and Practice, Vol. 118, No. 2, pp. 132138. Thibert, K. M. (2008). A Methodology for Assessing the Seismic Risk of Buildings. M Sc. Thesis, Department of Civil Engineering, University of British Columbia, Vancouver, BC, Canada. Todd, D.; Carino, N.; Chung, R. M.; Lew, H. S.; Taylor, A. W.; Walton, W. D.; Cooper, J. D.; and Nimis, R. (1994). 1994 Northridge Earthquake. Performance of Structures, Lifelines, and Fire Protection Systems. NIST Special Publication 862. Buildings and Fire Research Laboratory. National Institute of Standards and Technology. Townsend, F. F. (2006). The Federal Response to Hurricane Katrina: Lessons Learned. Washington, DC: Office of the Assistant to the President for Homeland Security and Counterterrorism. www.whitehouse.gov/reports/katrina-lessons-learned Tsuruta, M.; Goto, Y.; Shoji, Y. and Kataoka, S. (2008). Damage Propagation Caused by Interdependency Among Critical Infrastructures. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. UBC- Joint Interdependent Infrastructures Research Program, UBC-JIIRP final report (2009). Institute for Computing, Information and Cognitive Systems. UBC-JIIRP project, proposal (2003) Uemura, H. (1980). A Guideline to Evaluate Seismic Performance of Existing Medium and lowrise reinforced Concrete Buildings and its application. Proceedings of the 7 th World Conference On Earthquake Engineering, Istanbul, Turkey. van den Berg, B.; Grievink, L.; Yzermans, J. and Lebret, E (2005). Medically Unexplained Physical Symptoms in the Aftermath of Disasters: A systematic review. Epidemiol Rev. 2005;27:92e106. 192  Wang, Y. and Au, S. (2008). Critical Links of Water Supply to Crucial Water Consumers under an Earthquake. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October, 2008. Wijewickreme, D.; Honegger, D.; Mitchell, A. and Fitzell, Trevor (2005). Seismic Vulnerability Assessment and Retrofit of a Major Natural Gas Pipeline System: A Case History. Earthquake Spectra, Volume 21, No. 2, pages 539–567, May 2005; © 2005, Earthquake Engineering Research Institute. www.fema.gov. www.i2sim.ca www.wunderground.com Zakour, M. J. and Harrell, E.B. (2003). Access to Disaster Services: Social Work Interventions for Vulnerable Populations. Journal of Social Service Research, 30 (2): 27-54.  193  

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.24.1-0062902/manifest

Comment

Related Items