public infrastructure engineering vulnerability committee · this publication is the property of...

PUBLIC INFRASTRUCTURE ENGINEERING VULNERABILITY COMMITTEE

ENGINEERING PROTOCOL

BY ENGINEERS CANADA

PART I

VERSION 10 – OCTOBER 2011

Copyright© Canadian Council of Professional Engineers, 2011.

This publication is the property of Engineers Canada and is solely distributed by Engineers Canada. Reproduction or redistribution of this publication, in whole or in part, requires prior written permission from Engineers Canada/Public Infrastructure Engineering Vulnerability Committee Secretariat.

© Canadian Council of Professional Engineers, 2009. Engineers Canada is the business name of the Canadian Council of Professional Engineers. *The term Engineer is an official mark held by the Canadian Council of Professional Engineers. All Rights Reserved. Unauthorized duplication, in whole or in part, is strictly prohibited.

of 166

© Canadian Council of Professional Engineers 2011

Revision 10 - BETA

October 2011

PIEVC Engineering Protocol

For

Infrastructure Vulnerability Assessment and Adaptation to a Changing Climate

PIEVC Engineering Protocol For


Revision 10 -‐ BETA Oct 31, 2011

Page 2 of 166

© Canadian Council of Professional Engineers

2011

For further information about this Engineering Protocol or the National Engineering

Vulnerability Assessment Project please contact the Engineers Canada.

David Lapp, P.Eng. Manager, Professional Practice

Engineers Canada

1100-180 Elgin Street Ottawa, Ontario, Canada

K2P 2K3

[email protected]

(613) 232-2474 Ext. 240




Page 3 of 166


2011

Table of Contents

1 INTRODUCTION AND SCOPE 10

1.1 BACKGROUND 10

1.2 GUIDING PRINCIPLES 11

2 VULNERABILITY, ENGINEERING VULNERABILITY AND RISK ASSESSMENT 13

2.1 VULNERABILITY 13

2.2 ENGINEERING VULNERABILITY 13

2.3 RISK ASSESSMENT 15

2.4 ENGINEERING VULNERABILITY -‐ A SUBSET 16

3 VULNERABILITY ASSESSMENT PLANNING AND EXECUTION 17

3.1 PHASE I -‐ CONCEPT 19

3.2 PHASE II -‐ SCOPING 20 3.2.1 SIGNED LICENSE AGREEMENT 20 3.2.2 DECISION REGARDING PROJECT METHODOLOGY 20 3.2.3 IDENTIFY PROJECT MANAGER 21 3.2.4 PROJECT WORK STATEMENT 22 3.2.5 ENGAGING A CONSULTANT 22 3.2.6 MANAGEMENT APPROVAL 22

3.3 PHASE III -‐ TEAM BUILDING 23

3.4 PHASE IV -‐ EXECUTION 25 3.4.1 CONSULTANT/FACILITATOR RESPONSIBILITIES 25 3.4.2 OWNER RESPONSIBILITIES 25 3.4.3 JOINT RESPONSIBILITIES 26

3.5 PHASE V -‐ REPORTING 26 3.5.1 RECOMMENDATIONS AND CONCLUSIONS 26 3.5.2 STATEMENT OF VULNERABILITY / RESILIENCY 27

4 PROTOCOL OVERVIEW 29

4.1 STEP 1 -‐ PROJECT DEFINITION 31

4.2 STEP 2 -‐ DATA GATHERING AND SUFFICIENCY 31

4.3 STEP 3 -‐ RISK ASSESSMENT 32




Page 4 of 166


2011

4.4 STEP 4 -‐ ENGINEERING ANALYSIS 33

4.5 STEP 5 -‐ RECOMMENDATIONS AND CONCLUSIONS 35

5 THE TEAM 36

5.1 PRACTITIONER 36

5.2 PROFESSIONAL JUDGEMENT 36

5.3 A MULTI-‐DISCIPLINARY TEAM 37

5.4 THE TEAM LEADER 39

6 FUNDAMENTALS OF RISK AND RISK ASSESSMENT 40

6.1 HAZARD IDENTIFICATION – WHAT CAN HAPPEN? 41

6.2 PROBABILITY – HOW LIKELY IS IT TO HAPPEN? 42

6.3 SEVERITY – GIVEN THAT IT HAS HAPPENED, WHAT ARE THE CONSEQUENCES? 43

6.4 RISK – WHAT IS THE SIGNIFICANCE OF THE EVENT? 44

6.5 COMMON MYTHS AND MISCONCEPTIONS ABOUT RISK 44

6.6 SPECIAL CASES 46 6.6.1 VERY LOW PROBABILITY – VERY HIGH SEVERITY 46 6.6.2 VERY HIGH PROBABILITY – VERY LOW SEVERITY 46

6.7 THE RISK MATRIX 47

7 THE VULNERABILITY ASSESSMENT WORKSHOP 49

8 CLIMATE RESOURCES FOR USE IN PIEVC PROTOCOL EVALUATIONS 53

SECTION 1: RESOURCES AND TOOLS 53

8.1 OBSERVED DATA 53 8.1.1 CLIMATE NORMALS 55 8.1.2 CLIMATE INDICES 56 8.1.3 INTENSITY, DURATION AND FREQUENCY (IDF) CURVES 56 8.1.4 MAPPED AND GRIDDED DATA 57

8.2 MODEL PROJECTIONS 58 8.2.1 MODEL VARIETIES 59

8.3 MODEL OUTPUT 61 8.3.1 FINER SCALE 64




Page 5 of 166


2011

SECTION2 -‐ APPLICATION OF RESOURCES 66

8.4 WHO TO ASK FOR CLIMATE DATA 66

8.5 WHAT TO ASK FOR 66

8.6 WHAT TO EXPECT 66

8.7 CONCLUSION 70

8.8 REFERENCES 71

9 ESTABLISHING CHANGING CLIMATE PROBABILITY SCORES 72

9.1 BASIS 72

9.2 THE IMPORTANCE OF DIFFERING INTERPRETATION OF KEY WORDS ACROSS DISCIPLINES 72

9.3 DEFINITION 73 9.3.1 CLARIFICATION 73

9.4 PROFESSIONAL JUDGEMENT AND THE PRACTITIONER TEAM 73

9.5 DEFINING THRESHOLDS 74

9.6 FRAME OF REFERENCE 76

9.7 METHODOLOGIES FOR ASSIGNING PROBABILITY SCORES 76 9.7.1 TWO DIFFERENT APPROACHES 76 9.7.2 CONSIDERATIONS AFFECTING PROBABILITY SCORES 77 9.7.3 ASSIGNING THE PROBABILITY SCORE 79

9.8 PROBABILITY SCORING WORKSHEET 80

9.9 SCORING FOR PRESENT AND FUTURE CLIMATE SCENARIOS 82

10 ESTABLISHING SEVERITY SCORES 85

10.1 BASIS 85

10.2 DEFINITION 85

10.3 THE PRACTITIONER’S PROFESSIONAL JUDGEMENT AND THE PRACTITIONER TEAM 85

10.4 THRESHOLDS 86 10.4.1 CONSIDERATIONS AFFECTING SEVERITY SCORES 86

10.5 IMPLICATIONS OF DIFFERING RISK ASSESSMENT APPROACHES 89

11 ECONOMIC CONSIDERATIONS 90

12 PROTOCOL FOR CHANGING CLIMATE INFRASTRUCTURE VULNERABILITY ASSESSMENT 93




Page 6 of 166


2011

12.1 STEP 1 – PROJECT DEFINITION 95 12.1.1 PREPARE STEP 1 WORKSHEET 96 12.1.2 IDENTIFY THE INFRASTRUCTURE 96 12.1.3 IDENTIFY CLIMATE PARAMETERS 96 12.1.4 IDENTIFY THE TIME HORIZON 97 12.1.5 IDENTIFY THE GEOGRAPHY 97 12.1.6 IDENTIFY JURISDICTIONAL CONSIDERATIONS 97 12.1.7 SITE VISIT 97 12.1.8 ASSESS DATA SUFFICIENCY 97

12.2 STEP 2 – DATA GATHERING AND SUFFICIENCY 99 12.2.1 PREPARE STEP 2 WORKSHEET 100 12.2.2 STATE INFRASTRUCTURE COMPONENTS 100 12.2.3 STATE THE TIME HORIZON FOR THE ASSESSMENT 101 12.2.4 STATE THE GEOGRAPHY 101 12.2.5 STATE SPECIFIC JURISDICTIONAL CONSIDERATIONS 102 12.2.6 STATE OTHER POTENTIAL CHANGES THAT MAY AFFECT THE INFRASTRUCTURE 102 12.2.7 IDENTIFY RELEVANT CLIMATE PARAMETERS 102 12.2.8 IDENTIFY INFRASTRUCTURE THRESHOLD VALUES 103 12.2.9 IDENTIFY POTENTIAL CUMULATIVE OR SYNERGISTIC EFFECTS 104 12.2.10 STATE CLIMATE BASELINE 104 12.2.11 STATE THE CHANGING CLIMATE ASSUMPTIONS 105 12.2.12 ESTABLISH CHANGING CLIMATE PROBABILITY SCORES 106 12.2.13 ASSESS DATA SUFFICIENCY 111

12.3 STEP 3 – RISK ASSESSMENT 112 12.3.1 PREPARE STEP 3 WORKSHEET 113 12.3.2 ESTABLISH THE INFRASTRUCTURE OWNER’S RISK TOLERANCE THRESHOLDS 113 12.3.3 PREPARE FOR RISK ASSESSMENT WORKSHOP 115 12.3.4 CONDUCT A RISK ASSESSMENT WORKSHOP (FACILITATED OPTION) 116 12.3.5 CONFIRM CLIMATE PARAMETERS 116 12.3.6 CONFIRM INFRASTRUCTURE THRESHOLD VALUES 117 12.3.7 CONFIRM PROBABILITY SCORES 117 12.3.8 CONFIRM POTENTIAL CUMULATIVE OR SYNERGISTIC EVENTS 117 12.3.9 IDENTIFY RELEVANT INFRASTRUCTURE RESPONSES 118 12.3.10 COMPLETE YES/NO ANALYSIS 119 12.3.11 ESTABLISH INTERACTION SEVERITY 119 12.3.12 CALCULATE RISK SCORES 122 12.3.13 CONDUCT A RISK ASSESSMENT WORKSHOP (CONSULTANT OPTION) 122 12.3.14 ASSESS DATA SUFFICIENCY 123 12.3.15 CONFIRM THE INFRASTRUCTURE OWNER’S RISK TOLERANCE THRESHOLDS 123 12.3.16 DOCUMENT RISK PROFILE 123




Page 7 of 166


2011

12.3.17 REVIEW SPECIAL CASES 124 12.3.18 IDENTIFY NEXT STEPS 125

12.4 STEP 4 – ENGINEERING ANALYSIS 127 12.4.1 PREPARE STEP 4 WORKSHEET 128 12.4.2 CALCULATE THE EXISTING LOAD (LE) 129 12.4.3 CALCULATE CHANGING CLIMATE LOAD (LC) 129 12.4.4 CALCULATE OTHER CHANGE LOADS (LO) 130 12.4.5 CALCULATE TOTAL LOAD (LT) 130 12.4.6 CALCULATE THE EXISTING CAPACITY (CE) 130 12.4.7 CALCULATE THE PROJECTED CHANGE IN EXISTING CAPACITY (CΔE) 131 12.4.8 CALCULATE ADDITIONAL CAPACITY (CA) 131 12.4.9 CALCULATE THE PROJECTED TOTAL CAPACITY (CT) 132 12.4.10 CALCULATE VULNERABILITY RATIO 133 12.4.11 CALCULATE CAPACITY DEFICIT 133 12.4.12 ASSESS DATA SUFFICIENCY 134 12.4.13 EVALUATE NEED FOR ADDITIONAL WORK 134 12.4.14 IDENTIFY CONCLUSIONS AND RECOMMENDATIONS 135

12.5 STEP 5 – RECOMMENDATIONS AND CONCLUSIONS 136 12.5.1 PREPARE STEP 5 WORKSHEET 137 12.5.2 DECLARE ASSUMPTIONS REGARDING AVAILABLE INFORMATION, DATA SOURCES, UNCERTAINTIES AND RELEVANT LIMITATIONS 137 12.5.3 STATE CONCLUSIONS 137 12.5.4 STATE RECOMMENDATIONS 138 12.5.5 PREPARE STATEMENT OF VULNERABILITY / RESILIENCY 138

13 REFERENCES 140

APPENDIX A: SUGGESTED CLIMATE AND INFRASTRUCTURE THRESHOLD PARAMETERS 141

APPENDIX B: INFRASTRUCTURE RESPONSE CONSIDERATIONS 145

APPENDIX C: GLOSSARY 151

List of Figures

Figure 1: Vulnerability, Engineering Vulnerability and Risk ......................................................... 16 Figure 2: Vulnerability Assessment Roadmap ............................................................................. 18 Figure 3: Relevant Interactions between Climate and Infrastructure .......................................... 29 Figure 4: Overview of the Protocol .............................................................................................. 30




Page 8 of 166


2011

Figure 5: Team Member Skills and Interactions .......................................................................... 39 Figure 6: Example Risk Matrix ..................................................................................................... 48 Figure 7: Global Climate Temperature Map ................................................................................ 54 Figure 8: Climate Normals for St. John’s NL, 1971-‐2000 ............................................................. 56 Figure 9: Example IDF Curve ......................................................................................................... 57 Figure 10: Contour Map of Extreme Rainfall 24 Hour Duration over Atlantic Canada ................ 58 Figure 11: Conceptual Structure of a Coupled Atmosphere-‐Ocean GCM ................................... 60 Figure 12: Examples of GCMS ...................................................................................................... 61 Figure 13: A Simple Typology of Model Uncertainties ................................................................ 63 Figure 14: Example of Output from Statistical Downscaling ....................................................... 65 Figure 15: Example Gaussian Distribution ................................................................................... 70 Figure 16: Examples of Infrastructure Threshold Values ............................................................. 76 Figure 17: Example Probability Scoring Exercise ......................................................................... 82 Figure 18: Example Baseline Weather Event Probability Scoring Exercise .................................. 84 Figure 19: Step 1 – Project Definition Process Flowchart ............................................................ 95 Figure 20: Step 2 – Data Gathering and Sufficiency Process Flowchart ...................................... 99 Figure 21: Probability Score Definitions .................................................................................... 110 Figure 22: Step 3 – Risk Assessment Process Flowchart ............................................................ 112 Figure 23: Reference Risk Tolerance Thresholds ....................................................................... 114 Figure 24: Severity Score Definitions ......................................................................................... 121 Figure 25: Step 4 – Engineering Analysis Process Flowchart ..................................................... 128 Figure 26: Step 5 – Recommendations Process Flowchart ........................................................ 136




Page 9 of 166


2011

Part I

Background, Overview and Guidance




Page 10 of 166


2011

1 Introduction and Scope

1.1 Background The PIEVC Engineering Protocol for Infrastructure Vulnerability Assessment and Adaptation to a Changing Climate (Protocol) outlines a process to assess the infrastructure component responses to impacts of changing climate. Information developed through the assessment process will assist owners and operators to effectively incorporate adaptation to changing climate into design, development and management of their existing and planned infrastructure. The Protocol describes a step-by-step methodology of risk assessment and optional engineering analysis for evaluating the impact of changing climate on infrastructure. The observations, conclusions and recommendations derived from the application of this Protocol provide a framework to support effective decision-making about infrastructure operation, maintenance, planning and development. The Protocol was developed for owners and operators to assess public infrastructure. However, the principles and steps will be similar for assessing privately owned infrastructure. The Protocol was developed with funding contributions from Natural Resources Canada under the direction of the Public Infrastructure Engineering Vulnerability Committee (PIEVC). PIEVC was a national steering committee established by Engineers Canada in 2005. The committee consisted of senior representatives from federal, provincial and municipal levels of government in Canada along with several non-government organizations. It oversaw the first National Engineering Vulnerability Assessment project, a long-term initiative of the Canadian Engineering profession to assess the vulnerability of public infrastructure to the impacts of changing climatic conditions. This Protocol is one key product of PIEVC’s work. Engineers Canada (the business name of the Canadian Council of Professional Engineers) owns the intellectual property that is the Protocol. It may be used in Canada for Canadian-based infrastructure without charge, provided the user signs a license agreement with Engineers Canada. Parties residing outside of Canada wishing to use the Protocol should contact Engineers Canada to establish licensing arrangements. The Protocol is divided into two main sections: Part I

Description of the processes and organization for planning engineering vulnerability




Page 11 of 166


2011

assessments of public infrastructure Presentation of the basic principles of risk management that are applicable to this work,

along with technical references Procedural description of the five steps involved in executing the Protocol.

Part II

Detailed, step-by-step procedures Worksheets covering each step of the process

Part I of the Protocol is designed to provide a comprehensive background on the basic elements of a engineering vulnerability assessment and provides additional guidance on the philosophies, assumptions and critical elements that support a successful assessment project. Part II of the Protocol is a procedures document. Generally, it does not explain the rationale for the outlined tasks, as this can be found in Part I. Rather, Part II provides the practitioner with a sequential listing of the actions they must undertake to complete an assessment. The companion worksheets provide a location for the practitioner to document that they have addressed each step of the Protocol and record their findings.

1.2 Guiding Principles This Protocol is based on a set of guiding principles that establish a consistent methodology to assess infrastructure engineering vulnerability induced by changing climatic conditions. These principles include:

This is an engineering exercise. The purpose is to assess engineering vulnerability on a system-component basis and identify adaptation strategies based on this this analysis.

The assessment of climate induced engineering vulnerability is a multi-disciplinary

process requiring effective interdisciplinary collaboration. The process relies on professional judgment and as such it is critical that the assessment incorporate the input and advice of all relevant disciplines.

In the context of the Protocol, the word practitioner refers to the entire team of

professional, scientific, management, operations and maintenance personnel engaged in the assessment.

The process is results-oriented. Emphasis is placed on identifying pragmatic approaches




Page 12 of 166


2011

to answering the infrastructure vulnerability question within an established schedule and budgetary constraints. These constraints may affect the robustness of the overall assessment. The Protocol is designed to allow practitioners to assess vulnerability, document assumptions and identify the limitations of the assessment.

The Protocol specifically calls for documentation in a manner that allows for future

review of the infrastructure, allowing future teams to update the assessment with newer information.

Professional judgment and expert opinion are different. Expert opinion is a legal term

that defines a set of circumstances where a court or other judicial body may officially and legally rely upon an individual’s unique and specialized opinions within a limited area of expertise. Professional judgement is the interpretation and synthesis of data, facts and observations and the extrapolation of that analysis by the practitioner to provide a judgment of how the infrastructure may respond to a specific set of conditions.

Language may present hurdles in the effective execution of the Protocol. Wherever

possible we have defined the way we apply terms within the context of the Protocol. However, it is critical that the various professionals involved in the process communicate openly with each other. Different professional disciplines often use the same words to convey very different meanings. On multi-disciplinary teams this can often lead to confusion and conflict. Understanding these differentiations is critical to both the progress and efficiency of an assessment team. An open and frank assessment of underlying principles will often clarify matters and resolve apparent differences of opinion.

It is very important that practitioners do not underestimate the impact that differences in language usage between different disciplines may have on the effective execution of an interdisciplinary assessment. This is particularly evident in the language and culture of climate scientists as opposed to that of the engineering profession. To a scientist, being conservative means essentially not drawing conclusions beyond what can confidently be inferred from scientific analysis. For example, in sea level rise, this would generally mean “low estimates” to a scientist. However, to an engineer, largely interested in protection, being “conservative” would normally mean the exact opposite, i.e. high estimates. These differences can often lead to confusion and avoidable conflict unless the practitioner is particularly sensitive to the nuances of language and professional culture as they manage their team and work with other professionals from the infrastructure owner’s organization.




Page 13 of 166


2011

2 Vulnerability, Engineering Vulnerability and Risk Assessment This Protocol outlines a procedure for assessing the engineering vulnerability of infrastructure and identifying adaptive actions to reduce the identified vulnerability, as appropriate. A glossary of terms used in the Protocol is presented in Appendix C. However, it is important that practitioners have a clear understanding of the distinctions between vulnerability, engineering vulnerability and risk. We provide a brief explanation in the following sections.

2.1 Vulnerability In this protocol we define vulnerability as:

The degree to which a system is susceptible to, or unable to cope with, adverse effects of climate change, including climate variability and extremes. Vulnerability is a function of the character, magnitude, and rate of climate variation to which a system is exposed, its sensitivity, and its adaptive capacity.

Vulnerability captures all of the potential impacts posed by the climate affecting an infrastructure system. This includes not only impacts on the serviceability and functionality of the system but also broader socioeconomic and environmental effects. This definition is far-reaching and creates a large set of factors that may not be within the purview of the infrastructure owner or its engineering and management personnel. An all encompassing assessment of vulnerability, per se, would require an extensive range of expertise including not only engineering and climate specialties but also social scientists, economists, multiple levels of government and a very large group of external stakeholders. This type of study is well outside of the scope of most infrastructure owners and operators. Infrastructure owners require a more focused approach to determine the sub-set of hazards that are within their purview. This allows them to develop an effective adaptation strategy for factors that are within their direct management and budgetary control. This is the realm of engineering vulnerability.

2.2 Engineering Vulnerability Engineering vulnerability is a sub-set of vulnerability that is more focused on the structural and operational features of the infrastructure itself. In this Protocol we define engineering vulnerability as:




Page 14 of 166


2011

The shortfall in the ability of public infrastructure to absorb the negative effects, and benefit from the positive effects, of changes in the climate conditions used to design and operate infrastructure. Vulnerability is a function of:

1. Character, magnitude and rate of change in the climatic conditions to which infrastructure is predicted to be exposed;

2. Sensitivities of infrastructure to the changes, in terms of positive or negative consequences of changes in applicable climatic conditions; and

3. Built-in capacity of infrastructure to absorb any net negative consequences from the predicted changes in climatic conditions.

Vulnerability assessment will, therefore, require assessment of all three elements above.

Engineering vulnerability is a specific category of vulnerability that generally excludes factors outside of the direct management control of the infrastructure owner, operations, maintenance and engineering personnel. While the assessment of engineering vulnerability may consider broader socio-economic impacts, the assessment remains focused on identifying adaptation strategies directly applicable to the infrastructure system itself. The assessment may identify broader issues but typically leaves the development of adaptation responses to those issues to other experts in follow-up studies. For example, a highway vulnerability assessment may identify that fogging conditions will occur more frequently potentially affecting public safety. While outside of the scope of the current assessment, this may nonetheless be a serious concern that the practitioner identifies for further study requiring specialized expertise. There is an element of prediction imbedded in the definition of engineering vulnerability. The practitioner is asked to forecast both:

The change in climate conditions to which the infrastructure will be exposed; and The way the infrastructure may respond to those changes.

In engineering practice, analysis of the likelihood of hazard events and the impacts of those events on engineered systems is defined as risk assessment.




Page 15 of 166


2011

2.3 Risk Assessment Risk assessment is a process used to establish a measure of the risk created by identifiable hazards. In this Protocol we define risk as:

The possibility of injury, damage, loss, loss of function, or negative environmental impact created by a hazard. The significance of risk is a function of the probability of an unwanted incident and the severity of its consequence.

We provide a detailed backgrounder on risk in Section 6. It is important to understand that, in this Protocol, risk is applied as a measure of engineering vulnerability. Engineering vulnerability is a physical phenomenon. That is, either a structure is or is not vulnerable to changing climate. If it is, in an engineering sense, the infrastructure is subject to failure or impaired service function. The degree of failure depends upon the level of the vulnerability. Risk is a derived value that characterizes this vulnerability. Risk can be based on statistical analysis of failure frequency and failure mode analysis. When assessed in this manner, the process can be a highly numerical and precise analysis. However, just as often, due to technical, budgetary or other limitations, we do not have the capacity to measure engineered systems to a high level of precision. Nonetheless, risk assessment may still be conducted based on professional judgement and review of historic failure modes. In these cases, the resulting risk values cannot be viewed as precise, calculated, parameters. Rather the assessment is a scoring exercise that identifies areas of concern, facilitating the prioritization of management and engineering responses. For most changing climate infrastructure vulnerability assessments, the practitioner team will not have access to precisely measured or derived data. There is still a significant level of uncertainty associated with predicting how climate parameters will change within precisely defined geographic and temporal boundary conditions. As well, although the infrastructure in question may have experienced similar conditions in the past, it is often unclear how it may respond to the specific set of conditions identified by practitioner teams. For these reasons, most infrastructure vulnerability assessments are fundamentally a blend of some measure of site-specific data in conjunction with a scoring exercise based on professional judgment. This Protocol provides detailed guidance on the range and level of professional expertise required to successfully execute an evaluation.




Page 16 of 166


2011

2.4 Engineering Vulnerability -‐ A Subset Based on the above considerations, engineering vulnerability is a subset of overall vulnerability. We can gauge the range of engineering vulnerability through standard engineering tools, such as risk assessment. The Protocol outlines a risk scoring methodology based on professional judgment and applies standard risk assessment methodologies to characterize the engineering vulnerability of an infrastructure system. Many assumptions are used to back up the professional judgment. The Protocol assists in documenting these assumptions. The practitioner may revisit the assumptions in the future to update the overall risk assessment based on new or better information. Conceptually, the relationships between vulnerability, engineering vulnerability and risk assessment are depicted in Figure 1.

Figure 1: Vulnerability, Engineering Vulnerability and Risk




Page 17 of 166


2011

3 Vulnerability Assessment Planning and Execution Engineering vulnerability assessments normally involve a discrete infrastructure, subsystems or components rather than an entire inventory. The infrastructure(s) should be carefully selected to provide a representative sample of the inventory. If significant vulnerabilities are detected, and there is widespread variability in nature and severity of vulnerabilities, it may be necessary to assess all infrastructures in an inventory to determine what adaptive actions are required for an individual component of the infrastructure. There are five phases in planning and executing a changing climate infrastructure vulnerability assessment. These include:

Phase I – Concept Phase II – Scoping Phase III – Team Building Phase IV – Execution Phase V – Reporting

These phases are briefly described in the following sections and are presented graphically in Figure 2.



Revision 10.3 Oct 31, 2011

Page 18 of 166


2011

Figure 2: Vulnerability Assessment Roadmap



Revision 10.3 Oct 31, 2011

Page 19 of 166


2011

3.1 Phase I -‐ Concept All vulnerability assessments initiate with a single question:

Is this infrastructure at risk because of changing climate conditions? The question may come from a variety of sources. These include, but are not limited to:

Infrastructure management; Consultants approaching the infrastructure owner; Concerned public; Regulators; or Key decision-makers, including politicians.

Whatever the source, the fundamental objective of a changing climate vulnerability assessment is to directly answer that question in a formal, substantiated, manner. The answer must be based on solid scientific analysis, engineering practice and informed professional judgment. The Protocol provides a roadmap for answering the question. Other benefits accrue. For example, executing the Protocol will often identify actions that can be implemented readily with minimal cost or business disruption. In this way, the Protocol provides added value as a supplement to traditional criticality assessment efforts. At the concept stage, the question has been raised and, once raised, some form of action is deemed necessary. The project proponent resolves three key questions at this point in the process:

1. Which particular infrastructure are we concerned about?

Where is it? Who makes decisions about it? What does it do? How critical is the infrastructure to its user base?

2. What time horizon are we concerned about?

Only until the next refurbishment? Over the design lifetime of the infrastructure system?

3. Does senior management buy in?

Assessments typically identify areas where action is necessary. It is important that




Page 20 of 166


2011

managers are fully informed of the implications of agreeing to this process. There will likely be additional work arising from the findings of the assessment and it may be necessary to secure new funding to support the actions identified by an assessment.

With clear answers to these questions, the project proponent will be positioned to effectively plan for and establish the overall scope of the assessment. These answers not only define the project at its highest level, but also provide a preliminary indication of the project boundary conditions. For example, managers may give notional approval only if the project remains within specified budgetary limits. This would dictate the types of climate information and other analysis available to the project team and establish guidelines for addressing the questions raised during the Scoping phase of the assessment.

3.2 Phase II -‐ Scoping

3.2.1 Signed License Agreement During the initial stages of project scoping, the project proponent will sign a license agreement with Engineers Canada. This will provide access to the Protocol and the intellectual resources of the Engineers Canada network of infrastructure and vulnerability experts. The Protocol is the intellectual property of Engineers Canada, and owners/operators of infrastructure, as well as third-party users, (e.g. consultants) may not use it without the permission of Engineers Canada. Permission is normally granted through a license agreement. Part of this agreement would normally include the obligation to share the results of the assessment with the Federal Government of Canada and Engineers Canada. Ensuring that there is a signed license agreement allows Engineers Canada to maintain consistency between case studies, acquire additional information to inform national level infrastructure vulnerability assessments and obtaining feedback to improve successive versions of the Protocol.

3.2.2 Decision Regarding Project Methodology At this stage of the project, the proponent will assess the project delivery methodologies that are available to execute the vulnerability assessment. These decisions are dictated by availability of internal resources and overall budget limitations. Notionally, there are two fundamental approaches to executing the Protocol.




Page 21 of 166


2011

1. Consultant:

The infrastructure owner retains an external expert in vulnerability assessment to execute the project. This approach uses a minimum of internal resources, requiring staff involvement only in project management, data acquisition, workshops and review. The consultant team conducts the bulk of the work off-site. Although this approach uses a minimum of internal resources it is more expensive, requiring more consultant professional hours.

2. Facilitator:

The infrastructure owner executes the Protocol using internal resources. The services of a facilitator, conversant with the Protocol, are used to guide internal staff through the process. Internal staff conducts all key activities associated with the assessment. This approach is the least expensive in terms of overall project budget. However, it places a significant demand on internal resources throughout the project. Project timelines may extend somewhat as staff engagement is often secondary to their day-to-day activities.

PIEVC Protocol Vulnerability assessments have been executed successfully using both approaches. The key factors in making a decision regarding project methodology are budget, availability of internal resources that can be assigned to this task and project scheduling. It is important to be clear about the methodology as consultant/facilitator proposal scope and cost will be driven by this decision.

3.2.3 Identify Project Manager Based on the chosen Protocol execution strategy, the infrastructure owner will select an internal project manager. This decision would normally fall out of the deliberations regarding methodology, as the skill sets necessary to manage the process are slightly different depending upon the decision. Using a consultant to execute the process relieves the project manager of some of the technical demands of the project. The consultant should provide, or have access to, this expertise. However, if the facilitated approach is chosen, the project manager will have a greater responsibility to manage the internal team, direct and maintain data and information flow between the internal team and external consultants, and assess the technical arguments raised by




Page 22 of 166


2011

staff during the process. The facilitator may not have the expertise to draw conclusions about specific engineering issues.

3.2.4 Project Work Statement

Based on the Protocol execution strategy, the project manager will now have sufficient information to complete the initial stages of project definition in sufficient detail to prepare a project work statement suitable for procurement purposes. This would include a better understanding of the internal resource needs, overall project budget requirements and schedule. Obviously, at the project scoping stage, proponents will not have access to all of the data necessary to execute the Protocol. However, the methodology and underlying thought process will significantly help the proponent identify the key components that must be included in the project Work Statement to provide sufficient information to potential consultants to scope and cost the assessment. Engineers Canada has developed generic work statements and has assisted a number of infrastructure owners establish work statements for PIEVC assessments. Engineers Canada will share the most recent incarnations of these documents with serious case study proponents, upon request.

3.2.5 Engaging a Consultant Armed with the work statement, and a thorough understanding of the project execution strategy, the project manager can now engage a consultant or project facilitator. The owner may prepare and issue a request for proposal (RFP), as appropriate. The form of the RFP is dictated by the internal policies and procedures of the infrastructure owner. The owner may also choose to negotiate and directly assign the work.

3.2.6 Management Approval Based on a more thorough understanding of the project execution strategy, internal resource demands, scheduling and budget the senior manager can now make an informed preliminary approval decision regarding the vulnerability assessment. Throughout the scoping of the project, the proponent will uncover potential project sensitivities.




Page 23 of 166


2011

These may simply be project management concerns regarding budgeting and scheduling or they may have broader implications. As the infrastructure owner becomes more fully conversant with the level of assessment being contemplated, they may wish to impose further constraints on the overall range and scope of the work based on criteria such as available resources and budget, overlap with the scope of other projects, etc. This is a dialogue that should occur with the senior manager before the project gets underway, as it is often difficult to pull the assessment team back from sensitive issues after the fact. The scope should not only clearly define what IS to be considered by the assessment it should also clearly define what IS NOT to be considered. In this sense, the scope represents the boundary limits of the assessment. These matters should be fully discussed with the senior manager and guidelines established as part of the decision process. Good, clear, definition of boundary conditions is necessary for a successful project.

3.3 Phase III -‐ Team Building Retaining the consultant/facilitator is a key step in building the project team. Based on the skill sets presented by the consultant/facilitator, the project manager will need to secure other resources. This may be from within the organization, or may include retired staff and/or outside experts working on a volunteer basis, or other consultants. The skill sets necessary for successful execution of the Protocol are outlined in Section 6. Once the consultant and internal team have been assigned, the project manager can establish firm project schedules and budgets. Project timelines depend upon the scheduling constraints of both the consultant and other resources. Successful execution of the Protocol depends upon availability and commitment of each member of the team and the project manager will need to be flexible to ensure maximum participation. Budgeting should include not only consultant/facilitator fees but also allocations for data acquisition, data analysis/processing efforts, other retained services, workshop costs, etc. Once the project manager has a clear definition of where and when the workshop will occur and of the allocation of additional resources, they can finalize the project budget. It is also important at this stage of the project to clearly identify and secure internal data resources. Uncoordinated chasing of data can jeopardize budgetary control. Abundant and complete data is often difficult to secure for a vulnerability assessment, and where it cannot be provided, the consultant or team, may find it necessary to make assumptions or develop the data




Page 24 of 166


2011

using additional resources. It is important that the project manager has a clear understanding of what can be provided from within the organization to inform these considerations. It is very important at this stage of the project to identify the individuals or organizations that will be providing climate analysis support to the team. As a minimum, this individual or organization should possess the following key attributes:

Thorough knowledge of meteorology and/or climatology and related sciences. Knowledge of Earth's climate system and its interaction with the natural and built environment. Knowledge of climate change science and models, potential impacts of changing climate and possible measures to adapt to changing climate.

o This may be less critical as newer datasets are made available for regions.

Experience in working with large sets of meteorological and climatological data. Thorough knowledge of the characteristics and applications of data collection methodology for a variety of data streams including station data, remote sensing imagery and data. Knowledge of computer input and data manipulation and analysis techniques and software.

Experience with writing technical or policy documents, such as Standards of

Operation manuals or technical reports. PIEVC assessments are inherently multidisciplinary efforts with differences in approach, terminology and perceived sensitivities when each discipline reviews similar information. Each brings a necessary perspective It is important that parties work together to understand the different perspectives each discipline brings to the project. Because of these differences in approaches used by engineers and climate specialists, it is important that the two parties learn to work together and understand the constraints and limits of each expert’s respective field. The climate specialist should:

Be involved early on in the process to begin discussing climate data needs and clearly explain what is possible to provide in terms of information;

Consider ensemble results generated from multiple climate models (and model runs); Consider results generated using different methods to capture the whole range of

plausible scenarios of change (not just preferred methods); and Validate when possible the information with other climate specialists.

Equipped with all of this information, the project manager can now establish a project execution plan. This plan will not only outline the consultant/facilitator schedule and budget but should also consider the availability and scheduling constraints of internal resources and external




Page 25 of 166


2011

advisors, as appropriate. This plan should then be shared with the team, as appropriate, to ensure commitment to the project. Often, during the RFP stage of a project the climate specialist is not defined or selected. This can lead to the practitioner being asked to develop an execution plan prior to completing the team building activities outlined above. This can lead to budgetary and scope management issues later in the project. Project proponents are strongly urged to require definition of the climate data as one element of the RFP process.

3.4 Phase IV -‐ Execution The Protocol will guide the vulnerability assessment.

3.4.1 Consultant/Facilitator Responsibilities During execution, the consultant/facilitator should provide ongoing project status reports at regularly scheduled intervals. This may entail periodic meetings or teleconferences with the project manager. The consultant/facilitator should clearly identify issues that may have budgetary or scheduling implications as early as possible to allow the project manager the opportunity to address matters before they grow into significant problems. At project completion the consultant/facilitator should provide a project report. The scope of the report will depend upon the requirements of the infrastructure owner. It may be an overview of findings or a very detailed engineering report. At the end of the project the consultant/facilitator would normally report

Conclusions regarding the nature and severity of infrastructure component vulnerability;

Conclusions regarding the overall resiliency of the infrastructure; Recommendations for action, including areas of concern where further study is

necessary.; and A Statement of Vulnerability/Resiliency.

3.4.2 Owner Responsibilities During execution, the infrastructure owner will provide input into the development of an infrastructure inventory. The consultant/facilitator may not have a complete understanding of




Page 26 of 166


2011

how the infrastructure is managed or constructed. It is particularly important that the owner identify where the in-field infrastructure may differ from design drawings, specifications and standards. The owner will also provide a clear statement of the Owner’s Risk Tolerances, in terms of both safeguarding the condition of the infrastructure asset or a minimum tolerable level of service. The Protocol will guide this discussion. The consultant/facilitator should not dictate risk thresholds to the owner. The owner should establish their tolerances based on budget, resources, the needs of their stakeholders, and other considerations. This understanding should be clearly defined to the consultant/facilitator to guide their work in determining the overall climate risk profile.

3.4.3 Joint Responsibilities Either the consultant team or the owner may secure climate information. This will depend upon the availability of the data. Some owners have weather data and climate projections that are of interest to them. Others do not. It is important to clearly define who is responsible for this task. Overall project cost management will be significantly more difficult if coordination of climate data is not well executed. Acquiring climate information can often be the critical path activity for the entire project. Upon completion of the project, the project team and consultant/facilitator should clearly state:

The overall changing climate risk profile of the infrastructure; and Information gaps that affect the risk profile.

Information gaps that affect the risk profile often lead to recommendations for additional work, data gathering and studies following completion of the assessment. Consideration should be given to revisiting the recommendations in future when needed data may be secured.

3.5 Phase V -‐ Reporting

3.5.1 Recommendations and Conclusions At the completion of the vulnerability assessment the consultant and/or project team will provide a set of conclusions and recommendations relating to the climate impact and adaptation of the infrastructure. These conclusions and recommendations will fall into several categories, as




Page 27 of 166


2011

outlined in Section 4.5:

1. A report of infrastructure components that have been assessed to be vulnerable.

2. Initial recommendations regarding possible:

i. Remedial engineering actions; ii. Monitoring activities;

iii. Management actions; iv. Additional data collection; or v. Additional analysis of particular infrastructure-climate interactions that

may be necessary to determine extent and nature of vulnerabilities.

3. A summary of the infrastructure components that have been assessed to have sufficient adaptive capacity to withstand projected changing climate impacts; thus requiring no further action at this time.

4. A report on data gaps and availability; requiring additional work or studies.

5. Identification of infrastructure components that may be evaluated in the future.

6. A report on other conclusions, trends, insights and limitations.

3.5.2 Statement of Vulnerability / Resiliency The vulnerability assessment initiated with the question:

Is this infrastructure at risk because of changing climate conditions? It is very important that the consultant and/or team provide a clear concise answer to the question. This would normally be articulated as a summary statement outlining context and limitations. For example, the statement may say:

This infrastructure is generally resilient to changing climate impacts anticipated over the next twenty years with the exception of xxx that is vulnerable to projected changes in xxx over that time horizon. This opinion is based on information available to us at the time of our assessment.




Page 28 of 166


2011

The above statement is provided as an example only. The statement will depend on the actual findings and limitations of the assessment. Also, if the assessment identifies that the infrastructure is generally vulnerable, it should state that opinion and identify the most significant vulnerabilities.




Page 29 of 166


2011

4 Protocol Overview Traditionally, engineers have relied on historic climate data records to design infrastructure. Under climate change, this historic data may no longer be appropriate, as it does not capture trends that climate change can pose. This can translate to a more challenging operating environment for which the infrastructure was never designed. As a result, infrastructure may be vulnerable. Existing infrastructure may not have sufficient resiliency. New infrastructure may not be designed with sufficient load and adaptive capacity. The process set forth by the Protocol is designed to aid practitioners in characterizing any gaps between additional duty loads and its capacity to adapt to that challenge outside its original design. To assess changing climate infrastructure vulnerability, the practitioner must evaluate:

1. The infrastructure; 2. The climate (historic, recent and projected); and 3. Historic and projected responses of the infrastructure to the climate.

This interaction is depicted in Figure 3.

Figure 3: Relevant Interactions between Climate and Infrastructure




Page 30 of 166


2011

A great deal of information may be available to describe the infrastructure and the climate in the region. The Protocol sets out a procedure to sift the data to develop an understanding of how climate and infrastructure interact to create vulnerability. Not all climate and infrastructure data is necessary to complete the Protocol, nor may it be relevant to the infrastructure. Practitioners must always be asking fi the infrastructure will “notice” the change in a given climate parameter. The initial stages of the Protocol help the practitioner identify the key data necessary to complete the assessment. Throughout the Protocol the practitioner is directed to continuously evaluate the availability and quality of data sufficient to support conclusions and recommendations. The Protocol is divided into five steps, as illustrated in Figure 4. Each step of the Protocol is described in greater detail in Sections 4.1 through 4.5.

Figure 4: Overview of the Protocol




Page 31 of 166


2011

4.1 Step 1 -‐ Project Definition In Step 1 the practitioner will be asked to:

• Develop a general description of the infrastructure; o The location; o Historic climate; o Load; o Age; o Life Cycle; o Other relevant factors; and

Identify major documents and information sources. In this step the practitioner defines the boundary conditions for the vulnerability assessment.

4.2 Step 2 -‐ Data Gathering and Sufficiency In Step 2 the practitioner will be asked to provide more definition about:

1. Which parts of the infrastructure will be assessed; and 2. The particular climate factors that will be considered.

Step 2 is comprised of two key activities:

1. Identification of the features of the infrastructure that will be considered in the assessment:

Detailing physical components of the infrastructure;

o Number of physical components; o Location(s);

Other relevant engineering/technical considerations: o Material of construction; o Age; o Importance within the region; o Physical condition;

Existing and archival operations and maintenance practices; Operation and management of the infrastructure;

o Insurance considerations; o Policies;




Page 32 of 166


2011

o Guidelines; o Regulatory setting; and o Legal considerations.

2. Identification of applicable climate information. Sources of climate information include,

but are not limited to:

The National Building Code of Canada, Appendix C, Climate Information; Intensity - Duration – Frequency (IDF) curves; Flood plain mapping; Regionally specific climatic modeling and scenario development (IPCC, CCCSN.ca); Heat units (i.e. degree-days) (i.e. for agriculture, HVAC, energy use, etc.); Institute for Catastrophic Loss Reduction; and Others, as appropriate.

The practitioner will be required to exercise professional judgement based on experience and training. Step 2 is an interdisciplinary process requiring engineering, climatological, operations, maintenance, and management expertise. The practitioner must ensure that the right combination of expertise is represented either on the assessment team or through consultations with other professionals during the execution of the assessment.

4.3 Step 3 -‐ Risk Assessment In Step 3 the practitioner will identify the interactions between the infrastructure, the climate and other factors that could lead to vulnerability. These include:

Specific infrastructure components; Specific climate parameter values; and Specific minimum performance goals.

The Protocol requires the practitioner to identify which components of the infrastructure are likely to be sensitive to changes in particular climate parameters. They will be required to evaluate this sensitivity in the context of the performance expectations and other demands that are placed on the infrastructure. Infrastructure performance may be influenced by a variety of factors and the Protocol directs the practitioner to consider the overall environment that encompasses the infrastructure. At this point in the Protocol the practitioner will perform a risk assessment of the infrastructure’s




Page 33 of 166


2011

vulnerability to changing climate. The interactions identified will be evaluated based on the professional judgement of the assessment team. The risk assessment will identify areas of key concern. The practitioner will identify those interactions that need further evaluation. The assessment process does not require that all interactions be subjected to further assessment. In fact, in most assessments most of the interactions considered will ultimately be eliminated from further consideration. Some interactions may clearly present no, or negligible, risk. Some interactions may clearly indicate a high risk and a need for immediate action. Those interactions that do not yield a clear answer regarding vulnerability may be subjected to the further Engineering Analysis or recommended for additional study subsequent to the assessment. At this stage, the practitioner must also assess data availability and quality. If professional judgment identifies a potential vulnerability that requires data that is not available to the assessment team, the Protocol requires that the practitioner revisit Step 1 and/or Step 2 to acquire and refine the data to a level sufficient for risk assessment. The practitioner may determine that this process requires additional work outside of the scope of the assessment. Such a finding must be identified in the recommendations outlined in Step 5. This is a key decision point in the Protocol. The practitioner is required to determine:

Which interactions require additional assessment; Where data refinement is required; and Initial recommendations about:

o New research; o Immediate remedial action; or o Non-vulnerable infrastructure.

4.4 Step 4 -‐ Engineering Analysis In Step 4 the practitioner will conduct focused engineering analysis on climate/infrastructure interactions requiring further assessment, identified in Step 3. This step is optional. Not every interaction requires engineering analysis. Normally, the practitioner would designate items for Step 4 analysis when they deem that additional, more focused analysis will further resolve the risk profile. This may include, but is not limited to:

Interactions found to be medium risk during Step 3 that generated significant debate




Page 34 of 166


2011

amongst team members; Interactions that were found to be part of a pattern of vulnerability, regardless of the risk

assessment score; Areas where information gaps made Step 3 risk assessment problematic; or Areas where additional work would help identify mitigation responses that can be

immediately implemented. The decision to conduct Step 4 analysis is fundamentally driven by available budget, depth of study required and project scheduling constraints. It is reasonable for the practitioner to identify additional work that cannot be concluded within the project schedule and cite that work as part of the recommendations arising from the study, The Protocol sets out equations that direct the practitioner to numerically assess:

The total load on the infrastructure, comprising: o The current load on the infrastructure; o Projected change in load arising from changing climate effects on the

infrastructure; o Projected change in load arising from other change effects on the infrastructure;

The total capacity of the infrastructure, comprising:

o The existing capacity; o Projected change in capacity arising from aging and normal wear and tear of the

infrastructure; and o Other factors that may affect the capacity of the infrastructure.

Based on the numerical analysis:

A vulnerability exists when Total Projected Load exceeds Total Projected Capacity; and Adaptive capacity exists when Total Projected Load is less than Total Projected

Capacity. At this stage the practitioner must make one final assessment about data availability and quality. If, in the professional judgement of the practitioner, the data quality or uncertainty does not support clear conclusions from the Engineering Analysis, the Protocol directs the practitioner to revisit Step 2 to acquire and refine the data to a level sufficient for robust engineering analysis. The practitioner may determine that this process requires additional work outside of the scope of the assessment. Such a finding must be identified in the recommendations outlined in Step 5. Once the practitioner has established sufficient confidence in the results of the engineering




Page 35 of 166


2011

analysis, the Protocol directs them to make recommendations based on their analysis (Step 5).

4.5 Step 5 -‐ Recommendations and Conclusions In Step 5 the practitioner is directed to provide recommendations based on the work completed in Steps 1 through 4. Generally, the recommendations will fall into five major categories:

Remedial action is required to upgrade the infrastructure; Management action is required to account for changes in the infrastructure capacity; Continue to monitor performance of infrastructure and re-evaluate at a later time; No further action is required; and/or There are gaps in data availability or data quality that require further work.

The practitioner may identify additional conclusions or recommendations regarding the robustness of the assessment, the need for further work or areas that were excluded from the current assessment. In Step 5, the practitioner is also required to formulate a statement of overall infrastructure resiliency or vulnerability.




Page 36 of 166


2011

5 The Team

5.1 Practitioner Throughout the Protocol we use the term practitioner. The reader should interpret this to mean the entire assessment team. It is highly unlikely that a project proponent will identify a practitioner with all of the necessary attributes, skills, knowledge and experience in a single person.

5.2 Professional Judgement The execution of this Protocol relies on professional judgment. It is important to make a clear distinction between professional judgment and expert opinion. Expert opinion is a statement of belief, which is based on the unique expertise of the practitioner. Expert opinion is a legal term that defines a set of circumstances where a court or other judicial body may officially and legally rely upon the expert’s unique and specialized conclusions within a limited and focused area of expertise. Experts are generally challenged to establish their expertise through a detailed analysis of their credentials. More often than not, other experts, representing competing interests, will challenge their opinion. Professional judgment is a different thing, although common English language usage may often blur the boundaries between the two. Within this Protocol, when we refer to professional judgment we are referring to arriving at conclusions that are limited by the scope of experience and skills of the practitioner. This is defined by a Professional Engineer’s scope of practice. To practice within a specific discipline the engineer must demonstrate a minimum acceptable, not expert, level of training and experience. The individual engineer must only work in areas where they have the skills and training to provide service and are guided in this matter by their code of ethics. In executing the Protocol, professional judgment refers to the combined skills, training, expertise and experience of the entire team. Professional judgement is the interpretation and synthesis of data, facts and observations collected by the team and the extrapolation of that analysis to provide a judgment of how the infrastructure may respond to a specific set of conditions. The strength of the process is derived from the combined expertise of the entire team, members being limited by their own specific scope of practice. Given the multidisciplinary nature of changing climate infrastructure vulnerability assessment, it




Page 37 of 166


2011

would be exceptional to find the entire scope of skills necessary to establish professional judgment in any one individual. For example, engineers may find themselves limited in their ability to assess climate information while climate specialists may not be able to comment on the way infrastructure systems respond to specific weather events. However, together they can identify reasonable changing climate scenarios, relevant to a specific infrastructure component and pass judgment on how that infrastructure would likely respond to that particular stimulus. This is a very different situation than one expert expressing an opinion based on their unique expertise.

5.3 A Multi-‐Disciplinary Team When guided by a well-balanced team of qualified professionals, the Protocol is a very powerful tool, derived from standard risk management methodologies, tailored to assessing the impact of changing climate on infrastructure. It is quite common for practitioners to identify data gaps, poor data quality, or lack of relevant tools such as local results from regional climatic models. Often, lack of financial resources or project schedule commitments can affect the ability of the practitioner to completely address these concerns. The Protocol allows a number of avenues to proceed when these issues arise. For example,

The practitioner may identify the data gap and make a recommendation for further work outside of the context of the vulnerability assessment.

The practitioner may identify the data gap and table any further analysis on the affected parameters.

The practitioner may infill the missing data based on reasonable professional assumptions and/or applicable data from other sources, and precede with the analysis.

Lack of input data need not deter practitioners from making professionally based judgments and expressing opinions leading to recommendations. Of paramount importance in addressing the types of questions raised by the Protocol is a well-balanced team of professionals dedicated to the execution of the vulnerability assessment. The correct blend of professional and local expertise can support and validate assumptions that allow the practitioner to compensate for missing or poor quality data and account for the lack of other technical resources. Team composition and depth of experience has a very significant bearing on the veracity of the final assessment report. The following expertise is absolutely necessary on the assessment team:

Fundamental understanding of risk and risk assessment processes;




Page 38 of 166


2011

Directly relevant engineering knowledge of the infrastructure type; Climatic and meteorological expertise/knowledge relevant to the region, including as a

minimum: o Thorough knowledge of meteorology and/or climatology and related sciences.

Knowledge of Earth's climate system and its interaction with the natural and built environment. Knowledge of climate change science and models, potential impacts of climate change and possible measures to adapt to changing climate.

o Experience in working with large sets of meteorological and climatological data. Thorough knowledge of the characteristics and applications of data collection methodology for a variety of data streams including station data, remote sensing imagery and data. Knowledge of computer input and data manipulation techniques and software.

Hands-on operation experience with the specific infrastructure under assessment; Hands-on management knowledge with the specific infrastructure under assessment; and Local knowledge and history, especially regarding the nature of previous climatic events,

their overall impact in the region and approaches used to address concerns, arising. We cannot overstate the importance of local knowledge in conducting a vulnerability assessment. Local knowledge, filtered through the overall expertise of the assessment team, more often than not, will compensate for data gaps and provide a solid basis for professional judgment of the vulnerability of the infrastructure. Figure 5 outlines how the skill sets presented by the team membership interact to create a valid basis for professional judgment.




Page 39 of 166


2011

Figure 5: Team Member Skills and Interactions

5.4 The Team Leader The team leader should be an experienced professional with demonstrated experience in management of multi-disciplinary projects. In some cases, the team leader may also contribute some of the other technical and professional skills outlined above. However, in all cases the leader must be able to coordinate and prioritize the work of the rest of the team and have sufficient background and experience to consolidate findings from different disciplines and areas of expertise. These attributes are normally developed over years of professional practice. It is generally inadvisable to assign team leadership to a junior professional.




Page 40 of 166


2011

6 Fundamentals of Risk and Risk Assessment The Protocol is derived from standard risk assessment processes. As such, it is advantageous to review these concepts prior to initiating a vulnerability assessment. This will ensure that the entire team and workshop participants have a common understanding of the expectations established by the Protocol and of acceptable approaches for addressing questions that the practitioner may identify throughout the exercise. Risk is defined as the possibility of injury, loss or negative environmental impact created by a hazard. The significance of risk is a function of the probability of an unwanted incident and the severity of its consequence1. In mathematical terms:

R = P × S

Where: R = Risk P = Probability of a negative event S = Severity of the event, given that it has happened

In risk assessment, practitioners answer three questions2:

1. What can happen? 2. How likely is it to happen? 3. Given that it has happened, what are the consequences?

The Protocol guides the practitioner through a process designed to answer these questions. In risk analysis, practitioners are cautioned to ensure that their assessment of probability does not affect their assessment of severity, and vice versa. The consequence of an event is independent

1 Paul R. Amyotte, P.Eng.

& Douglas J. McCutcheon, P.Eng.; Risk Management – An Area Of

Knowledge For All Engineers; Engineers Canada, 2006

2 Tim Bedford and Roger Cooke; Probabilistic Risk analysis: Foundations and Methods; Cambridge University Press; Fourth Printing 2006




Page 41 of 166


2011

from the likelihood that the event will occur. By separating probability and severity in this way, the practitioner is able to dissect the factors that contribute to risk. Ultimately, this can yield very useful information to guide recommendations regarding approaches to risk mitigation. Practitioners can identify steps that reduce:

• The probability of an event; • The severity of an event; or • Both.

6.1 Hazard Identification – What can happen? In the Protocol, hazards are identified as interactions between identified climatic events and components of the infrastructure. The practitioner identifies conceivable climatic events that could occur in the region within the time horizon of the vulnerability assessment.

For example, the practitioner could identify that an event of 50 mm of rain in one hour is conceivable during the remaining service life of the infrastructure.

The practitioner will then review the infrastructure and determine the components and sub-components that comprise the infrastructure. This requires professional judgement. If the component analysis is not sufficiently detailed, the assessment may miss potential vulnerabilities. However, if the component analysis is overly detailed, the scope of the assessment can mushroom and become unmanageable or very expensive. Once the component analysis and climate analysis are completed the practitioner consolidates the lists. The consolidated list yields a set of interactions between climatic events and infrastructure components.

For example, the list may suggest that, during the time horizon of the evaluation, it is conceivable that the 50 mm rain event could impact culverts within the infrastructure system.

As a final step of the hazard identification the practitioner will perform a pre-screening of the identified interactions. They will judge if the identified interactions could conceivably occur. It is imperative that at this stage of the assessment the practitioner does not establish a numerical value for the likelihood of the interaction. They are assessing the reasonableness or




Page 42 of 166


2011

conceivability of the interaction. Based on professional judgment, this “sniff test” can significantly reduce the number of interactions considered in further evaluation. At the end of the hazard analysis, the Protocol will yield a set of interactions, or hazards, that will be assessed further for likelihood and severity, finally yielding a value for risk.

Hazard analysis does not identify risks.

Hazard analysis identifies a specific set of circumstances that could potentially result in a negative outcome. In the following analysis, the practitioner will establish just how likely the interaction is and the consequences of the interaction, should it actually occur.

6.2 Probability – How likely is it to happen? To determine risk, the practitioner must first assign a probability of the weather event or climate trend occurring. In some circumstances, historical data or statistics are available to guide this assessment. However, more often than not, this guidance is not available. In such cases, the probability can be assigned based on professional judgment. This is a normal procedure in risk assessment. A lack of measured data should not impose an impediment to completing the vulnerability assessment. Standard risk assessment textbooks state:

… judgment techniques are useful for quantifying models in situations in which, because of either cost, technical difficulties or the uniqueness of the situation under study, it has been impossible to make enough observations to quantify the model with “real data”.2

The Protocol provides guidance on assigning changing climate probability scores based on the available information, confidence in the information sources, and uncertainty. The Protocol uses a standardized probability score of 0 to 7, where 0 means that the event will never occur and 7 means that the event is certain. Further, the Protocol provides two different approaches to assigning these factors. Finally, the Protocol allows the practitioner to use other methods to assess probability, should these methodologies be justified given the circumstances of the current assessment. It is important to ensure that sufficient expertise, experience and knowledge be accessed to ensure a balanced and reliable estimate of the probability.




Page 43 of 166


2011

Often, a sub-committee that has specific skills and knowledge regarding local weather and climate projections may assign probability scores. In any event, at the Vulnerability Assessment Workshop (described in Section 7) participants should systematically assess the reasonableness of the probability scores. Sometimes, workshop participants will uncover inconsistencies in the probability scoring that can materially affect the overall risk profile. The combined expertise and experience of the workshop participants is designed to yield a pragmatic and realistic estimate of the probability of occurrence of weather events and climate trends.

6.3 Severity – Given that it has happened, what are the consequences? The second step in establishing a value for risk is to assess the consequences of an event, given that the event has happened. In some circumstances, historical data or statistics are available to guide this assessment. However, more often than not, this guidance is not available. In such cases, the severity can be assigned based on professional judgment. It is important to ensure that sufficient expertise, experience and knowledge be accessed to ensure a balanced and reliable estimate of the severity. In the Vulnerability Assessment Workshop, participants systematically assess each of the interactions deemed to be conceivable and reasonable by the practitioner. The combined expertise and experience of the workshop participants is designed to yield a pragmatic and realistic estimate of the severity of an infrastructure – climate interaction, given that event has occurred. The Protocol provides guidance regarding the selection of severity values. The Protocol uses a standardized severity score of 0 to 7, where 0 means no negative consequences, should the interaction occur and 7 means significant failure, should the interaction occur. Further, the Protocol provides two different approaches to assigning these factors. Finally, the Protocol allows the practitioner to use other methods to assess severity, should these methodologies be justified given the circumstances of the current assessment. During execution of the Protocol the practitioner may find that different individuals on the team or within the infrastructure owner’s organization, may have decidedly different perspectives on the severity of events. This is both normal and desirable. It is important that the practitioner identifies the rationale for these differences and considers these factors prior to assigning a final severity score.




Page 44 of 166


2011

6.4 Risk – What is the significance of the event? Finally, the practitioner is directed to determine the risk for each interaction. As previously stated, risk is a function of the probability of an unwanted incident and the severity of its consequence. Logistically, the Protocol directs the practitioner to multiply the probability and severity values derived above to establish a value for risk. If the practitioner uses the recommended probability and severity scores, the risk analysis will yield a set of risk values ranging between 0 and 49. Since, the score factors are unitless, the resulting risk values are also unitless. The Protocol then goes on to help the practitioner define criteria for further screening the risks. Low risk interactions are eliminated from further evaluation. Medium risk interactions may be subjected to further engineering analysis (Step 4 of the Protocol). High risk interactions are normally passed forward to conclusions and recommendations (Step 5 of the Protocol). In simple terms, low risk interactions pose minimal threat. Medium risk interactions MAY be significant and MAY require further refinement and analysis before the practitioner passes final judgement. High risk interactions pose a material threat and require remedial action. The Protocol identifies categories of recommendations for high risk items including, but not limited to, management action, retirement, or re-engineering and retrofit. The concept of tolerance to risk is inherent in the predefined cut-offs suggested by the Protocol. The Protocol assumes that infrastructure owner accepts a level of risk simply by operating the infrastructure. The owner accepts this level of risk as a normal consequence of the operation and may already have procedures in place to manage the risk. No activity is risk free, but a minimal level of risk is acceptable. The Protocol also assumes that as risk values increase, the owner’s tolerance to the risk decreases and they are likely to undertake risk mitigation activities to address the concern and reduce the risk to a level within their risk tolerance. At the highest level, the risk exceeds the boundaries of the owner’s risk tolerance and they will take urgent action. The Protocol allows the practitioner to adjust the cut-off values, as appropriate, based on their professional judgment and consultation with the infrastructure owner. The Protocol provides guidance to assist in “triaging” these risks.

6.5 Common Myths and Misconceptions About Risk It is important for practitioners to understand the implications of common myths and misconceptions about risk. In the Protocol, there is a significant level of involvement of laypeople. Understandably, the average layperson does not have a technical understanding of




Page 45 of 166


2011

risk. The practitioner has the responsibility to guide the layperson through the process in a technically rigorous manner. It is important to be able to identify and address the most common problems associated with risk analysis. Some of these common myths and misconceptions include:

“Hazard is risk.” It is very common for the average person to confuse the conceivability of an event with its risk. Simply because an event can be conceived does not mean that, in the real world, it will actually occur. Risk assessment considers the likelihood of an event in association with its consequence. Hazard assessment simply asks the question: “What events can I imagine that could result in a negative outcome.” “Probability is risk.” Often the average person will confuse the likelihood of an event with risk. Likelihood, or probability, is only one factor that constitutes risk. The severity of the event must also be considered. When probability is confused with risk, the impact of the event is neglected. It is possible to label high probability - low impact events as high risk. This can lead to unnecessary management action. Conversely, it is possible to label high severity – low probability events as low risk, resulting in little or no mitigative action when action is actually necessary. For example, the probability of a flood control dam experiencing a 1 in 200 year rainfall event may be small but the severity of the event could lead to catastrophic failure of the dam. Based on probability alone, this event may be identified as a very low risk whereas a more thorough analysis would reveal a much more significant level of risk. “Severity is risk.” The average person may confuse the severity of an event with its risk. High severity events are considered to be high risk regardless of their likelihood. Similarly, low severity events are considered to be low risk even though they may occur quite frequently. As above, by neglecting one key factor of risk the actual risk may not be properly assessed or managed. “Probability and severity are dependent (linked) variables.” This misconception is often the most difficult to address with a layperson. It is very challenging for the average person to separate the likelihood of an event from its consequences. For example, if they can conceive of the event, then it must be serious. The problem with this view is that it does not allow the practitioner to assess probabilities and impacts in a clinical manner. Properly executed, a risk assessment must treat severity and probability as independent variables. Although, the average person may see probability and severity as causally linked, the probability of the event is in no way related to the severity of the consequence. Severity does not cause probability, nor does probability cause severity. Probability is a function of frequency. Severity is a function of the physical nature and




Page 46 of 166


2011

physics of the infrastructure and climatic event. Risk assesses the combined implications of the two. This perspective allows the practitioner to rank the likelihood of events and the severity of events separately in order to rigorously evaluate the implications.

These concepts are technically complex and outside of the experience of the average person. It is the practitioner’s duty to be vigilant in the execution of the Protocol. They must ensure that these myths and misconceptions do not creep into the mindset of the practitioner team or workshop participants and compromise the veracity of the assessment results.

6.6 Special Cases Two situations arise in changing climate risk assessments that require special attention. These include cases where the assessment team identifies:

Very low probability and very high severity; or Very high probability and very low severity.

6.6.1 Very Low Probability – Very High Severity These situations are characterized by probability scores of “1” and severity scores of “7”. The risk score is “7”, which indicates low risk. However, should these events actually occur they are potentially devastating, resulting loss of asset or even loss of life. Given the severity and the fact that the team deems the event to be possible, even if unlikely, warrants special attention. The Protocol requires that these cases pass one additional level of scrutiny. The practitioner is required to determine if the infrastructure owner has emergency response procedures that could accommodate these rare events and, if not, the practitioner is directed to comment on the possible need for such procedures. Within the context of a changing climate risk assessment, it is reasonable for the owner or team to determine that the event is so rare that it warrants no further action. However, it is critical within the context of the risk assessment process that the question is asked and the rationale, one way or the other, clearly documented. Further, if the practitioner deems that further action is required, the Protocol requires that a recommendation regarding this matter be included within the final report.

6.6.2 Very High Probability – Very Low Severity These situations are characterized by probability scores of “7” and severity scores of “1”. The




Page 47 of 166


2011

risk score is “7”, which indicates low risk. Although these interactions are not characterized by severe weather they nonetheless warrant close scrutiny. These interactions may indicate a situation where the infrastructure could experience increased weathering resulting in high maintenance demands, increased costs or reduced overall capacity. Although they score low risk, the Protocol requires that the practitioner assess the infrastructure owner’s response to these possible outcomes. Within the context of a changing climate risk assessment, it is reasonable for the owner or team to determine that the interaction is so minor that it warrants no further action. However, it is critical within the context of the risk assessment process that the question is asked and the rationale, one way or the other, clearly documented. Further, if the practitioner deems that further action is required, the Protocol requires that a recommendation regarding this matter be included within the final report.

6.7 The Risk Matrix In risk assessment it is common to present risk results within the context of a risk matrix. The risk matrix is a Cartesian chart with probability scores listed on the x-axis and severity scores listed on the y-axis. Risk scores are presented within the body of the chart. Within the chart, areas of low, medium and high risk can be denoted with colour coding. An example risk matrix is presented in Figure 6. In this example, the areas of low, medium and high risk correspond to the suggest risk tolerance thresholds outlined in the Protocol. The special cases, described in Section 6.6 are also highlighted. The risk matrix is a visual representation of the risk profile of the infrastructure. It clearly denotes the circumstances leading to high risk interactions and areas of little immediate concern. This can be a valuable tool, assisting the practitioner in identifying interactions that are potentially very sensitive to the assumptions underlying professional judgement. For example, in the case outlined Figure 6, interactions receiving risk scores of 35 or 36 might merit closer attention. In these cases, very minor shifts in the assumptions leading to probability and severity scores may result in shifting an interaction from medium risk to high risk.




Page 48 of 166


2011

Figure 6: Example Risk Matrix




Page 49 of 166


2011

7 The Vulnerability Assessment Workshop Step 3 of the Protocol requires that the practitioner execute a workshop with representatives from the infrastructure ownership and operations teams. This is a way to draw on the combined experience of the practitioner and people who have direct contact and history with the infrastructure. This method allows the team to apply professional judgment in a transparent and consistent manner. This can be done in a technically rigorous way and yield results that can withstand professional scrutiny. Where data exists, the practitioner is directed to use it. However, if the data is missing or suspect in any manner, the practitioner is directed to rely on the professional judgment of the practitioner team and workshop participants. Thus, the workshop represents the most important phase of the evaluation. The practitioner can execute the workshop in two fundamentally different ways.

1. Practitioner Risk Assessment

At the workshop the practitioner reviews the results of their risk assessment and invites participants to review and assess the probabilities and severities of the interactions identified by the practitioner. This is a working session designed to test the practitioner’s risk profile against the experience and expertise of workshop participants.

2. Facilitated Risk Assessment

At the workshop the facilitator guides participants through a process of reviewing interactions, confirming probability scores and assigning severity scores. This is a working session designed to execute the risk assessment and establish a risk profile for the infrastructure.

Although the Protocol allows the practitioner to conduct the risk assessment through a series of one-on-one meetings where necessary, experience to date demonstrates that a properly executed workshop yields the most robust risk analysis. It is therefore strongly recommended that the practitioner use a workshop unless there are significant, compelling and material reasons to the contrary. Given the importance of the workshop, it is critical that the right mix of knowledge, experience and professional skills be present. If the practitioner team has been structured properly, the professional skills and experience should be available to the workshop. However, the




Page 50 of 166


2011

practitioner team may be missing hands-on experience with this particular infrastructure. Local knowledge regarding weather events and how the infrastructure and operations team responded to those events is critical information of interest to the practitioner team. Participants at the workshop can fill these gaps. It must be stressed that it is not sufficient to include only management and engineering staff from the infrastructure owner. Operations and maintenance staff must also participate. It is not uncommon for operations staff and management/engineering staff to have a distinctly different perspective of climate-infrastructure interactions. Events that the management team view to be very significant may already have been encountered and addressed by the operations team. This is critical input to your PIEVC assessment.

For example, the management team may view that a severe snow event could prevent operations staff from executing their duties, while the operations staff have already

experienced snow events of equal or greater severity and developed methods to address the problems they encountered. As often as not, these procedures are not formally documented

and can only be described by the affected staff.

Although these perspectives may seem trivial on the surface, they are very significant indicators of how the staff will respond during severe weather events that affect their operations responsibilities. This should emerge during the workshop discussions and forms a substantive input to the local knowledge data used by the practitioner to establish the risk profile. Generally, participants at the workshop should include:

The practitioner team; Representatives from the infrastructure management team; Representatives from the infrastructure engineering team; Representatives from the infrastructure operations team; Local expertise/knowledge regarding severe weather events in the region and climatic

trends that may have affected the infrastructure; Representatives from the organization providing climate information; Representatives from any advisory groups or technical experts who may be supporting

the vulnerability assessment; and Others deemed necessary by the infrastructure owner or practitioner team.

The workshop should follow a consistent agenda. Given the number of laypeople who may be involved, it is important to provided sufficient background on the exercise to all participants and establish the expected outcomes from the meeting. Generally, the workshop agenda should include:




Page 51 of 166


2011

A brief presentation on climatic change and the implications for the region; A brief presentation on risk and risk assessment; A brief presentation on the work completed by the practitioner to date;

o As a minimum, identifying the key interactions to be considered by workshop participants;

Introduction of the spreadsheet or matrix developed by the practitioner in compliance with Step 3 of the Protocol;

o Explanation of the infrastructure components weather events and climate trends that the practitioner deems to be relevant;

o Polling of the workshop to determine if potentially relevant infrastructure components, weather events or climate trends have been missed;

A tabletop exercise, drawing on the expertise of workshop participants, establishing probability and severity for each relevant interaction identified by the practitioner. This could be done by:

o Assigning groups to input data to hard copies of the matrix distributed to the workshop;

o Assigning groups to input data to laptops distributed throughout the workshop; o As a single facilitated discussion filling in a master spreadsheet projected to the

entire workshop; or o Other methods as deemed appropriate.

If appropriate, a site visit or tour of the infrastructure or of specific components of the infrastructure; and

A summary of findings arising from the workshop. Because of the length of the agenda, and the need for rigorous discussion, the practitioner should plan the workshop for one complete eight-hour day. Depending on assessment scope, staff availability and budget the practitioner may consider expanding the workshop to two full eight-hour days. Given the amount of professional, billable, hours that will be consumed at the workshop, it is critical that the practitioner:

Carefully plan the event in consultation with the infrastructure management and operations teams;

Schedule it to maximize productive outcomes; o Not before screening analysis is complete or before all necessary and relevant

data has been accumulated; and Provide as much validated data and background information as possible.




Page 52 of 166


2011

PIEVC has gained considerable experience in applying this Protocol to assess the vulnerability of many different types of infrastructure. The workshop approach outlined above has consistently unearthed issues that would otherwise have escaped the notice of

practitioners. For this reason, we STRONGLY recommend that practitioners use a workshop within the vulnerability assessment process. Only in cases where there are compelling and

material reasons to use alternative approaches should these alternatives be considered. Even then, we recommend that findings derived from the alternative still be reviewed with the

infrastructure owner in a workshop environment.




Page 53 of 166


2011

8 Climate Resources for use in PIEVC Protocol Evaluations

Note to Readers

This section is provided to aid practitioners’ understanding of the wealth of climate data that

is available to an infrastructure vulnerability assessment and their understanding of the strengths and weaknesses of these data formats. By no means should a practitioner attempt to employ every source of climate data nor should they necessarily endeavor to access the most

precise and expensive data set. The form of the climate data is primarily dictated by the overall technical and financial scope of the assessment at hand. In some cases, provincial governments may have already published “off-the-shelf” climate data sets for a particular

region that may be totally acceptable for the current project.

This section is intended primarily as information to provide context to the practitioner and not as a prescriptive definition of data acceptability for their assignment.

Section 1: Resources and Tools

8.1 Observed Data

Raw data – observations of weather phenomena made by human or machine at varying hours of the day, month and year at numerous locations.

Observed weather ranges in type and quality depending on site and observing method. With the beginning of weather services in Canada in the late 1800’s, there still exist a number of observing sites that have records longer than 100 years. Since World War II there have been a number of sites that supported airstrips that have evolved into major airport sites. Most of these sites are “manned” since trained human observers have been using instrumentation to observe and note the weather on a minute-by-minute basis. In comparison, climate reference sites are also “manned” but volunteers take observations only once or twice a day. The remaining sites are managed by part-time humans or are “automatic” where all of the weather is observed by instrumentation directly. The result is a mixture of data quality and quantity that poses challenges when data on a change in weather across large geographical areas is required for climate change research. To overcome




Page 54 of 166


2011

some of those challenges, national weather services, provincial environment programs and private sector concerns have instituted observational programs, specifically to increase the amount of observational data in certain geographical areas. For example, Environment Canada maintains a national archive of data, accessible to users through their Internet sites. To sustain that archive, EC manages a large national observing network in Canada and quality controls all information it gathers. Even with this effort, gaps and inaccuracies may affect the archive. In order to improve the overall utility of archival data and provide a database that could be used to determine trends in the data, EC adopted a program of homogenization that is now in use for 210 observing sites in Canada. The Global Historical Climatology Network (GHCN) is one of the primary reference compilations of temperature data used in climatology. The map presented in Figure 7 shows the 7280 fixed temperature stations in the GHCN catalogue, colour coded by the length of available record.

Figure 7: Global Climate Temperature Map

Source: WikiCommon & Goddard Institute for Space Studies (GISS) USA. http://data.giss.nasa.gov/gistemp/




Page 55 of 166


2011

Adjusted homogenized daily data – data from 210 observing sites in Canada that have been statistically adjusted to remove unclimatic trends in the data set. In order to have confidence in tracking trends in observed data, any trend that may be induced in the data set by other non-climate forces must be dealt with. For example, data sets that are composed of observations from periods where 2 different observers or sets of equipment were used must be corrected. Several techniques over the years have been used to create the homogenized data. Most recently EC has completed a statistical approach that provides a suite of 210 sites that are corrected and can be used to detect climatic trends.

8.1.1 Climate Normals In order to best identify trends or characteristics in the climate database, utilizing records of at least 10 years in length is recommended. Under the World Meteorological Organization’s Technical Regulations, (WMO, 2011.) climatological standard normals are averages of climatological data computed for the following consecutive periods of 30 years: 1 January 1901 to 31 December 1930, 1 January 1931 to 31 December 1960, and so forth. The most recent sets for Canada, of temperature and precipitation, are for recent decades 1961-1990 and 1971-2000. The next set, 1981-2010, is currently being produced and should be available in 2012. Normal values range from averages to extremes of both temperature and precipitation on a monthly, yearly or decadal scale. They are easily accessible for a number of sites across Canada but are limited to those sites where consistent observational records are available. An example of climate normals for St. John’s Newfoundland is presented in Figure 8.




Page 56 of 166


2011

Figure 8: Climate Normals for St. John’s NL, 1971-‐2000

Source: Environment Canada, 2010

8.1.2 Climate Indices Building on observed data and expanding the idea that climate records can be understood and displayed in various ways, there are also values available that can be of use to more specific users. Climate indices that range from heat wave duration to growing season can be accessed from sites that maintain national archives and normals. These values require specific calculation based on either temperature or precipitation data (or sometimes both) in order to arrive at the correct value. For example, growing degree-days can be defined as the number of days that exceed 5o C.

8.1.3 Intensity, Duration and Frequency (IDF) Curves Much like the climate indices, IDF curves are a calculated product developed from precipitation records for specific sites. They provide a graphical representation of the nature of a range of precipitation amounts for that site. By choosing various axes on a graph one can determine the frequency of heavy precipitation based on the historical observed data for that site. This can be




Page 57 of 166


2011

very valuable information when developing infrastructure design. However this approach still has the limitation that it does not provide projections of change in the future. Some consulting teams may develop site-specific IDF curves for an assessment, depending on data availability and the overall scope of the assessment. An example IDF curve is presented in Figure 9.

Figure 9: Example IDF Curve

Curves for St. John’s NL, 1949-1996. Source: Environment Canada, 2010.

8.1.4 Mapped and Gridded Data In order to better visualize climate data on a geographical scale, routines have been created that can prepare the data, either raw observed data or climate indices, on a grid to be mapped at a defined scale. This provides the user with an opportunity to view the data on a spatial scale in comparison to the usual temporal scale. Such visualizations have been used with great success to map combinations of indices and determine causal relationships between various climate parameters. All of the above-mentioned data resources are based on observational data taken at sites across




Page 58 of 166


2011

Canada over the past 150 years. They are all subject to data quality issues related to type of observation technique and location. As well, decisions made on this data are limited by the historical past that the data represent. An example contour map of extreme rainfall is presented in Figure 10.

Figure 10: Contour Map of Extreme Rainfall 24 Hour Duration over Atlantic Canada

Source: Atmospheric Hazards Website, Environment Canada, 2010.

8.2 Model Projections In order to study potential changes in future climate appropriately, mathematical models must be utilized. These models allow researchers to simulate climate and compare results with past observations. Once these models are validated in this way, researchers can then be confident that future climate projections of various climate parameters can be useful in understanding future change. While infrastructure design depends heavily on past climate to define design parameters,




Page 59 of 166


2011

decision-making based solely on past climate will not adequately capture the entire operating environment and satisfy the designers need for due diligence. Model results that provide the user with estimates of future climate parameter changes are necessary.

8.2.1 Model Varieties Even though climate models have been used by researchers since the late 1800’s, by the 1980’s researchers began using climate models to not only represent recent climate but to provide projections into the future. Early models were focused on the global energy balance and were simplified mathematical representations. As development on weather forecasting models advanced and computer power increased, climate models became more sophisticated and were designed to provide much more detail and generate more specific parameters. Modeling science has now advanced to the point where regional climate models are now being used to determine much finer scale results than ever before. The conceptual structure of a coupled atmosphere-ocean general circulation model is presented in Figure 11.




Page 60 of 166


2011

Figure 11: Conceptual Structure of a Coupled Atmosphere-‐Ocean GCM

Source: Environment Canada, Climate Research Branch (from H. Hengeveld, Climate Change CD, Part 3, slide 6),

2006 The most common model in use currently is the Global Climate Model (GCM) (or General Circulation Model). Over a dozen different countries run versions of this model, utilizing various mathematical techniques, computer power and physical characterizations of climate. Most of these models have a “coupled” feature where the oceans interact with the atmosphere above them in key ways. All of these models have been validated against past climate and can be used to determine future trends with confidence. Examples of GCMs from different national modeling centers are outlined in Figure 12.




Page 61 of 166


2011

Figure 12: Examples of GCMS

8.3 Model Output The nature of climate model output differs greatly from that determined by weather models and even seasonal models based on climate model structure. Weather forecasts are determined by initializing a model with observed data and running forward. The results begin to differ from reality soon after initial time and these models need to be re-initialized with data every 12 hours to improve the forecast in the first 24 hours. Climate predictions (seasonal forecasts) also depend on input of atmospheric and ocean conditions at the initial time to best determine that prediction, forward as much as a decade. In comparison, climate model projections are driven by “external forcing” such as greenhouse gas (GHG) emissions and concentrations over the next century. Instead of “next day forecasts”, climate models provide projections of climate parameters over 30-year periods (tri-decades). Those are defined as the 2020s (2011-2039), 2050s (2040-2069) and 2080s (2070-2099). Models

Modeling Centre Country Model(s)

Commonwealth Scientific and Industrial Research Organization (CSIRO)

Australia CSIRO-Mk2

Max Planck Institüt für Meteorologie (formerly Deutsches Klimarechenzentrum, DKRZ)

Germany ECHAM4/OPYC ECHAM3/LSG

Hadley Centre for Climate Prediction and Research

UK HadCM2 and HadCM3

Canadian Centre for Climate Modeling and Analysis (CCCMA)

Canada CGCM1, 2 & 3

Geophysical Fluid Dynamics Laboratory (GFDL)

USA GFDL-R15 and GFDL-R30

National Centre for Atmospheric Research (NCAR)

USA NCAR DOE-PCM

Center for Climate Research Studies (CCSR) and National Institute for Environmental Studies (NIES)

Japan CCSR-NIES




Page 62 of 166


2011

are run every year, out to a specific time horizon (typically 100 years), with concentrations of GHGs varying (increasing) each time, providing values in the tri-decades. Since the entire model run continues to the end of the century, climate parameter calculations will vary from reality as it proceeds in time, creating uncertainty in the projections. When the model projects a value, for example, for temperature, that value has 2 specific uncertainties; first, it has generated this number from its own base climate and, second, it has applied the GHG scenario chosen. The model’s base climate will always be different from reality. Models used for climate projections have been validated against current climate, thereby, ensuring they simulate climate within reasonable criteria. However that doesn’t mean they match current climate exactly, especially for specific locations. Therefore the projected value will never provide an exact match for reality. Standard procedure to account for this uncertainty is to use the “delta” of the climate parameter rather than the absolute value. Subtracting the projected value from the model’s base climate will provide that delta value. It can then be added to current observed climate to get an appropriate scenario value for that location. Climate Model Uncertainty At issue when determining the appropriate future values for infrastructure planning is how much uncertainty exists in the climate model results. Typically the amount of uncertainty, or potential variance from reality, may be different from one model to the next. Climate model result uncertainty has several sources (see Table 1 for more detail):

1. Unpredictability 2. Structural Uncertainty 3. Value Uncertainty

To deal with these uncertainties, several approaches can be taken:

1. Ranges of results derived from choosing specific scenarios including multiple GHG emission scenarios.

2. Use ensemble techniques to compare models in order to understand potential structural differences.

3. Determine best quality data sets. Use probability distribution techniques to display results.

Each model research facility has conducted experiments where individual models were run with scenarios of GHG emissions and multi-model comparisons were completed. Results were reported in 2007 in the IPCC Fourth Assessment of Climate Change. Model output from all of




Page 63 of 166


2011

these experiments is available on-line from individual model sites as well as the IPCC site. Figure 13 outlines the range of uncertainties that models may exhibit.

Figure 13: A Simple Typology of Model Uncertainties

Type Indicative Examples of Sources Typical Approaches or Considerations

Unpredictability

Projections of human

behavior not easily amenable to prediction (e.g. evolution of political systems).

Chaotic components of complex systems.

Use of scenarios spanning a

plausible range, clearly stating assumptions, limits considered, and subjective judgments.

Ranges from ensembles of model runs.

Structural

Uncertainty

Inadequate models Incomplete or competing

conceptual frameworks Lack of agreement on

model structure Ambiguous system

boundaries or definitions Significant processes or

relationships wrongly specified or not considered.

Specify assumptions and system

definitions clearly Compare models with

observations for a range of conditions

Assess maturity of the underlying science and degree to which understanding is based on fundamental concepts tested in other areas.

Value

Uncertainty

Missing, inaccurate or

non-representative data Inappropriate spatial or

temporal resolution Poorly known or

changing model parameters.

Analysis of statistical properties

of sets of values (observations, model ensemble results, etc)

Bootstrap and hierarchical statistical tests

Comparison of models with observations.

From Guidance Notes for Lead Authors of the IPCC Fourth Assessment Report on Addressing Uncertainties.




Page 64 of 166


2011

8.3.1 Finer Scale Most GCM’s operate on a geographical scale of 400km x 300km. This means that an area the size of New Brunswick is covered by only one model “grid” value representing all of that geographical area. This presents a problem for specific regions of the country since local climates can vary dramatically, especially in the coastal or mountain regions. In order to resolve some of this issue, researchers have been taking two approaches. Regional climate models (RCMs) have been designed to provide climate parameter results on a 50km grid scale. They depend on the larger GCMs to initiate the model process then they focus on producing parameters on the smaller scale. This process is referred to as dynamic downscaling. Statistical downscaling is a much less intense computational process. It uses a statistical relationship between the observed climate at a specific site and the climate viewed through the GCM at the larger scale. Since the GCM has an excellent ability to project forward, once the relationship is created, the user can determine future values using the combined power of the relationship and the GCM. An example of the output from a typical statistical downscaling exercise is presented in Figure 14.




Page 65 of 166


2011

Figure 14: Example of Output from Statistical Downscaling

Example of annual mean screen (2m) temperature (°C) in 1971-‐1990 simulated by CRCM3.6.1. Note detail in more mountainous areas over western Canada.

Source: Environment Canada, 2010 Both of these techniques have advantages. While RCMs require large amounts of computer power they provide much regional detail. Statistical processes use relationships that are static and must be assumed across future projections but can be tuned to specific location climatology. Both provide an opportunity to get climate parameter values at finer geographical scales.




Page 66 of 166


2011

Section2 -‐ Application of Resources

8.4 Who to Ask for Climate Data To build a team that will appropriately address changing climate issues as they pertain to infrastructure design, the practitioner must acquire expertise based on:

Thorough knowledge of meteorology and/or climatology and related sciences; Knowledge of Earth's climate system and its interaction with the natural and built

environment; Knowledge of climate changes science and models, potential impacts of changing

climate and possible measures to adapt to changing climate; Experience in working with large sets of meteorological and climatological data; Thorough knowledge of the characteristics and applications of data collection

methodology for a variety of data streams including station data, remote sensing imagery and data; and

Knowledge of computer input and data manipulation techniques and software.

While there may be a variety of experience most climate consultants who match this criteria will be able to access the data and tools noted in Section 1 and provide “scientifically defensible” approaches to developing model projections.

8.5 What to ask for Request for climate projections should be based on the climate baseline work and thresholds that have been determined from experience with infrastructure sensitivity. Each project will dictate what specific projections and thresholds are appropriate. In some cases, information may have been developed for other projects that could be used in the current assessment. As well, some provincial jurisdictions are creating generic climate information that may be acceptable for the assessment. Please refer to Appendix A for suggested climate and infrastructure parameters.

8.6 What to expect In order to proceed properly with assigning event probabilities and ultimately, a risk score, the practitioner expects to receive climate values and advice that will provide the necessary




Page 67 of 166


2011

information to determine those probabilities. Most likely they will rely on the climate expertise of the consultant hired to gather and interpret the climate data and projections. However there are aspects of this process that make delivery of those values not as straightforward as the practitioner may assume or expect. Some of these aspects are outlined in the following points.

1. Variances in availability of data sets may lead to there being “holes” in the overall understanding of the climate baseline for the site in question. Usually this applies to regions of the country that are more sparsely populated, away from networks of roads, electrical power capability. As well, given the vast geographical area of Canada it is fairly obvious that we simply cannot afford to monitor climate everywhere, all the time. Strategies for acquiring the most information possible may have to include remote sensing (satellite, radar, etc.) to “fill in the holes” especially regarding precipitation.

2. Length and quality of data sets will vary based on the source of that data. The more

reliable sources tend to be government managed observing networks. However some private sector data sets, including climate projections published by the insurance sector, have been maintained for long periods of time and could be quite reliable. Some rigor should be applied in examining data sets made available from differing sources. For example, data sets shorter than 10 years are not suitable for examining climate trends. As noted earlier, WMO has set a 30-year criteria for data sets upon which “normals” are calculated. As well, quality control processes that examine the validity of the data (based on various statistical tests) should have been applied to the data sets.

3. Approaches used to determine appropriate climate projections must follow standard

(best) practices as determined in current scientific literature, so as to be “scientifically defensible”.

4. Since these approaches are constantly being tested in the research community and

evolving into various forms over time it can be seen as a “moving target” to the practitioner.

To aid the practitioner in understanding the “moving target” nature of determining climate model projections, here are some specific aspects of the process to keep in mind.

5. The background, training and focus of each climate researcher vary. Some researchers focus on designing climate models, with an objective to lessen the uncertainty in the




Page 68 of 166


2011

model results as much as possible. Others focus on individual physical processes that lead to climate change on a global or regional scale. Since that focus is much narrower, there are usually differences of opinion in how to incorporate the physical processes into models to improve results. These differences are normally debated and then tested through the peer-review process at the basis of scientific endeavor. However from a practitioner perspective such debate may simply appear to be pointless scientific discussion.

6. In order for such discussion and debate to be productive, the climate science community

uses a suite of climate terms that they understand and that refer to specific scientific methods. However the use of these terms may lead to confusion when this discussion becomes part of a multi-disciplinary effort such as the Protocol.

7. For example, climate scientists comment on what they think future climate will be based

on climate “sensitivity”. Climate sensitivity is a measure of how responsive the temperature of the climate system is to a change in the radiative forcing, such as increasing greenhouse gases, and is usually expressed as the temperature change. Such terminology can be construed or misunderstood when used in a multi-disciplinary situation.

8. In order to generate climate projections, complex mathematical models must be used.

While these models have improved exponentially over the past decade, they are still models and, by definition, contain a level of uncertainty and error that has to be accounted for if we wish to understand and apply the results.

9. To attain the best results (i.e. the least uncertain results) a “scenario” approach must be

taken. In this approach it is recognized that each model result, however different from each other, is still a valid future climate projection. To take advantage of all that these models offer, they are “run” over the same time horizons, using common greenhouse gas emission scenarios. This “ensemble” technique not only provides a “mean” value for each parameter but that value has less error (uncertainty) than values generated separately by each individual model. Please refer to http://cccsn.ca/ for in depth discussion of scenario building and ensemble results.

10. Even so, these values are still projections and not forecasts. Uncertainty still exists based

on the combined weaknesses in the models, the choice of emission scenario, geographical area of study, etc. In other words, these models are not designed to provide a forecast




Page 69 of 166


2011

daytime high temperature for July 15, 2050. A more appropriate model result would be an average high temperature for any day in any July during the tri-decade that spans 2041-2070 (2050s).

11. The accuracy of the model results is partially dependent on the uncertainty existing in the

approach used to create the projections. If the practitioner is seeking “error bars” in order to best classify the results, they are not easily produced with these types of model results or approaches. Utilizing a scenario approach allows the climate researcher to provide results that have the “highest likelihood” of occurring with the least uncertainty. However measuring that level of uncertainty can be problematic. Ultimately the results can be provided with some “confidence level” that advises the practitioner how they may proceed, utilizing that specific climate projection.

12. One indicator of accuracy could be the range of values available from a scenario process.

For example, results for Atlantic Canada based on an ensemble process (CCCSN) are a range based on low, medium and high emission scenarios. Projected annual temperature change across that range is (on average) 2C to 2.8C. Therefore it can be stated that the most likely result, given current science and modeling techniques, has an uncertainty of 0.8C. To be clear, this is not an “error”; simply the range of results based on specific model assumptions and a scenario approach.

13. Another indicator could be a probability distribution of future results. Most climate

parameters, when examined as a “probability of occurrence”, provide a distribution that is mathematically described as Gaussian. An example of a Gaussian distribution is presented in Figure 15. From such a distribution statistical values such as the standard deviation can be calculated and used as a measure of accuracy. In this case, the accuracy is a value of the variability of that result. For example, the median of the distribution provides the most frequent occurrence while values in the “tails” of the distribution are the most rare occurrences.




Page 70 of 166


2011

Figure 15: Example Gaussian Distribution

Combination of shift in mean and change in variance creates new future climate. Source: IPCC 2001. Climate Change 2001: The Scientific Basis Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change [Houghton,J.T. et al. (eds.)]. Cambridge University Press, Cambridge, UK. Figure 2.32.

8.7 Conclusion As stated earlier, utilization of climate model projections can aid infrastructure planning greatly. If care is taken in acquisition, interpretation and application of this information, development of infrastructure plans on 20, 50 or 100 year time horizons can be very successful.




Page 71 of 166


2011

8.8 References

i. Personal Communication, Flato, Greg & Lines G. June 13, 2011.

ii. IPCC, 2007: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

iii. IPCC, 2005. Guidance Notes for Lead Authors of the IPCC Fourth Assessment Report on Addressing Uncertainties. Supporting Material, July 2005

iv. IPCC-TGICA, 2007: General Guidelines on the Use of Scenario Data for Climate Impact

and Adaptation Assessment. Version 2. Prepared by T.R. Carter on behalf of the Intergovernmental Panelon Climate Change, Task Group on Data and Scenario Support for Impact and Climate Assessment, 66pp.

v. CICS (Canadian Institute for Climate Studies), 2003.

http://www.cics.uvic.ca/scenarios/WMO, 2011. Guide to Climatological Practices, 2011 Edition, 117pp. WMO-No. 100.




Page 72 of 166


2011

9 Establishing Changing Climate Probability Scores

9.1 Basis The practitioner must first clearly define the set of circumstances for which they are assigning probability scores. Results from earlier studies clearly indicate that this analysis requires much more than simply identifying whether a particular climate parameter will change over the time horizon of the assessment. The idea of identifying hazards assumes that there is a preliminary screening of the potential events to determine those that could result in the infrastructure (infrastructure component) failing to meet its operational objectives. This preliminary screening is accomplished through the Infrastructure Response and Yes/No Analysis outlined in the Protocol. Only interactions that have potential infrastructure responses to a specific weather event and pass the Yes/No analysis should be considered. Otherwise, the entire process may become cumbersome and resource intensive.

9.2 The Importance of Differing Interpretation of Key Words Across Disciplines Interdisciplinary practitioner teams can encounter difficulties arising from differences in team members’ uses and understanding of common terms as they apply to their areas of expertise. These differences can lead to disagreements within the team that, when evaluated, are not material differences in opinion but are rather communication gaps arising from different definitions of common language. Of particular note in executing an engineering vulnerability assessment is the meaning and use of the word “probability”. In pure science, probability is an exact numerical value, normally reported with explicit precision and accuracy. It is normally a calculated value with assigned statistical error bands and associated analysis. In operations and management, probability is often used more loosely to express knowledge or belief that an event will occur. Although, these differences are subtle, they can lead to considerable debate among team members when they are asked to assign probability scores within an assessment. These issues may stall the evaluation process for a time while the team sorts out these differences. For example, some members may be very comfortable assigning a score from the scales provided within the Protocol while other team members may argue that this approach is neither scientific




Page 73 of 166


2011

nor meaningful. Based on this understanding, we recommend that the practitioner clearly define to the team that the probability scoring exercise is not a quantitative process. Rather, the Protocol outlines a process based on informed professional judgment and decision-making. The Protocol is asking the team to assign a SCORE for the probability of the events under consideration.

9.3 Definition The Protocol defines probability as:

The likelihood of an event occurring 9.3.1 Clarification The practitioner should be careful not to focus on the consequence of the event when establishing the likelihood score. To better define the probability scoring process, the practitioner should specifically evaluate the likelihood of:

• A climate event; Triggering a defined threshold; During the time horizon of the assessment.

9.4 Professional Judgement and the Practitioner Team Professional judgement is a critical element of assigning changing climate probability scores. Within the context of a vulnerability assessment, professional judgement refers to the combined professional expertise, knowledge and wisdom of the entire team. It is critical that as much information and insight from different professional backgrounds be applied in forming the ultimate professional judgment. Climate specialists can provide a fundamental understanding of climate scenarios and climate science that is well outside of the normal expertise of the rest of the team. Similarly, operations and management personnel can provide insight regarding the “real-world” behaviour of the infrastructure under weather conditions similar to those projected by the climate specialists. Finally, the engineering team can provide insight into how the infrastructure is supposed to operate under design conditions and the limits of the infrastructure’s functional capacity.




Page 74 of 166


2011

Decisions based on any one element of the team’s expertise alone may miss critical information from the other areas of expertise resulting in a potentially erroneous assessment of probability. One area that deserves special attention is the treatment of climate projections. Climate specialists will provide projections and, if asked, will provide context regarding the uncertainties and the reliability of the projections based on their unique understanding of the modelling scenarios, mathematics and other assumptions inherent in the modeling. They may provide this information as very precise estimates of statistical uncertainty. For their part, the engineering team may be less concerned with precision and much more concerned about the accuracy of the projected trend. The vulnerability assessment provides a coarse screening of risks into three categories – High, Medium and Low. Some of the precision of the modeling may be lost in the trending approaches used by the engineering and management team. This is not to say that precision in not important. However, the screening analysis conducted in the risk assessment (Step 3 of the Protocol) may not warrant detailed statistical precision. Once Step 3 is complete, the team may identify areas where much more detailed and precise climate information is necessary to fully resolve an identified risk. Under these circumstances the team may ask for focused studies in those specific areas, either as part of Step 4 of the Protocol or as additional follow-up study. This provides some perspective on the interaction between the various players on the team. At this stage of the assessment, the engineering team is looking for trends and the scientific team may wish to provide specific and highly contextual forecasts. It is important that the team fully appreciates both:

The context of the climate projection that is provided; and The experience of the engineering community with managing uncertainty.

The team leader must be vigilant to ensure that all team members are provided the opportunity to contribute and that the rest of the team understands their contributions. With this understanding, the team can make sound judgments regarding interaction likelihood scores.

9.5 Defining Thresholds In applying the Protocol, each climate parameter is assigned an associated threshold value that is specific to the infrastructure being considered, for example, the number of days with temperature greater than 30oC. Thus, the assessment does not simply evaluate the impact of higher temperatures on infrastructure. Rather, the assessment considers the frequency and magnitude of




Page 75 of 166


2011

climate parameters triggering defined threshold values. These thresholds may be defined from a variety of sources including, but not limited to:

Design standards; Operational standards; Rules of thumb; Maintenance guidelines; Codes of practice; Engineering/design practice literature; Experience (past events); Professional judgement Etc.

The practitioner should note that some threshold values define maximum conditions while others define minimum conditions relative to a particular infrastructure service level. For example, the threshold may define a maximum temperature above which the infrastructure may start to exhibit loss of function. Conversely, the threshold could be a minimum temperature below which the infrastructure may start to exhibit loss of function. For this reason, we advise that the practitioner use the word “trigger” when discussing these interactions. This avoids the confusion of language that, for example, could suggest that a minimum temperature exceeds a minimum threshold value. Although this language is technically correct, it may lead to confusion among members of the team. For each climate parameter, the practitioner should define a corresponding threshold value. These threshold values should be shared with the climate specialists who can then tailor their efforts to provide climate projections relevant to the specified threshold. Examples of thresholds are presented in Figure 16. Practitioners should note that the thresholds applied in any given assessment are specific to the infrastructure under consideration and the area where the infrastructure is located (thresholds for rain or high temperatures will vary according to region) and that the examples provided in Figure 16 may not apply to all infrastructure assessments. These examples are provided for reference only.




Page 76 of 166


2011

Figure 16: Examples of Infrastructure Threshold Values

Climate Parameter Infrastructure Indicator

High Temperature Number of days with maximum temperature exceeding 30o C

Low Temperature Number of days with minimum temperature below -24o C

Temperature Variability Number of days with daily temperature variation of more than 24o C

Freeze / Thaw 17 or more days where maximum temperature > 0o C and minimum temperature <0o C

Climate events that do not interact with the infrastructure, do not present the opportunity to trigger a threshold, or that are irrelevant to the normal functioning of the infrastructure, are not assigned risk scores. Normally, these interactions are screened out of the evaluation process through the Yes/No and Infrastructure Response Analysis.

9.6 Frame of Reference If a projected changing climate event results in more frequent triggering of a threshold value, the practitioner would assign a higher probability score. However, if the change were projected to result in a significant decrease in trigger events, the practitioner would assign a low likelihood score. It is important that the practitioner maintain this frame of reference throughout the execution of the assessment. Should positive and negative likelihood outcomes be mixed together, assessing the overall risk to the infrastructure may become very misleading and confusing.

9.7 Methodologies for Assigning Probability Scores

9.7.1 Two Different Approaches Probability scoring can be worked in two ways.

1. Probability that, over the time horizon of the assessment, the climate parameter will change in a way that triggers an infrastructure-specific threshold.




Page 77 of 166


2011

2. Probability of a climate event triggering an infrastructure-specific threshold in the current

environment. Probability of a climate event triggering an infrastructure-specific threshold in the future environment. Probability of change is the difference between the two.

Both methods are acceptable, yielding slightly different cuts at the projection and interpretation. However, the changing climate implications are exposed by either approach.

9.7.2 Considerations Affecting Probability Scores Assignment of probability scores is an informed decision based on the professional judgment of the practitioner validated through the expertise of participants at the risk assessment workshop. Where appropriate, input from climate specialists can significantly improve overall confidence in the scoring. Practitioners should note that these processes are not precise numerical computations. Rather, this is a consultation process designed to generate dialogue between the various professionals engaged in the assessment. The intent is to draw on the combined professional judgment of the team to score (or rate) the likelihood of projected climatic events. This draws upon standard engineering practices that evaluate statistical, technological and resource limitations to assign functional parameter values that allow the advancement of an issue. Most commonly, this is associated with the application of safety factors that accommodate the uncertainties associated with a parameter to ensure estimates that err on the side of the overall integrity of the engineered system and, most importantly, public safety. Generally, professional judgment on scoring may be guided by five considerations. A: Will climate conditions, relevant to the infrastructure, change over the time horizon of the assessment? The practitioner may consider input data from a number of sources (grey or published literature, experience, etc.) in addressing this issue, including, but not limited to:

Weather and Climate Science o Scenarios based on climate model output o Analysis of weather patterns o Statistical analysis o Local knowledge




Page 78 of 166


2011

Professional judgement of the members of the team B: Will thresholds be triggered more often, the same as, or less than in current conditions? Change that results in more frequent triggering of thresholds supports a higher likelihood score. Change that results in the same number of triggering events supports no change in likelihood scoring. Change that results in fewer triggering events supports a lower likelihood score. The practitioner may consider input data from a number of sources in addressing this issue, including, but not limited to:

Weather and Climate Science o Scenarios based on climate model output o Analysis of weather patterns o Statistical analysis

Professional judgement of the members of the team C: What is the impact of the projected change in magnitude of the climate event on the frequency of trigger events? Climatic events that become more intense may have a greater likelihood of triggering thresholds more frequently in the future. If the climatic event is projected to decrease in magnitude this may result in fewer trigger events in the future, yielding a lower likelihood score overall. The practitioner may consider input data from a number of sources in addressing this issue, including, but not limited to:


Professional judgement of the members of the team D: What is the projected impact of the change in frequency of climate events on the




Page 79 of 166


2011

frequency of trigger events? Climatic events that become more frequent may have a greater likelihood of triggering thresholds. The practitioner may consider input data from a number of sources in addressing this issue, including, but not limited to:


Professional judgement of the members of the team E: How robust are the results of the climate projections? The practitioner’s confidence in the climate projection can have a mitigating effect on the overall likelihood score. For example, regional climate projections may indicate a change in climatic parameters that would exceed design or operational thresholds. Notionally, this would support a higher probability score. However, the climate specialists may state that the model results are highly variable or that there is a high level of uncertainty associated with the forecast. In considering this information the practitioner may lower the likelihood score to account for the uncertainty of the forecast. Conversely, the climate specialists may anticipate an improvement in climate conditions but with high level of uncertainty. This may result in the practitioner increasing the overall likelihood score. The practitioner may consider input data from a number of sources in addressing this issue, including, but not limited to:


Professional judgement of the members of the team

9.7.3 Assigning the Probability Score The final probability score is a function of all of the input parameters that the practitioner




Page 80 of 166


2011

considers.

P = ⨍ (A, B, C, D, & E) There is no quantitative methodology for executing this analysis. Rather, the practitioner team must weigh:

Each factor; and The influence of the factors on each other.

Based on this evaluation the practitioner team can make an informed decision regarding the overall probability score. The score is a practitioner-assigned value reflecting the probability of changing climate causing a change in threshold triggering events.

9.8 Probability Scoring Worksheet The practitioner is encouraged to use the climate-infrastructure interaction Probability Scoring Worksheet that is included in the Step 2 Worksheet package associated with the Protocol. For each climate parameter – infrastructure threshold pair, the Protocol and worksheet pose the following questions: A: Will climate conditions, relevant to the infrastructure, change over the time horizon of the assessment??

Yes No

B: Will thresholds be triggered more often, the same as, or less than current operation?

+ = More 0 = Same - = Less

C: What is the impact of the projected change in magnitude of the climate event on the frequency of trigger events?




Page 81 of 166


2011

H = High M = Medium L = Low Not Applicable

D: What is the projected impact of the change in frequency of climate events on the frequency of trigger events?

H = High M = Medium L = Low Not Applicable

E: How robust are the results of the climate forecasts?

H = High M = Medium L = Low

The worksheet provides open cells to allow the practitioner to document their considerations and finally a location for recording the assigned likelihood score. In some cases, the practitioner may deem that one (or more) of the questions is not relevant to the climate-threshold pair being considered. In such cases, the method allows an answer of not applicable. The method is not designed to be prescriptive. Rather, it outlines a series of considerations that the practitioner may wish to incorporate into their professional judgment and decision-making process. The practitioner is free to include other considerations within their assignment of probability scoring, as they deem appropriate. The important feature of this method is the requirement to document the rationale for the risk score assignment. An example probability assignment is outlined in Figure 17.




Page 82 of 166


2011

Figure 17: Example Probability Scoring Exercise

Will the Interaction

Change Over Tim

e Horizon of

Assessment?

More-‐Sam

e-‐Less?

Projected Change in Magnitude?

Projected Change in Frequency

Robustness of Forecast?

Professional Judgment

Probability Score

Y/N

+ 0 -‐

H M L

H M L

H M L

Comments 0-‐7

☞ ☞ ☞ ☞ ☞ ☞ P =⨍ (A,B,C,D, & E)

Climate Parameter

Infrastructure Indicator A B C D E ☞ P

High Temperature

Number of Days with maximum temperature exceeding 30oC

Y + H H H Regional climate modeling forecast suggests that both frequency and magnitude will increase from an average of 0.6 to ~ 3 days per year. The climate specialists have a high level of confidence in the projection. This is a significant change. The projection is consistent with other predictions. This suggests that the likelihood of change is relatively high.

6

9.9 Scoring for Present and Future Climate Scenarios Practitioners who are using the alternate approach to assessing changing climate risk may apply the same fundamental methodology. In this case, the questions would be somewhat modified, as follows:




Page 83 of 166


2011

A: Are the thresholds triggered?

Yes No

B: What is the impact of the magnitude of the climate event on the frequency of trigger events?


C: What is the impact of the frequency of climate events on the frequency of trigger events?


D: How robust is the climate projection and/or weather data?


Probability scoring sheets would be completed for both the present and future climates and risk analysis completed for both cases. Changing climate impacts are determined from the change in overall risk profile from the baseline to the future climate case. In some cases, using these questions to establish current probability scores will be moot. If the practitioner has access to established weather records, they may be able to simply apply that information to establish the score for the current environment. However, often the team will calibrate the baseline and projected climate through climate scenarios analysis. In this case, it may be necessary to use the questions to assess the robustness of the data being used for the current climate risk scoring process. These are decisions that the practitioner would make on an assessment-by-assessment basis and the approach will be dictated by the type of climate data being used and the team’s confidence in that data. An example of assigning a score to a current (baseline) weather event is presented in Figure 18.




Page 84 of 166


2011

Figure 18: Example Baseline Weather Event Probability Scoring Exercise

Thresholds Triggered?

Magnitude of Event

Frequency of Event

Robustness of Forecast

Professional Judgment

Probability Score

Y N

H M L

H M L

H M L

Comments 0-‐7

☞ ☞ ☞ ☞ P =⨍ (A,B,C,D, & E)

Climate Parameter

Infrastructure Indicator B C D E ☞ P

High Temperature

Number of Days with maximum temperature exceeding 30oC

Y H H H Meteorological data suggests that the event occurs frequently and that temperatures are often well in excess of the threshold value. Environment Canada reports a high level of statistical confidence in the data. This suggests that the likelihood of the event triggering the infrastructure threshold is relatively high.

6




Page 85 of 166


2011

10 Establishing Severity Scores

10.1 Basis Severity scoring follows many of the same principles as those described in Section 9. As applied in the Protocol, severity scoring is fundamentally an exercise in the application of professional assessment and judgment. This is informed by the expertise and experience of the assessment team and though consultation with infrastructure owner staff who may have experience in dealing with similar situations to those under consideration. The practitioner should never underestimate the value of site-specific experience in finalizing the severity scoring. As well, practitioners should avoid the assumption that the majority opinion regarding the severity of an event is correct. Sometimes, only one individual may have hands-on experience relevant to their particular severity score. It is very important for the practitioner to test the basis for the severity score values proposed both within the practitioner team and from other sources. This process can unearth circumstances overlooked by the rest of the team that could dramatically change the final severity score result.

10.2 Definition The Protocol defines severity as:

The severity of an event, given that it has happened.

10.3 The practitioner’s Professional Judgement and the Practitioner Team Professional judgement is a critical element of assigning severity scores. Within the context of a vulnerability assessment, professional judgement refers to the combined professional expertise, knowledge and wisdom of the entire team. It is critical that as much information and insight from different professional backgrounds be applied in forming the ultimate professional judgment. Operations and management personnel can provide insight regarding the “real-world” behaviour of the infrastructure under weather conditions similar to those projected by the climate specialists. The engineering team can provide insight into how the infrastructure is supposed to operate under design conditions and the limits of the infrastructure’s functional capacity.




Page 86 of 166


2011

Decisions based on any one element of the team’s expertise alone may miss critical information from the other areas of expertise resulting in a potentially erroneous assessment of severity.

10.4 Thresholds As describe in Section 9, in applying the Protocol, each climate parameter is assigned an associated threshold value that is specific to the infrastructure being considered. These thresholds may be defined from a variety of sources including, but not limited to:

Design standards; Operational standards; Rules of thumb; Maintenance guidelines; Codes of practice; Engineering/design practice literature; Experience (past events); Professional judgement Etc.

The practitioner should note that some threshold values define maximum conditions while others define minimum conditions relative to a particular infrastructure service level. For example, the threshold may define a maximum temperature above which the infrastructure may start to exhibit loss of function. Conversely, the threshold could be a minimum temperature below which the infrastructure may start to exhibit loss of function. For this reason, we advise that the practitioner use the word “trigger” when discussing these interactions. This avoids the confusion of language that, for example, could suggest that a minimum temperature exceeds a minimum threshold value. Although this language is technically correct, it may lead to confusion among members of the team. Climate events that do not interact with the infrastructure, do not present the opportunity to trigger a threshold, or that are irrelevant to the normal functioning of the infrastructure, are not assigned risk scores. Normally, these interactions are screened out of the evaluation process through the Yes/No and Infrastructure Response Analysis.

10.4.1 Considerations Affecting Severity Scores Assignment of severity scores is an informed decision based on the professional judgment of the




Page 87 of 166


2011

practitioner validated through the expertise of participants at the risk assessment workshop. Practitioners should note that these processes are not precise numerical computations. Rather, this is a consultation process designed to generate dialogue between the various professionals engaged in the assessment. The intent is to draw on the combined professional judgment of the team to score (or rate) the severity of projected climatic events. This draws upon standard engineering practices that evaluate statistical, technological and resource limitations to assign functional parameter values that allow the advancement of an issue. Most commonly, this is associated with the application of safety factors that accommodate the uncertainties associated with a parameter to ensure estimates that err on the side of the overall integrity of the engineered system and, most importantly, public safety. The Protocol directs practitioners to identify a set of infrastructure responses for each infrastructure component. These responses define ways that the infrastructure could conceivably react to external stimuli. For example, more maintenance may be required. The practitioner would normally consider the identified infrastructure responses to inform their deliberations and guide their insight into how severe an interaction may be. It is important to understand that often the practitioner will identify a number of outcomes from an identified interaction. These may include reactions in several of the identified infrastructure responses. For example, the event may require more frequent maintenance and also reduce the service life of the infrastructure. All of these factors enter into the professional judgment used to assign severity scores. Generally, professional judgment on scoring may be guided by five considerations.

A. Has the practitioner or infrastructure owner experienced similar events in the past?

What where the outcomes? Is cost information available? Does any member of the team or participant at the workshop have direct, hands-on,

experience with the event being considered?

Often individuals will have knowledge of practices or events that has not been shared with the rest of the organization or documented in any formal way. This information can dramatically affect a severity score result.

B. Does the infrastructure owner have initiatives or programs in place that address the

issue?

Do current programs or initiatives touch on matters that can affect the severity of an event, even tangentially?




Page 88 of 166


2011

Will these programs or initiatives be in place throughout the time horizon of the assessment?

The practitioner should consider the impact of these initiatives on the severity scores. A program that is not initially believed to be relevant to the assessment may touch on matters that affect the serviceability of an infrastructure component. For example, a routine upgrade program may result in replacing infrastructure components such that they are more resilient to the weather event being considered.

C. Have other organizations experienced the event?

What information is available from the literature, conferences or personal

communication that can inform a severity score? Can the practitioner obtain cost, serviceability, reliability and other information

regarding hands-on experience from these organizations? Is the comparison relevant to this assessment?

The practitioner should evaluate how similar events have affected other organizations. This will inform their severity considerations for the current assessment. In doing this, the practitioner must ensure that there is good correlation between this assessment and the event being considered. The age of the infrastructures, the owners’ management systems and budgetary processes, size of the community, etc. can all affect the severity of an event. Nonetheless, evaluating the experience of third parties can often reveal sensitivities that should be considered in the current assessment.

D. Does design information accurately reflect the installed infrastructure?

Have design data and drawings been amended to reflect “as-built” conditions? Have operations and maintenance practices resulting in physical changes to the

infrastructure not document in design data?

The practitioner should ensure that their design assumptions accurately reflect the installed infrastructure. Often, during construction, operation or maintenance physical changes are made that are not fully documented. These changes could potentially result in the infrastructure being more, or less, resilient to the weather event than may be indicated by the design assumptions.

E. Will local jurisdictional considerations impact the outcome of an interaction?

Will local codes, standards and/or regulations change during the time horizon of the




Page 89 of 166


2011

assessment such that the infrastructure will become more, or less, resilient to the weather event?

Will these changes drive the owner to make physical changes to the infrastructure? Will these changes affect budgetary processes in a way that reduces routine

maintenance activities, increases service demands or delays anticipated infrastructure upgrade, replacement or expansion activities?

Vulnerability assessments make assumptions regarding the regulatory and management setting within which the infrastructure operates. If these conditions materially change over the time horizon of the assessment, severity scores can also change. For example, if anticipated maintenance activities are reduced, the severity of a weather event could be higher.

10.5 Implications of Differing Risk Assessment Approaches As described in Section 9, the Protocol supports two different risk-scoring approaches based on the probability scoring exercise.

1. Probability that, over the time horizon of the assessment, the climate parameter will change in a way that triggers an infrastructure-specific threshold.

2. Probability of a climate event triggering an infrastructure-specific threshold in the current

environment. Probability of a climate event triggering an infrastructure-specific threshold in the future environment. Probability of change is the difference between the two.

In assigning severity scores, the practitioner must ensure that they do not fall into the trap of assuming that severity outcomes remain constant over the time horizon of the assessment. It is possible that as the infrastructure ages, operations and maintenance practices evolve, population demands change, etc. that the impact of a particular weather event may be significantly different in the baseline than in the projected future conditions. This potential shift in severity should be considered regardless of the probability scoring methodology. In either case, the practitioner should consider how present and future conditions differ not only climatically but also from a regulatory, management and user perspective. In many cases the severity score will not be affected. However, in those cases where there is an impact there can be a material impact on the overall risk profile for the infrastructure. As an added benefit, understanding how these other factors affect the severity score can inform the practitioner’s conclusions and recommendations arising from the assessment.




Page 90 of 166


2011

11 Economic Considerations Economic considerations permeate changing climate infrastructure vulnerability assessment. At the project level, the infrastructure owner must establish a scope for the project and work that scope within budgetary limitations. This may drive decisions regarding the use of regional climate modeling, which can be expensive, and the overall depth and reach of the assessment. Thus, economics may dictate a smaller, more focused, assessment. Under such constraints, it is the practitioner’s responsibility to work with the infrastructure owner to establish a scope of work that both addresses the owner’s immediate issues while maximizing the opportunity to extrapolate assessment results to other areas of interest to the infrastructure owner. That is, the practitioner must work with the owner to maximize the “bang for the buck”. During the execution of the assessment, practitioners will often identify data gaps. When this occurs, the practitioner and infrastructure owner must assess the available mechanisms for obtaining or improving the data. This can also be an expensive exercise and must be evaluated based on the economic return associated with the task. For example, the data may be necessary to fully understand a risk associated with one sub-component of the infrastructure. If this sub-component is deemed to be critical with a significant economic penalty associated with its loss, the team may decide that the costs are justifiable. That is, the cost of the potential risk significantly outweighs the cost of filling the data gap. On the other hand, the data may be desired to characterize a risk that, in the grand scheme of things, is relatively minor. In this case, the team may decide to forego the expense of additional data acquisition. That is, the cost of the potential risk is much less than the cost of filling the data gap. These examples establish economic boundary conditions. During the actual execution of an assessment, significant professional judgment and consultation with the infrastructure owner may be required. It should be noted that acquiring 100% of the data desired to support a vulnerability assessment is normally outside of the economic reach of the assessment. Missing data is common and filling the gap can be very expensive. The Protocol directs practitioners to use professional judgment to address these issues. One key element of this judgment is the economic implication of the methodologies the practitioner recommends to address the gap. Finally, the practitioner may identify recommendations to address vulnerabilities identified by the assessment. Once again, the practitioner should take economic factors into consideration. For example, one potential solution to an identified vulnerability could be replacement of the infrastructure, with major capital expenditure. Since the assessment does not normally evaluate the engineering alternatives to address vulnerabilities at any depth, the practitioner should evaluate the implications of such a recommendation, in consultation with the owner, to assess the




Page 91 of 166


2011

economic feasibility. Practitioners must not shy away from reporting identified vulnerabilities, but should take to care state their recommendations within the context of reasonable, economic constraints. In the example above, although full replacement may be ideal other, more cost effective, approaches may be available and should be considered. Ultimately, these considerations will play a role in the final acceptance of the assessment and its recommendations.

public infrastructure engineering vulnerability committee · this publication is the property of...

Documents