International Construction Specialty Conference of the Canadian Society for Civil Engineering (ICSC) (5th : 2015)

A process for the assessment of infrastructure related risk due to natural hazards Hackl, Jürgen; Adey, Bryan T.; Heitzler, Magnus; Iosifescu-Enescu, Ionut; Hurni, Lorenz Jun 30, 2015

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

Item Metadata

Download

Media
52660-Hackl_J_et_al_ICSC15_246_A_Process_For.pdf [ 1.63MB ]
52660-Hackl_J_et_al_ICSC15_246_A_Process_For_slides.pdf [ 6.93MB ]
Metadata
JSON: 52660-1.0076348.json
JSON-LD: 52660-1.0076348-ld.json
RDF/XML (Pretty): 52660-1.0076348-rdf.xml
RDF/JSON: 52660-1.0076348-rdf.json
Turtle: 52660-1.0076348-turtle.txt
N-Triples: 52660-1.0076348-rdf-ntriples.txt
Original Record: 52660-1.0076348-source.json
Full Text
52660-1.0076348-fulltext.txt
Citation
52660-1.0076348.ris

Full Text

5th International/11th Construction Specialty Conference 5e International/11e Conférence spécialisée sur la construction    Vancouver, British Columbia June 8 to June 10, 2015 / 8 juin au 10 juin 2015   A PROCESS FOR THE ASSESSMENT OF INFRASTRUCTURE RELATED RISK DUE TO NATURAL HAZARDS Jürgen Hackl1,3, Bryan T. Adey1, Magnus Heitzler2, Ionut Iosifescu-Enescu2, Lorenz Hurni2 1 Institute for Construction Engineering and Management, Swiss Federal Institute of Technology, Zürich, Switzerland 2 Institute of Cartography and Geoinformation, Swiss Federal Institute of Technology, Zürich, Switzerland 3 hackl@ibi.baug.ethz.ch Abstract: The determination of network related risks for transport infrastructure systems, such as road or railway networks, is a challenging task. Due to such complex systems, it is generally impossible to abstract the global behavior from the analysis of single components, especially under conditions such as failures or damages. People who manage infrastructure have to handle these risks. The proposed overarching risk assessment process is constructed in a way so that computational support can be constructed in modules. This allows to couple the process with detailed sub-processes to achieve varying levels of detail in the risk assessment. The use of the overarching risk assessment process is demonstrated by using it to evaluate infrastructure related risk due to natural hazards for an example region in Switzerland. 1 INTRODUCTION Transport infrastructure is a key element for the economic growth and development and it plays a fundamental role in the modern world. Over the past decades it has become obvious that the analysis and understanding of large-scale infrastructure networks is important for research, engineering and society. The failure or damage of an infrastructure system could cause huge social disruption. It could be out of all proportion to the actual physical damage (Vespignani 2010). Thus, understanding the phenomenon of failure is critical for evaluating the systems risk and vulnerability and for designing robust infrastructure (Schneider et al. 2011). People who manage infrastructure, herein referred to as infrastructure managers, have to handle these risks. Each infrastructure manager relies on his own risk management processes. These processes are systematic, timely and structured processes that when followed will provide the infrastructure manager with a better understanding of what may go wrong with the system in which the infrastructure is embedded, the probability of this happening and the associated consequences. This risk assessment process is particularly challenging for managers of infrastructure networks, due to the large number of scenarios that need to be analysed in order to assess the risks appropriately, the spatial and temporal correlations between these events (MOVE 2011), and the correlation between event occurrences, or so called cascading events (Garcia-Aristizabal & Marzocchi 2011). In addition to the challenges in the physical world, the process is made even more complex because the risk assessment process requires that persons work together from many different disciplines who each 246-1 has their own discipline based approaches to risk assessment that are not always harmonious with those in other disciplines. This makes it so that independent risk assessments from different persons are not always easy to aggregate to a level that is useful for the infrastructure manager. The overarching process presented in this article is meant to be helpful to infrastructure managers who want to assess the infrastructure related risks due to natural hazards. It is to be used to help bring together people from many different disciplines so that they can provide information in a way that will be useful to an infrastructure manager. It has been specifically developed to deal with road and rail infrastructure networks but it is believed to be generally applicable to all types of infrastructure networks. The proposed overarching process is meant to fit within the risk management process of any infrastructure owner. This process is developed so that it can be coupled with detailed sub-processes to achieve varying levels of detail in risk assessment. This flexibility ensures that the overarching process is applicable for different types of infrastructure, different types of hazards, different levels of detail in the assessment, different sizes of regions, different types of regions and different levels of abstraction. It is also developed to ensure that the temporal and spatial correlation of events can be considered. The work was carried out in the scope of the European project INFRARISK, with the aim to develop reliable stress tests to establish the resilience of European road and rail network infrastructure to rare low frequency extreme events and to aid decision making in the long term regarding robust infrastructure development and protection of existing infrastructure. This article is a summary of Hackl et al. (2014). Additional information also can be found in the report Adey et al. (2014) which was submitted as a deliverable in the INFRARISK project. The work builds on that done for the Swiss Federal Roads Authority in 2005 (Adey et al. 2009; Adey et al. 2010). 2 METHODOLOGY The overarching risk assessment process is based on ISO 31000 (2009), including different principle activities: communicating and consulting, establishing the context, and identifying, analysing, evaluating, treating, monitoring and reviewing risk. Beside the basic concepts of ISO 31000, the proposed framework has been extended to allow explicit consideration of the spatial and temporal correlation between hazards as well as the modelling of the functional interdependencies between multiple objects in the infrastructure networks, including physical dependencies, cybernetic dependencies, geographical dependencies and the modelling of impacts. The process is described using generic definitions of sources, hazards, objects of the network and the network itself, which eases the application to different decision-making situations. It is constructed keeping in mind that for many decision-making situations it will be desired to have the process be computer supported, for example to model specific parts of the system. It has also been constructed keeping in mind that different decision situations will require the use of different types of models and models that will provide different levels of detail. In the following, a brief overview of the different sub-processes of the overarching risk assessment is given. 2.1 Problem Identification The first step is to identify the question to be answered. This step includes the generation of preliminary thoughts on the area to be investigated. It is only once this question is identified that a meaningful risk assessment can be conducted. It will affect the system definition, the requirements of the assessment in terms of both input, e.g. man-power, and output, e.g. the accuracy of the results or the number and types of scenarios to be investigated. It will also affect the scope of the assessment and the level of detail. 2.2 System Definition The system representation is a model of the relevant part of reality used for the evaluation and consists of all relevant realizations of stochastic processes within the investigated time period. It includes sufficiently good representations of the hazards, infrastructure, and consequences, as well as the interaction between them so that it can be reasonably certain that there is an appropriate understanding of the system and that the risks and the effectiveness of the strategies can be determined. 246-2 The system to be modelled includes all things required to assess risk, including the natural environment, e.g. amount of rain, amount of water in rivers, the physical infrastructure, e.g. the behaviour of a bridge when subjected to high water levels, and human behaviour, e.g. traffic patterns when a road bridge is no longer functioning. As it is necessary to model the system over time, it is necessary to also model the spatial and temporal correlation between events and activities within the investigated time period. This includes the consideration of assumptions, agreements as to how the system will react in specific situations, and drawing fixed system boundaries where it is clear that the things outside the considered system are not being modelled. It also includes the consideration of cascading events. 2.2.1 Boundaries By establishing spatial boundaries, the part of the natural and man-made environment to be specifically modeled is determined. In addition to the definition of the geographical space, this includes specification of where the objects are located, where the events and hazards can occur and where the consequences could take place. It is usually easy to specify the possible locations of the events, hazards and objects. It is more difficult to, however, determine how they are related, e.g. heavy rain causes a flood hazard. This becomes even more difficult when the location of possible consequences is to be specified. Consequences can be far away from the location of the events, hazards, and infrastructure, and may be outside the direct area of responsibility of the infrastructure manager (e.g. the collapse of a highway bridge on a trans-European highway network can have consequences on the free flow of goods in many countries). By establishing temporal boundaries, the time period over which risk is to be assessed is fixed, as well as how this time period is to be subdivided for analysis purposes. With respect to time, the system representation can be made either: static or dynamic. In the case of a dynamic representation, the model evolves over time whereas in the case of a static representation time is not explicitly modelled. 2.2.2 Elements It is proposed to group the system elements from initiating events to the events that are considered to be quantifiable and no further analysis is required. It is considered that the element types can be further grouped as either elements to which no value can be directly assigned or elements to which a value can be assigned. In the assessment of risk related to infrastructure due to natural hazards, one can label these further as “hazard elements” and “consequence elements”. Although the number of element types to be considered vary depending on the type of problem and the desired level of detail. Each element type is considered to correspond with events, which can be considered to have a probability of occurrence. Five basic element types, or event types, that should be regularly considered are:  Source events, or initiating events, are events, such as rainfall, tectonic plates movements, ground movement etc. The occurrence of such an event does not necessarily mean that a hazard will be triggered. Hazard events, or loading events, are events related to any earlier event that may lead to consequences. A hazard always has a source event. It may also trigger another one (e.g. earthquake triggers landslide). Most hazards evolve through space and time and interact with their environment. The time frame can vary from a few seconds (e.g. earthquake) to over a few days (e.g. flood) to several months (e.g. drought). The area that is affected can range from very local, to global. In defining the hazards to be considered it is important to define the intensities of the hazards to be considered. This should include consideration of the return period of the hazards to be used, e.g. 1/500 year flood or earthquake, and the loads to which the infrastructure will be subjected, e.g. the amount of water in the river during a flood, the magnitude of ground motions during an earthquake, the amount of displaced soil during a landslide. Infrastructure events include all the objects and the condition states of these objects to be considered, e.g. a bridge collapse is an infrastructure event. How the infrastructure networks to be modelled are subdivided into infrastructure objects depends on the specific problem and the level of detail desired in the risk assessment. For example, a 10 km road link may be modelled as one element, although it consists of 3 bridges, 4 road sections and a tunnel, or it may be subdivided to explicitly consist of all eight of these objects. If more detail is required then each object could be subdivided. For example, one of the bridges could be seen as being composed of columns, bearings, decks, etc. In the development of the 246-3 define the events to be investigated, e.g. the water height above which a flood event is considered to have occurred. At each branch in the event tree a decision is required to determine the value of the intensity measures, which allow classification of the event. The number of intensity measures used to describe the events depends on the problem being investigated and the level of detail required in the analysis. A very simple example is given in Figure 1. As can be deduced from this simple example, there is an infinite number of ways to represent reality. Due to this, particular care needs to be used in the development of an appropriate system representation. The necessary detail to be used depends on the specific problem and the level of detail desired. If events at any level, or complete ranges of the values of intensity measures are excluded, it should be explicitly explained and documented why, because in the following risk assessment, the risk coming from those hazards cannot be taken into account. 2.2.3 Relationships In order to estimate the likelihood of each subsequent event in the causal chain of events appropriate models of the relationship between them are to be developed. For example, in order to determine the amount of water coming in contact with a bridge during a flood, it is necessary to model how the water which falls as rain reaches the river, taking into consideration, for example, the amount of water that seeps into the ground or evaporates, or is held in temporary retention ponds. The amount of effort to be invested in this depends on the exact problem and the level of detail desired. For example, in some cases it may be sufficient to use one dimensional vulnerability curves based on expert opinion to estimate the amount of damage that a single object might incur during an earthquake. In other cases, it may be desirable to use multidimensional vulnerability curves based on detailed finite element models to estimate the amount of damage a large dam might incur during an earthquake. In general, extra effort should be spent to achieve more detail when it is suspected that the results will add additional clarity for decision-making. If additional clarity is not provided the extra effort is not worth it. Although specific examples are given here, the general thoughts apply to all system elements, i.e. initiating events, hazard events, infrastructure events, network use events and societal events. If possible the availability of data to be used to model the relationships should be taken into consideration in determining the level of detail to be used. 2.3 Risk Identification In the previous step emphasis is made on identifying the correct system elements to be used in the risk assessment and how to model the relationships between these. In its most extensive form the definition of these elements and relationships will provide all possible scenarios, or risks. As it is unrealistic to attempt to quantify all of these it is necessary to identify the specific scenarios that are to be part of the risk assessment. Each branch in Figure 1 is a scenario which has an associated risk. The identification of the scenarios should be done in this step without an explicit estimation of their probability of occurrence or putting a value on the consequences. The starting point for the development of this set of scenarios is all combinations of the system elements in the system representation. It is useful in the identification of scenarios to first determine for who the risk assessment is to be done, and then to: • start with the initiating events and think through how the infrastructure will be affected and then how humans will react to this, • to start with the consequences and think through how the infrastructure would have to behave to something to cause these consequences, and • to start with infrastructure behaviour and think in the other two directions. Comprehensive identification of relevant scenarios is critical, because scenarios excluded in this step will not be included in further analysis and may result in an underestimation of risk. To minimize the possibility of this happening it is important that experts in each area are involved. 246-5 2.4 Risk Analysis The analysis of risk has to do with estimating the probability of occurrence of the scenarios and the value of the consequences of the scenario if it occurs. It is only through doing this that an infrastructure manager can decide if action needs to be taken and if multiple options are available, which one is the best. It can be done using a qualitative or a quantitative approach. In both cases, however, the goal is to gain a better understanding of the probability of occurrence of a scenario and the consequence of that scenario. Risk analysis, as with risk identification, can be undertaken with varying degrees of detail, depending on the specific problem, the information, data and resources available. Analysis can be qualitative, semi-quantitative or quantitative, or a combination of these, depending on the circumstances. The certainty with which both the probabilities of occurrence of each of the scenarios and the consequences can be estimated, as well as the sensitivity of these values to the modelling assumptions, need to be given appropriate consideration in interpreting the results. Indicators of the sensitivity of these values are the divergence of opinion among experts, the availability of information, the quality of information, the level of knowledge of the persons conducting the risk analysis, and the limitation of the models used.  2.5 Risk Evaluation Risk evaluation has to do with verifying the meaning of the estimated risk to persons that may be affected, i.e. stakeholders. This is true regardless if a qualitative or a quantitative approach is used. A large part of this evaluation is the consideration of how people perceive risks and the consideration of this over- or under-valuation with respect to the analyst’s point of view used in the risk analysis step of the risk assessment. Through the risk evaluation there is the possibility to bring into the risk assessment aspects that have not been explicitly modelled in the risk analysis step. The risk evaluation step help to bring decision makers closer to finding a solution that is more acceptable to all stakeholders. One possible result of this step is that the risk analysis needs to be redone with more detailed system representations, improved models and different values. Another possible result is that it is decided that the risks are acceptable and no exploration of possible interventions are required (ISO 31000 2009). 3 EXAMPLE In this section, the use of the overarching process is demonstrated by using it to evaluate infrastructure related risk due to natural hazards for an example region. For the sake of simplicity, the example is presented in a sequential manner, although the process itself is highly iterative. The results of this example should be treated with care since only very simple physical models are used to evaluate the risk.  3.1 Problem Identification The target area is located around the city of Chur, the local capital of the easternmost Canton of Switzerland, Graubünden. Chur is located in a valley between several mountains with many watercourses draining into the main river Rhine. The addressee of this risk assessment is the city administration being interested in damage, cost and other consequences resulting from a rare flood event. It is here important to note that this example was made using public data and was not done in collaboration with, or for, the city of Chur. The results, although they are realistic and serve to demonstrate how the methodology works, are not to be used for other purposes. 3.2 System Definition 3.2.1 Boundaries The spatial boundary of the system has been selected to be that shown in Figure 2. The risk assessment is done for a flood hazard with a return period of 500 years. The occurrence of this hazard takes three 246-6 days. To compare the risk with other cities and regions, the losses resulting from this analysis are converted into an average annualized loss.  Figure 2: Overview of the area of interest. 3.2.2 Elements Source event precipitation: The model of precipitation was constructed using the precipitation data from a historical event which occurred in year 2007 and is scaled in such a way that it corresponds to a precipitation event resulting in a flood with a return period of 500 years. Hazard event flood: The model of the amount of water on each land surface area and in the rivers was developed using a set of interrelated tools. Infrastructure event hospitals: In the area of interest, only one institution is present for ambulant care, the hospital of the Canton of Graubünden. Infrastructure event road segments: Since road geometries for the target area can have lengths up to several hundred metres, these are partitioned in such a way that a spatial analysis can be undertaken on a feasible resolution. Network use events: The road network for the target area is extracted from the VECTOR25 dataset. Each road is represented by a linear geometry with assigned attributes on their type (swisstopo 2014). Societal events: Societal events are how the traffic behaves on the network when it is not fully operational. It is estimated using traffic simulations to estimate how much additional time is required to travel from anywhere in the hospital catchment area to the hospital. 3.2.3 Relationships Source-Hazard-Interaction: For reasons of simplicity and efficiency only a simple hydrological model for the runoff calculation is used. The ModClark model (Kull & Feldman 1998) is used to estimate the discharge during the precipitation event. Hazard-Infrastructure-Interaction: To estimate damage resulting from inundation, simple damage curves are used from Deckers, et al. (2010). Infrastructure-Society-Interaction: It is assumed that if infrastructure is damaged that it would be restored to the condition it had prior to being damaged. The unit values used are taken from Kutschera (2008). Infrastructure-Network-Interaction: Since this connectivity changes during the scenario due to node failure, for each time step a distinct network had to be created. Network-Society-Interaction: The quantification of consequences related to travelling across the network resulting from the failure of infrastructure network nodes was undertaken in terms of the following non-exhaustive list of examples: travel time costs (e.g. man hours of work time lost), vehicle operating costs (e.g. increase of fuel needed), accident costs (e.g. number and type of injuries/deaths), environmental costs (amount of additional nose/pollution) (Adey et al. 2012). 3.3 Risk Identification The target area has been historically prone to the mentioned natural hazards flooding and landslides. Information on past events are stored in the database "Unwetterschadens-Datenbank" (Hilker et al. 2009) In addition, two more recent projects, AquaProtect and SilvaProtect (Losey & Wehrli 2013) provide model based information on regions vulnerable to floods and landslides. For the sake of simplicity, only one scenario is considered. This scenario is comprised of the following events: Source event is rainfall, the hazard events are floods, defined as being more severe as the largest volume of water expected in the main river expected in 500 years. The infrastructure events are derived from the road sections and 246-7 3.5 Risk Evaluation In this paper, risk evaluation is not performed. If a complete risk management process is being conducted this work would need to be done in conjunction with the city administration of Chur. The results coming from the risk analysis would support this task in order to plan further analyses, safety measures or risk treatments. 4 DISCUSSION The example demonstrates that the proposed overarching risk assessment process is useful to assess infrastructure related risk due to natural hazards. Computer systems can highly accelerate its distinct steps so that the results can be delivered to infrastructure managers in a timely manner. However, in order to refine the results, the methodology needs to be applied to a greater number of scenarios. The process can be used for a wide range of different problems at different levels of detail. In addition, the changes over time and interactions between different events can be modeled as shown in the example. Although the proposed overarching risk assessment process can be used conceptually for all kinds of different problems, its usefulness depends on the quality of available models and data. Often the physical models do not take into account interaction with their environment. For example, if a bridge collapses, the cross-section of the river will be changed, too. In the presented example a relatively deterministic point of view was chosen. In order to take the numerous uncertainties into account a probabilistic approach seems more suitable, especially when dealing with natural hazards. If one associates a probability of occurrence with the occurrence of the particular precipitation then one could quantify the risk. A more sophisticated example will require the consideration of the not only the probability of occurrence of different rain patterns, but also given the rain fall patterns, the probability of different water run-off events, different levels of water in different parts of the rivers, different behavior of the infrastructure objects in the network, and different behavior of the vehicles on the network. It would also require consideration of larger periods of time, in which multiple rain events occur and perhaps even different types of source events that may result in consequences. In the expansion of the example to do this there are substantial hurdles with respect to the infinite number of scenarios possible, the uncertainties associated with many different models to be used to make approximations and the temporal changes in the probabilities of event occurrences. 5 CONCLUSIONS This article describes a generic overarching risk assessment process as well as an example of how it can be used and how it can be implemented using a GIS framework. Even in its current form it is believed that this process would be useful to infrastructure managers in the assessment of their infrastructure related risks due to natural hazards. It is applicable for different types of infrastructure, different types of hazards and different types of consequences and can take into consideration both simple and complex system representations. The overarching risk assessment process will be further improved by taking into account multiple scenarios, including multiple initiating events, multiple hazards, multiple infrastructure events, multiple network use events and multiple societal events. It will also be expanded to deal properly with the spatial and temporal consideration in the estimation of the probability of occurrence of scenarios and the establishment of the scenarios. More work is required to emphasis the human interaction in conducting the risk assessment. Acknowledgements This project has received funding from the European Union’s Seventh Programme for research, technological development and demonstration under grant agreement No 603960. 246-9 References Adey B.T., T. Herrmann, K. Tsafatinos, J. Luking, N Schindele, and R. Hajdin., 2012, Methodology and base cost models to determine the total benefits of preservation interventions on road sections in Switzerland. Structure and Infrastructure Engineering. 8(7): 639–654. Adey B.T., R. Hajdin, J. Birdsall, 2009, Methodology to determine optimal intervention strategies for structures adversely affected by latent processes, Transportation Association of Canada, Annual Conference, Vancouver, Canada, October 19-22. Adey B.T., J. Birdsall, R. Hajdin, 2010, Methodology to estimate risk related to road links, due to latent processes, 5th International Conference on Bridge Maintenance, Safety and Management, IABMAS, Philadelphia, USA, July 11-15. Adey, B.T., J. Hackl, M. Heitzler, and I. Iosifescu-Enescu, 2014, Preliminary Model, Methodology and Information Exchange. Eidgenössische Technische Hochschule Zürich, CH, INFRARISK Deliverable D4.1. Deckers P., W. Kellens, J. Reyns, W Vanneuville, and P. Maeyer, 2010, A GIS for Flood Risk Management in Flanders. In: Showalter PS, Lu Y, editors. Geospatial Techniques in Urban Hazard and Disaster Analysis. vol. 2. Springer Netherlands; 51–69. Garcia-Aristizabal, A., and W. Marzocchi, 2011, Deliverable D3.1: Review of existing procedures for multi-hazard assessment. New methodologies for multi-hazard and multi-risk assessment (MATRIX). Hackl, J., Adey, B.T., Heitzler, M., Iosifescu-Enescu, I. (2015), An overarching risk assessment process to evaluate the risks associated with infrastructure networks due to natural hazards, Special Issue of International Journal of Performability Engineering on Transport System Safety, Risk and Asset Management, 11(2). Hilker N, A. Badoux, and C. Hegg, 2009, The Swiss flood and landslide damage database 1972-2007. Nat Hazards Earth Syst Sci. 9(3): 913–925. IEEE, 1990, IEEE Standard Glossary of Software Engineering Terminology. IEEE Std 610.12-1990. ISO 31000, 2009, Risk management - Principles and guidelines. International Organization for Standardization (ISO). Kull D., and A. Feldman, 1998, Evolution of Clark’s unit graph method to spatially distributed runoff. Journal of Hydrologic Engineering, 3(1): 9–19. Kutschera G, 2008, Analyse der Unsicherheiten bei der Ermittlung der Schadenspotentiale infolge Überschwemmung. PhD Thesis, Technischen Hochschule Aachen. Losey, S, and A. Wehrli, 2013, Schutzwald in der Schweiz: Vom Projekt SilvaProtect-CH zum harmonisierten Schutzwald. Bern, CH: Bundesamt für Umwelt (BAFU). MOVE, 2011, Assessing vulnerability to natural hazards in Europe: From Principles to Practice: A manual on concept, methodology and tools. Methods for the Improvement of Vulnerability Assessment in Europe (MOVE). Schneider C. M., A. A. Moreira, J. S. Andrade, S. Havlin, and H. J. Herrmann, 2011, Mitigation of malicious attacks on networks. Proc. Natl. Acad. Sci., 108(10): 3838-3841. Swisstopo, 2014, Bundesamt für Landestopografie swisstopo, online: http://www.swisstopo.admin.ch/ (accessed on 24 May 2014) Vespignani A., 2010, The fragility of interdependency. Nature 464(15): 984–985.  246-10  5th International/11th Construction Specialty Conference 5e International/11e Conférence spécialisée sur la construction    Vancouver, British Columbia June 8 to June 10, 2015 / 8 juin au 10 juin 2015   A PROCESS FOR THE ASSESSMENT OF INFRASTRUCTURE RELATED RISK DUE TO NATURAL HAZARDS Jürgen Hackl1,3, Bryan T. Adey1, Magnus Heitzler2, Ionut Iosifescu-Enescu2, Lorenz Hurni2 1 Institute for Construction Engineering and Management, Swiss Federal Institute of Technology, Zürich, Switzerland 2 Institute of Cartography and Geoinformation, Swiss Federal Institute of Technology, Zürich, Switzerland 3 hackl@ibi.baug.ethz.ch Abstract: The determination of network related risks for transport infrastructure systems, such as road or railway networks, is a challenging task. Due to such complex systems, it is generally impossible to abstract the global behavior from the analysis of single components, especially under conditions such as failures or damages. People who manage infrastructure have to handle these risks. The proposed overarching risk assessment process is constructed in a way so that computational support can be constructed in modules. This allows to couple the process with detailed sub-processes to achieve varying levels of detail in the risk assessment. The use of the overarching risk assessment process is demonstrated by using it to evaluate infrastructure related risk due to natural hazards for an example region in Switzerland. 1 INTRODUCTION Transport infrastructure is a key element for the economic growth and development and it plays a fundamental role in the modern world. Over the past decades it has become obvious that the analysis and understanding of large-scale infrastructure networks is important for research, engineering and society. The failure or damage of an infrastructure system could cause huge social disruption. It could be out of all proportion to the actual physical damage (Vespignani 2010). Thus, understanding the phenomenon of failure is critical for evaluating the systems risk and vulnerability and for designing robust infrastructure (Schneider et al. 2011). People who manage infrastructure, herein referred to as infrastructure managers, have to handle these risks. Each infrastructure manager relies on his own risk management processes. These processes are systematic, timely and structured processes that when followed will provide the infrastructure manager with a better understanding of what may go wrong with the system in which the infrastructure is embedded, the probability of this happening and the associated consequences. This risk assessment process is particularly challenging for managers of infrastructure networks, due to the large number of scenarios that need to be analysed in order to assess the risks appropriately, the spatial and temporal correlations between these events (MOVE 2011), and the correlation between event occurrences, or so called cascading events (Garcia-Aristizabal & Marzocchi 2011). In addition to the challenges in the physical world, the process is made even more complex because the risk assessment process requires that persons work together from many different disciplines who each 246-1 has their own discipline based approaches to risk assessment that are not always harmonious with those in other disciplines. This makes it so that independent risk assessments from different persons are not always easy to aggregate to a level that is useful for the infrastructure manager. The overarching process presented in this article is meant to be helpful to infrastructure managers who want to assess the infrastructure related risks due to natural hazards. It is to be used to help bring together people from many different disciplines so that they can provide information in a way that will be useful to an infrastructure manager. It has been specifically developed to deal with road and rail infrastructure networks but it is believed to be generally applicable to all types of infrastructure networks. The proposed overarching process is meant to fit within the risk management process of any infrastructure owner. This process is developed so that it can be coupled with detailed sub-processes to achieve varying levels of detail in risk assessment. This flexibility ensures that the overarching process is applicable for different types of infrastructure, different types of hazards, different levels of detail in the assessment, different sizes of regions, different types of regions and different levels of abstraction. It is also developed to ensure that the temporal and spatial correlation of events can be considered. The work was carried out in the scope of the European project INFRARISK, with the aim to develop reliable stress tests to establish the resilience of European road and rail network infrastructure to rare low frequency extreme events and to aid decision making in the long term regarding robust infrastructure development and protection of existing infrastructure. This article is a summary of Hackl et al. (2014). Additional information also can be found in the report Adey et al. (2014) which was submitted as a deliverable in the INFRARISK project. The work builds on that done for the Swiss Federal Roads Authority in 2005 (Adey et al. 2009; Adey et al. 2010). 2 METHODOLOGY The overarching risk assessment process is based on ISO 31000 (2009), including different principle activities: communicating and consulting, establishing the context, and identifying, analysing, evaluating, treating, monitoring and reviewing risk. Beside the basic concepts of ISO 31000, the proposed framework has been extended to allow explicit consideration of the spatial and temporal correlation between hazards as well as the modelling of the functional interdependencies between multiple objects in the infrastructure networks, including physical dependencies, cybernetic dependencies, geographical dependencies and the modelling of impacts. The process is described using generic definitions of sources, hazards, objects of the network and the network itself, which eases the application to different decision-making situations. It is constructed keeping in mind that for many decision-making situations it will be desired to have the process be computer supported, for example to model specific parts of the system. It has also been constructed keeping in mind that different decision situations will require the use of different types of models and models that will provide different levels of detail. In the following, a brief overview of the different sub-processes of the overarching risk assessment is given. 2.1 Problem Identification The first step is to identify the question to be answered. This step includes the generation of preliminary thoughts on the area to be investigated. It is only once this question is identified that a meaningful risk assessment can be conducted. It will affect the system definition, the requirements of the assessment in terms of both input, e.g. man-power, and output, e.g. the accuracy of the results or the number and types of scenarios to be investigated. It will also affect the scope of the assessment and the level of detail. 2.2 System Definition The system representation is a model of the relevant part of reality used for the evaluation and consists of all relevant realizations of stochastic processes within the investigated time period. It includes sufficiently good representations of the hazards, infrastructure, and consequences, as well as the interaction between them so that it can be reasonably certain that there is an appropriate understanding of the system and that the risks and the effectiveness of the strategies can be determined. 246-2 The system to be modelled includes all things required to assess risk, including the natural environment, e.g. amount of rain, amount of water in rivers, the physical infrastructure, e.g. the behaviour of a bridge when subjected to high water levels, and human behaviour, e.g. traffic patterns when a road bridge is no longer functioning. As it is necessary to model the system over time, it is necessary to also model the spatial and temporal correlation between events and activities within the investigated time period. This includes the consideration of assumptions, agreements as to how the system will react in specific situations, and drawing fixed system boundaries where it is clear that the things outside the considered system are not being modelled. It also includes the consideration of cascading events. 2.2.1 Boundaries By establishing spatial boundaries, the part of the natural and man-made environment to be specifically modeled is determined. In addition to the definition of the geographical space, this includes specification of where the objects are located, where the events and hazards can occur and where the consequences could take place. It is usually easy to specify the possible locations of the events, hazards and objects. It is more difficult to, however, determine how they are related, e.g. heavy rain causes a flood hazard. This becomes even more difficult when the location of possible consequences is to be specified. Consequences can be far away from the location of the events, hazards, and infrastructure, and may be outside the direct area of responsibility of the infrastructure manager (e.g. the collapse of a highway bridge on a trans-European highway network can have consequences on the free flow of goods in many countries). By establishing temporal boundaries, the time period over which risk is to be assessed is fixed, as well as how this time period is to be subdivided for analysis purposes. With respect to time, the system representation can be made either: static or dynamic. In the case of a dynamic representation, the model evolves over time whereas in the case of a static representation time is not explicitly modelled. 2.2.2 Elements It is proposed to group the system elements from initiating events to the events that are considered to be quantifiable and no further analysis is required. It is considered that the element types can be further grouped as either elements to which no value can be directly assigned or elements to which a value can be assigned. In the assessment of risk related to infrastructure due to natural hazards, one can label these further as “hazard elements” and “consequence elements”. Although the number of element types to be considered vary depending on the type of problem and the desired level of detail. Each element type is considered to correspond with events, which can be considered to have a probability of occurrence. Five basic element types, or event types, that should be regularly considered are:  Source events, or initiating events, are events, such as rainfall, tectonic plates movements, ground movement etc. The occurrence of such an event does not necessarily mean that a hazard will be triggered. Hazard events, or loading events, are events related to any earlier event that may lead to consequences. A hazard always has a source event. It may also trigger another one (e.g. earthquake triggers landslide). Most hazards evolve through space and time and interact with their environment. The time frame can vary from a few seconds (e.g. earthquake) to over a few days (e.g. flood) to several months (e.g. drought). The area that is affected can range from very local, to global. In defining the hazards to be considered it is important to define the intensities of the hazards to be considered. This should include consideration of the return period of the hazards to be used, e.g. 1/500 year flood or earthquake, and the loads to which the infrastructure will be subjected, e.g. the amount of water in the river during a flood, the magnitude of ground motions during an earthquake, the amount of displaced soil during a landslide. Infrastructure events include all the objects and the condition states of these objects to be considered, e.g. a bridge collapse is an infrastructure event. How the infrastructure networks to be modelled are subdivided into infrastructure objects depends on the specific problem and the level of detail desired in the risk assessment. For example, a 10 km road link may be modelled as one element, although it consists of 3 bridges, 4 road sections and a tunnel, or it may be subdivided to explicitly consist of all eight of these objects. If more detail is required then each object could be subdivided. For example, one of the bridges could be seen as being composed of columns, bearings, decks, etc. In the development of the 246-3 define the events to be investigated, e.g. the water height above which a flood event is considered to have occurred. At each branch in the event tree a decision is required to determine the value of the intensity measures, which allow classification of the event. The number of intensity measures used to describe the events depends on the problem being investigated and the level of detail required in the analysis. A very simple example is given in Figure 1. As can be deduced from this simple example, there is an infinite number of ways to represent reality. Due to this, particular care needs to be used in the development of an appropriate system representation. The necessary detail to be used depends on the specific problem and the level of detail desired. If events at any level, or complete ranges of the values of intensity measures are excluded, it should be explicitly explained and documented why, because in the following risk assessment, the risk coming from those hazards cannot be taken into account. 2.2.3 Relationships In order to estimate the likelihood of each subsequent event in the causal chain of events appropriate models of the relationship between them are to be developed. For example, in order to determine the amount of water coming in contact with a bridge during a flood, it is necessary to model how the water which falls as rain reaches the river, taking into consideration, for example, the amount of water that seeps into the ground or evaporates, or is held in temporary retention ponds. The amount of effort to be invested in this depends on the exact problem and the level of detail desired. For example, in some cases it may be sufficient to use one dimensional vulnerability curves based on expert opinion to estimate the amount of damage that a single object might incur during an earthquake. In other cases, it may be desirable to use multidimensional vulnerability curves based on detailed finite element models to estimate the amount of damage a large dam might incur during an earthquake. In general, extra effort should be spent to achieve more detail when it is suspected that the results will add additional clarity for decision-making. If additional clarity is not provided the extra effort is not worth it. Although specific examples are given here, the general thoughts apply to all system elements, i.e. initiating events, hazard events, infrastructure events, network use events and societal events. If possible the availability of data to be used to model the relationships should be taken into consideration in determining the level of detail to be used. 2.3 Risk Identification In the previous step emphasis is made on identifying the correct system elements to be used in the risk assessment and how to model the relationships between these. In its most extensive form the definition of these elements and relationships will provide all possible scenarios, or risks. As it is unrealistic to attempt to quantify all of these it is necessary to identify the specific scenarios that are to be part of the risk assessment. Each branch in Figure 1 is a scenario which has an associated risk. The identification of the scenarios should be done in this step without an explicit estimation of their probability of occurrence or putting a value on the consequences. The starting point for the development of this set of scenarios is all combinations of the system elements in the system representation. It is useful in the identification of scenarios to first determine for who the risk assessment is to be done, and then to: • start with the initiating events and think through how the infrastructure will be affected and then how humans will react to this, • to start with the consequences and think through how the infrastructure would have to behave to something to cause these consequences, and • to start with infrastructure behaviour and think in the other two directions. Comprehensive identification of relevant scenarios is critical, because scenarios excluded in this step will not be included in further analysis and may result in an underestimation of risk. To minimize the possibility of this happening it is important that experts in each area are involved. 246-5 2.4 Risk Analysis The analysis of risk has to do with estimating the probability of occurrence of the scenarios and the value of the consequences of the scenario if it occurs. It is only through doing this that an infrastructure manager can decide if action needs to be taken and if multiple options are available, which one is the best. It can be done using a qualitative or a quantitative approach. In both cases, however, the goal is to gain a better understanding of the probability of occurrence of a scenario and the consequence of that scenario. Risk analysis, as with risk identification, can be undertaken with varying degrees of detail, depending on the specific problem, the information, data and resources available. Analysis can be qualitative, semi-quantitative or quantitative, or a combination of these, depending on the circumstances. The certainty with which both the probabilities of occurrence of each of the scenarios and the consequences can be estimated, as well as the sensitivity of these values to the modelling assumptions, need to be given appropriate consideration in interpreting the results. Indicators of the sensitivity of these values are the divergence of opinion among experts, the availability of information, the quality of information, the level of knowledge of the persons conducting the risk analysis, and the limitation of the models used.  2.5 Risk Evaluation Risk evaluation has to do with verifying the meaning of the estimated risk to persons that may be affected, i.e. stakeholders. This is true regardless if a qualitative or a quantitative approach is used. A large part of this evaluation is the consideration of how people perceive risks and the consideration of this over- or under-valuation with respect to the analyst’s point of view used in the risk analysis step of the risk assessment. Through the risk evaluation there is the possibility to bring into the risk assessment aspects that have not been explicitly modelled in the risk analysis step. The risk evaluation step help to bring decision makers closer to finding a solution that is more acceptable to all stakeholders. One possible result of this step is that the risk analysis needs to be redone with more detailed system representations, improved models and different values. Another possible result is that it is decided that the risks are acceptable and no exploration of possible interventions are required (ISO 31000 2009). 3 EXAMPLE In this section, the use of the overarching process is demonstrated by using it to evaluate infrastructure related risk due to natural hazards for an example region. For the sake of simplicity, the example is presented in a sequential manner, although the process itself is highly iterative. The results of this example should be treated with care since only very simple physical models are used to evaluate the risk.  3.1 Problem Identification The target area is located around the city of Chur, the local capital of the easternmost Canton of Switzerland, Graubünden. Chur is located in a valley between several mountains with many watercourses draining into the main river Rhine. The addressee of this risk assessment is the city administration being interested in damage, cost and other consequences resulting from a rare flood event. It is here important to note that this example was made using public data and was not done in collaboration with, or for, the city of Chur. The results, although they are realistic and serve to demonstrate how the methodology works, are not to be used for other purposes. 3.2 System Definition 3.2.1 Boundaries The spatial boundary of the system has been selected to be that shown in Figure 2. The risk assessment is done for a flood hazard with a return period of 500 years. The occurrence of this hazard takes three 246-6 days. To compare the risk with other cities and regions, the losses resulting from this analysis are converted into an average annualized loss.  Figure 2: Overview of the area of interest. 3.2.2 Elements Source event precipitation: The model of precipitation was constructed using the precipitation data from a historical event which occurred in year 2007 and is scaled in such a way that it corresponds to a precipitation event resulting in a flood with a return period of 500 years. Hazard event flood: The model of the amount of water on each land surface area and in the rivers was developed using a set of interrelated tools. Infrastructure event hospitals: In the area of interest, only one institution is present for ambulant care, the hospital of the Canton of Graubünden. Infrastructure event road segments: Since road geometries for the target area can have lengths up to several hundred metres, these are partitioned in such a way that a spatial analysis can be undertaken on a feasible resolution. Network use events: The road network for the target area is extracted from the VECTOR25 dataset. Each road is represented by a linear geometry with assigned attributes on their type (swisstopo 2014). Societal events: Societal events are how the traffic behaves on the network when it is not fully operational. It is estimated using traffic simulations to estimate how much additional time is required to travel from anywhere in the hospital catchment area to the hospital. 3.2.3 Relationships Source-Hazard-Interaction: For reasons of simplicity and efficiency only a simple hydrological model for the runoff calculation is used. The ModClark model (Kull & Feldman 1998) is used to estimate the discharge during the precipitation event. Hazard-Infrastructure-Interaction: To estimate damage resulting from inundation, simple damage curves are used from Deckers, et al. (2010). Infrastructure-Society-Interaction: It is assumed that if infrastructure is damaged that it would be restored to the condition it had prior to being damaged. The unit values used are taken from Kutschera (2008). Infrastructure-Network-Interaction: Since this connectivity changes during the scenario due to node failure, for each time step a distinct network had to be created. Network-Society-Interaction: The quantification of consequences related to travelling across the network resulting from the failure of infrastructure network nodes was undertaken in terms of the following non-exhaustive list of examples: travel time costs (e.g. man hours of work time lost), vehicle operating costs (e.g. increase of fuel needed), accident costs (e.g. number and type of injuries/deaths), environmental costs (amount of additional nose/pollution) (Adey et al. 2012). 3.3 Risk Identification The target area has been historically prone to the mentioned natural hazards flooding and landslides. Information on past events are stored in the database "Unwetterschadens-Datenbank" (Hilker et al. 2009) In addition, two more recent projects, AquaProtect and SilvaProtect (Losey & Wehrli 2013) provide model based information on regions vulnerable to floods and landslides. For the sake of simplicity, only one scenario is considered. This scenario is comprised of the following events: Source event is rainfall, the hazard events are floods, defined as being more severe as the largest volume of water expected in the main river expected in 500 years. The infrastructure events are derived from the road sections and 246-7 3.5 Risk Evaluation In this paper, risk evaluation is not performed. If a complete risk management process is being conducted this work would need to be done in conjunction with the city administration of Chur. The results coming from the risk analysis would support this task in order to plan further analyses, safety measures or risk treatments. 4 DISCUSSION The example demonstrates that the proposed overarching risk assessment process is useful to assess infrastructure related risk due to natural hazards. Computer systems can highly accelerate its distinct steps so that the results can be delivered to infrastructure managers in a timely manner. However, in order to refine the results, the methodology needs to be applied to a greater number of scenarios. The process can be used for a wide range of different problems at different levels of detail. In addition, the changes over time and interactions between different events can be modeled as shown in the example. Although the proposed overarching risk assessment process can be used conceptually for all kinds of different problems, its usefulness depends on the quality of available models and data. Often the physical models do not take into account interaction with their environment. For example, if a bridge collapses, the cross-section of the river will be changed, too. In the presented example a relatively deterministic point of view was chosen. In order to take the numerous uncertainties into account a probabilistic approach seems more suitable, especially when dealing with natural hazards. If one associates a probability of occurrence with the occurrence of the particular precipitation then one could quantify the risk. A more sophisticated example will require the consideration of the not only the probability of occurrence of different rain patterns, but also given the rain fall patterns, the probability of different water run-off events, different levels of water in different parts of the rivers, different behavior of the infrastructure objects in the network, and different behavior of the vehicles on the network. It would also require consideration of larger periods of time, in which multiple rain events occur and perhaps even different types of source events that may result in consequences. In the expansion of the example to do this there are substantial hurdles with respect to the infinite number of scenarios possible, the uncertainties associated with many different models to be used to make approximations and the temporal changes in the probabilities of event occurrences. 5 CONCLUSIONS This article describes a generic overarching risk assessment process as well as an example of how it can be used and how it can be implemented using a GIS framework. Even in its current form it is believed that this process would be useful to infrastructure managers in the assessment of their infrastructure related risks due to natural hazards. It is applicable for different types of infrastructure, different types of hazards and different types of consequences and can take into consideration both simple and complex system representations. The overarching risk assessment process will be further improved by taking into account multiple scenarios, including multiple initiating events, multiple hazards, multiple infrastructure events, multiple network use events and multiple societal events. It will also be expanded to deal properly with the spatial and temporal consideration in the estimation of the probability of occurrence of scenarios and the establishment of the scenarios. More work is required to emphasis the human interaction in conducting the risk assessment. Acknowledgements This project has received funding from the European Union’s Seventh Programme for research, technological development and demonstration under grant agreement No 603960. 246-9 References Adey B.T., T. Herrmann, K. Tsafatinos, J. Luking, N Schindele, and R. Hajdin., 2012, Methodology and base cost models to determine the total benefits of preservation interventions on road sections in Switzerland. Structure and Infrastructure Engineering. 8(7): 639–654. Adey B.T., R. Hajdin, J. Birdsall, 2009, Methodology to determine optimal intervention strategies for structures adversely affected by latent processes, Transportation Association of Canada, Annual Conference, Vancouver, Canada, October 19-22. Adey B.T., J. Birdsall, R. Hajdin, 2010, Methodology to estimate risk related to road links, due to latent processes, 5th International Conference on Bridge Maintenance, Safety and Management, IABMAS, Philadelphia, USA, July 11-15. Adey, B.T., J. Hackl, M. Heitzler, and I. Iosifescu-Enescu, 2014, Preliminary Model, Methodology and Information Exchange. Eidgenössische Technische Hochschule Zürich, CH, INFRARISK Deliverable D4.1. Deckers P., W. Kellens, J. Reyns, W Vanneuville, and P. Maeyer, 2010, A GIS for Flood Risk Management in Flanders. In: Showalter PS, Lu Y, editors. Geospatial Techniques in Urban Hazard and Disaster Analysis. vol. 2. Springer Netherlands; 51–69. Garcia-Aristizabal, A., and W. Marzocchi, 2011, Deliverable D3.1: Review of existing procedures for multi-hazard assessment. New methodologies for multi-hazard and multi-risk assessment (MATRIX). Hackl, J., Adey, B.T., Heitzler, M., Iosifescu-Enescu, I. (2015), An overarching risk assessment process to evaluate the risks associated with infrastructure networks due to natural hazards, Special Issue of International Journal of Performability Engineering on Transport System Safety, Risk and Asset Management, 11(2). Hilker N, A. Badoux, and C. Hegg, 2009, The Swiss flood and landslide damage database 1972-2007. Nat Hazards Earth Syst Sci. 9(3): 913–925. IEEE, 1990, IEEE Standard Glossary of Software Engineering Terminology. IEEE Std 610.12-1990. ISO 31000, 2009, Risk management - Principles and guidelines. International Organization for Standardization (ISO). Kull D., and A. Feldman, 1998, Evolution of Clark’s unit graph method to spatially distributed runoff. Journal of Hydrologic Engineering, 3(1): 9–19. Kutschera G, 2008, Analyse der Unsicherheiten bei der Ermittlung der Schadenspotentiale infolge Überschwemmung. PhD Thesis, Technischen Hochschule Aachen. Losey, S, and A. Wehrli, 2013, Schutzwald in der Schweiz: Vom Projekt SilvaProtect-CH zum harmonisierten Schutzwald. Bern, CH: Bundesamt für Umwelt (BAFU). MOVE, 2011, Assessing vulnerability to natural hazards in Europe: From Principles to Practice: A manual on concept, methodology and tools. Methods for the Improvement of Vulnerability Assessment in Europe (MOVE). Schneider C. M., A. A. Moreira, J. S. Andrade, S. Havlin, and H. J. Herrmann, 2011, Mitigation of malicious attacks on networks. Proc. Natl. Acad. Sci., 108(10): 3838-3841. Swisstopo, 2014, Bundesamt für Landestopografie swisstopo, online: http://www.swisstopo.admin.ch/ (accessed on 24 May 2014) Vespignani A., 2010, The fragility of interdependency. Nature 464(15): 984–985.  246-10  | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 1 J. Hackl, B. T. Adey, M. Heitzler, I. Iosifescu-Enescu and L. Hurni  ICSC 15 The CSCE International Construction Specialty Conference June 8 – 10, 2015 | Vancouver, Canada A process for the assessment of infrastructure related risk due to natural hazards | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 2 | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 3 Risk Management Process (ISO 31000:2009) Establishing the Context Risk Identification Risk Analysis Risk Evaluation Risk Treatment Communication and Consultation Monitoring and Review Risk Assessment | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 4 Source events Hazard events Infrastructure events Network use events Societal events | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 5 Intensity l Damage Cons Cons Cons Cons Cons Interruption No Interruption Intensity m Hazard Elements Consequence Elements | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 6 “Flow” of information # # # # # # # # ## # # # | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 7 Example Precipitation 500 year flood Roads, buildings Inaccessible network Travel time, hospital accesability | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 8 | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 9 | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 10 Uncertainty quantification Interdependent systems Multiple scenarios Resilience | | D-BAUG | IBI | Infrastructure Management Group  2015-07-23 Hackl et al. 11  ETH Zurich Infrastructure Management Group  Stefano-Franscini-Platz 5 IBI - HIL, Room F 25.3 8093 Zurich, Switzerland t: +41 44 63 36510 f: +41 44 63 31088 e: hackl@ibi.baug.ethz.ch 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52660.1-0076348/manifest

Comment

Related Items