Modeling the behavior of interdependent infrastructure, business unit and household systems under multiple disruptions
Open access
Author
Date
2019Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
Urban systems are growing into a fabric of interdependent systems-of-systems that are increasingly demonstrating the behavior of complex systems, particularly emergence and regime shifts. System-of-systems perspectives pose a new challenge, raising the question how the high degree of within and between system interdependencies affects their behavior under a broad range of disruptions. Though many models are “single system models,” “system of system models” demands for a new modeling approach that represents the interaction of systems with different purposes, lifecycles and governance structures. The present study takes up this challenge and aims to develop a proof-of-concept of a distributed simulation model, representing sets of business/household units, the metabolism of which is linked to a set of infrastructure systems, which together are exposed to a broad range of disruptions.
In particular, this study aimed: (1) to develop an agent representing the metabolism of a socioeconomic unit such as a household or a business; (2) to develop a network model of infrastructure lifeline systems and socio-economic entities, the interdependencies of which are represented by the flows of goods and services, which are facing a broad range of disruptions; (3) to investigate how synchronization of constituent systems of a system-of-systems model should be designed to ensure proper mapping of disruptions between systems; (4) to explore the application of the system-of-systems model of infrastructures, businesses, and households in a real-world use case.
The study resulted in three major findings. First, it developed and verified a proof-of-concept of a distributed simulation model, representing the metabolism of business/household units, important lifeline infrastructure systems, and disruptions that are interacting in an adaptive way. The price mechanism represented the self-adaptive capability of the overall system, where price increases signaled increasing disruption magnitudes. Simulation experiments yielded that disruptions of the water and power network have significantly higher impact than those of the transportation network wherein flows can be reconfigured. They additionally demonstrated that the concurrent disruption of the water and the power network has the highest impact on the system-of-systems, and that the impacts to the systems exhibit emergent behavior.
Second, simulation experiments explored how different time granularities of interacting systems affect the simulation model. The simulation experiments demonstrated that the time granularity of a specific system has to be finer than the expected length of an average resilience cycle. The experiments additionally suggested that if the time granularity is too coarse, then the model does not yield propagation effects appropriately. A recommendation for the time granularity of a system-of-systems simulation of infrastructures is to select time granularity similar in magnitude to the smallest expected recovery time from a major disruption of the constituent systems.
Third, the system-of-systems simulation model was applied to a real-world use case, the Clementi area of Singapore. Because the availability of infrastructure systems data was limited, Moore neighborhood was used to represent the lifeline system based on expert judgment. Simulation experiments demonstrated that disruptions of utility provider systems and networks such as power grid and water supply are costlier than those of transportation systems or other businesses. Furthermore, the model identified geographical areas, which are especially affected by disruptions being introduced into the system, and quantified the impact in areas of Moore neighborhood.
Finally, this study identified several open questions that should be addressed in future research. These include: (1) the study of another area with different infrastructure topologies; (2) the analysis of alternative resource allocation systems; (3) the development and verification of sophisticated disruption generators that allow generation of a wider range of unexpected disruption patterns; (4) the development of novel, clear and comprehensive methods for presentation of simulation results. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000381705Publication status
publishedExternal links
Search print copy at ETH Library
Publisher
ETH ZurichSubject
infrastructure modeling; Input-output model; Complex Networks; System-of-Systems (SOS); Agent Based Modelling; infrastructure resilience; Resilience assessment; disruption modelingOrganisational unit
08058 - Singapore-ETH Centre (SEC) / Singapore-ETH Centre (SEC)03331 - Heinimann, Hans-Rudolf (emeritus) / Heinimann, Hans-Rudolf (emeritus)
More
Show all metadata
ETH Bibliography
yes
Altmetrics