Abstract: - Critical infrastructure is the basis of our modern life, and the resilience of critical infrastructure is a

serious issue for maintaining our safe and secure society. After several remarkable events like terrorists’ attack

in US in 2001 and the Great East Japan Earthquake in 2012, a shared recognition that more comprehensive

approaches for crisis management is necessary. For resilience enhancement of critical infrastructure, we

have to understand the response of critical infrastructure to a crisis. Computer simulation is a

Sensitivity studies were conducted based on the 4R framework of resilience to examine the functionality of the

simulation system. Finally, it has been shown that a human-centered viewpoint is essential for assessing the

resilience of critical infrastructure.

Key-Words: - critical infrastructure, lifeline, resilience, interdependency analysis, service system, socio-

technical systems, computer simulation

1 Introduction

Critical infrastructure, which consists of a multiple

attack in US on September 11

2013, and Japanese government is now promoting

the National Resilience Policy for enhancing

national resilience of Japan [1][2].

infrastructure, we have to understand the response

of critical infrastructure to a crisis. Computer

simulation is a promising approach for this purpose,

but how to model interdependencies that exist

various threat scenarios considering their

interdependencies [3]. We expect the project will

contribute to the crisis management policy of Japan,

presenting options for the risk governance strategy

referring to the outcomes of technological analysis.

This paper, however, will focus on technological

aspects of the project and present the modeling

architecture of simulation as well as preliminary

results.

network in the metropolitan area of Tokyo,

considering interdependencies and developing a

system for simulating its complex behaviors under

various threat scenarios. The system consists of

detailed models for separate sectors, and the

integrated model including all of them.

2.1 Integrated Model

In the conventional approach of interdependency

the people are services like commodity supply,

medical service, administration, finance service, and

so on, which are provided relying on these facilities.

Operations of lifelines are also services. Since the

value of these services depends on civic life, the

resilience of service systems is further

interdependent with civic life.

using single point approximation for the power grids

of the surrounding areas.

as an earthquake is evaluated by probabilistic

planning by minimizing the total cost, which is the

sum of the facility cost and the expected variable

cost for various risk scenarios [9]. This method

enables us to obtain the best energy mix in the

metropolitan area of Tokyo considering the

shutdown risk of the power plants concentrated

Referring to the annual probability of earthquake

that directly hits Tokyo area, preliminary

assessment was done for four scenarios of 0%, 1%,

3%, and 5% risk. Fig. 4 show an example of

optimized energy mix in a day after a seismic

disaster.

were obtained. As the earthquake probability

increases, (1) new construction sites of combined

robustness of the power grid in this area from an

economic viewpoint. If the value of earthquake risk

improvement contributed to cutting 98% of the

computing time for travel search.

developed with virtual road network models was

increases, but this performance does not hold for the

model of Tokyo. Close investigation of this

performance revealed that a large difference in the

road density between different segmentation zones

resulted in failure in balancing computational loads

among many processors and then it caused the poor

performance in parallel computation. Now we are

implementing a new scheme of zone segmentation

that reflects difference in road density.

2.4 Water Supply Network Model

Since water supply tubes are installed along the

primary roads, the water supply network of the

metropolitan area of Tokyo was constructed from

the road network model for MATES. In addition,

the parts of network where seismic reinforcement

each critical facility and the nearest water supply

point is seismically resistant. Fig. 6 shows the water

supply network model of Tokyo with the locations

of critical facilities.

great earthquake that directly hits Tokyo was

The call loss probability evaluated by this

simulation under an overloading condition of

communication demands agreed well with the value

evaluated theoretically by Erlang-B formula as

shown in Fig. 7. In the next stage, each connection

in the physical ring was destroyed and the call loss

probability was evaluated for a surge of ten times

communication demands after crisis. Fig. 8 shows

only lifeline systems but also service and civic life

systems as subsystems. We can thereby simulate the

response of critical infrastructure more realistically

under a socio-technical context. Test simulation

using a virtual structure of the lifeline networks

demonstrated that considering service and civic life

is necessary to evaluate the resilience of critical

infrastructure from a human-centered viewpoint.

approach can provide technical insights useful for

decision makers in the emergence response policy.

Acknowledgement:

This research was supported by R&D Program:

Science of Science, Technology and Innovation

Policy, RISTEX, JST.

[1] http://www.oecd.org/governance/3rdoecdhighle

[5] K. Gopalakrishnan, S. Peeta, Sustainable and

[6] A. V. Gheorghe, M. Schläpfer, Critical Infra-

[7] D. Krackhardt, K. M. Carley, A PCANS Model