PSAM 16 - Session F04 Overview

PSAM 16 Conference Session F04 Overview

Session Chair: Zoltan Kovacs (kovacs@relko.sk)

Paper 1 FL13

Lead Author: Floris Goerlandt Co-author(s): Lauryne Rodrigues, Lauryne.Rodrigues@dal.ca Luana Souza Almeida, Luana.Almeida@dal.ca Ronald Pelot, Ronald.Pelot@dal.ca

Assessing risk to marine transportation assets and multi-modal community supply in a Cascadia M9.0 megathrust earthquake and tsunami event
Natural disasters such as earthquakes can severely impact multi-modal logistics networks. On the Canadian West Coast, a Cascadia Subduction Zone earthquake could lead to severe damage to critical infrastructures on a large regional scale, both due to the direct impacts of ground shaking and those caused by a subsequent tsunami. In the immediate disaster response phase, affected communities would furthermore require emergency supplies such as fuel, food, and medicine. An improved understanding of the vulnerabilities and impacts of a Cascadia earthquake disaster scenario, in particular to marine transportation assets, multi-modal logistic networks, and coastal communities, is important to improve disaster preparedness risk management. This article presents an integrated set of modeling approaches to analyse these risks, providing a high-level overview of the models’ objectives, rationale, and outputs. Subsequently, selected results are presented for a plausible Cascadia M9.0 megathrust earthquake and tsunami, focusing on the impacts on marine assets and the maritime dimension of the multi-modal logistics chain near Vancouver Island, British Columbia. Finally, a discussion provides insights into possible future research directions.

Assessing risk to marine transportation assets and multi-modal community supply in a Cascadia M9.0 megathrust earthquake and tsunami event

Natural disasters such as earthquakes can severely impact multi-modal logistics networks. On the Canadian West Coast, a Cascadia Subduction Zone earthquake could lead to severe damage to critical infrastructures on a large regional scale, both due to the direct impacts of ground shaking and those caused by a subsequent tsunami. In the immediate disaster response phase, affected communities would furthermore require emergency supplies such as fuel, food, and medicine. An improved understanding of the vulnerabilities and impacts of a Cascadia earthquake disaster scenario, in particular to marine transportation assets, multi-modal logistic networks, and coastal communities, is important to improve disaster preparedness risk management. This article presents an integrated set of modeling approaches to analyse these risks, providing a high-level overview of the models’ objectives, rationale, and outputs. Subsequently, selected results are presented for a plausible Cascadia M9.0 megathrust earthquake and tsunami, focusing on the impacts on marine assets and the maritime dimension of the multi-modal logistics chain near Vancouver Island, British Columbia. Finally, a discussion provides insights into possible future research directions.

Paper FL13 | Download the paper file. | Download the presentation pdf file.

Name: Floris Goerlandt (floris.goerlandt@dal.ca)

Bio: Dr. Floris Goerlandt is Assistant Professor in the Industrial Engineering Department at Dalhousie University. He currently also holds the Canada Research Chair (Tier 2) in Risk Management and Resource Optimization for Marine Industries. He co-lead the Maritime Risk and Safety (MARS) Group, working together with Dr. Ron Pelot. His research interests cover applied risk management for marine industries, authorities, and associated stakeholders. Current projects focus on analysis and modeling of risks to coastal communities due to shipping disruptions caused by major natural hazards, oil spill risk assessment and management, and risks of Arctic shipping. He also has a strong interest and commitment to foundational issues in risk research and safety science. He has (co-)authored over 100 peer-reviewed journal and conference articles, and several other publications. He serves on the Editorial Board of Safety Science, Transportation Safety and Environment, and Multimodal Transportation.

Country: CAN
Company: Dalhousie University
Job Title: Assistant Professor, Canada Research Chair

Paper 2 JH103

Lead Author: Jin Ho Lee Co-author(s): Hieu Van Nguyen, nvh10chelsea@gmail.com, Jung Han Kim, jhankim@pusan.ac.kr

Probabilistic Site Response Analysis Considering Variability of Soil Properties
Because seismic waves are influenced by variability at local soil sites, their effects on site responses are studied by means of a probabilistic site response analysis based on the random vibration theory. Monte Carlo simulations are employed to take into consideration the effects of variability of the layer thickness, low-strain shear-wave velocity, and nonlinear dependence of the shear modulus and hysteretic damping on shear strain at soil sites. The probabilistic approach is applied to generic soil sites to evaluate their soil-amplification functions. Seismic hazard curves and uniform hazard response spectra (UHRS) can be obtained using the amplification functions and the hazard curves of outcropping bedrock motions. Then, the seismic risk for structures at the sites is calculated to produce ground motion response spectra (GMRS) for a seismic design. It can be observed from the simulations that the variability in the layer thickness and the low-strain shear-wave velocity influences the soil-amplification functions and the resulting UHRS and GMRS values for the soil sites significantly.

Probabilistic Site Response Analysis Considering Variability of Soil Properties

Because seismic waves are influenced by variability at local soil sites, their effects on site responses are studied by means of a probabilistic site response analysis based on the random vibration theory. Monte Carlo simulations are employed to take into consideration the effects of variability of the layer thickness, low-strain shear-wave velocity, and nonlinear dependence of the shear modulus and hysteretic damping on shear strain at soil sites. The probabilistic approach is applied to generic soil sites to evaluate their soil-amplification functions. Seismic hazard curves and uniform hazard response spectra (UHRS) can be obtained using the amplification functions and the hazard curves of outcropping bedrock motions. Then, the seismic risk for structures at the sites is calculated to produce ground motion response spectra (GMRS) for a seismic design. It can be observed from the simulations that the variability in the layer thickness and the low-strain shear-wave velocity influences the soil-amplification functions and the resulting UHRS and GMRS values for the soil sites significantly.

Paper JH103 | Download the paper file. | Download the presentation pdf file.

Name: Jin Ho Lee (jholee0218@pknu.ac.kr)

Bio:

Country: KOR
Company: Pukyong National University
Job Title: Associate Professor

Paper 3 KO263

Lead Author: Zoltan Kovacs Co-author(s): Tibor Zold, tibor.zold@seas.sk

Seismic re-evaluation of the Unit 1&2 of the Mochovce NPP
The Slovak Nuclear Regulatory Authority (UJD SR) requires the licensee to keep the plant seismic capacity to a level generally accepted by the international community. The re-evaluation of the seismic capacity of an existing plant is generally due to the following reasons: • evidence of a seismic hazard at the site that is greater than the design basis earthquake, arising from new or additional data and • regulatory requirements stricter than those valid at the time of design and construction and take into account the state of knowledge and the actual condition of the plant. The original design basis earthquake of the plant with VVER440 type reactors was 0.06 g. The first project of seismic re-evaluation in 1997 increased the RLE (Review Level Earthquake) to 0.10 g. Significant number of plant upgrading measures were implemented. As a consequence of a new hazard evaluation carried out at the site, a new project was launched by the licensee, which further upgrades the seismic capacity of the plant to the RLE = 0.15 g (SL-2 value). The seismic re-evaluation is being performed on the basis of the IAEA guideline “Evaluation of Seismic Safety for Existing Nuclear Installation (NS-G-2.12)”. The paper summarizes the basic assumptions and methods being applied for the re-evaluation, which follows the generic methodology of the Seismic Margin Assessment (SMA) combined with a Seismic Probabilistic Safety Assessment (SPSA).

Seismic re-evaluation of the Unit 1&2 of the Mochovce NPP

The Slovak Nuclear Regulatory Authority (UJD SR) requires the licensee to keep the plant seismic capacity to a level generally accepted by the international community. The re-evaluation of the seismic capacity of an existing plant is generally due to the following reasons: • evidence of a seismic hazard at the site that is greater than the design basis earthquake, arising from new or additional data and • regulatory requirements stricter than those valid at the time of design and construction and take into account the state of knowledge and the actual condition of the plant. The original design basis earthquake of the plant with VVER440 type reactors was 0.06 g. The first project of seismic re-evaluation in 1997 increased the RLE (Review Level Earthquake) to 0.10 g. Significant number of plant upgrading measures were implemented. As a consequence of a new hazard evaluation carried out at the site, a new project was launched by the licensee, which further upgrades the seismic capacity of the plant to the RLE = 0.15 g (SL-2 value). The seismic re-evaluation is being performed on the basis of the IAEA guideline “Evaluation of Seismic Safety for Existing Nuclear Installation (NS-G-2.12)”. The paper summarizes the basic assumptions and methods being applied for the re-evaluation, which follows the generic methodology of the Seismic Margin Assessment (SMA) combined with a Seismic Probabilistic Safety Assessment (SPSA).

Paper KO263 | Download the paper file. | Download the presentation PowerPoint file.

Name: Zoltan Kovacs (kovacs@relko.sk)

Bio: Mr. Zoltan Kovacs, PhD is Director of RELKO Ltd, Engineering and Consulting Company. He is actively involved in the area of safety assessment of NPPs with more than 40 years of experience. He is the project manager of the level 1 and level 2 living PSA and PSA application projects being performed for the Slovak NPPs in operation and under construction. He has been involved also in several projects initiated to evaluate and improve safety of NPPs in Eastern and Central Europe. He has also a close co-operation with IAEA in Vienna. He has lectured the PSA training courses, organised by IAEA. Such training courses were organized in Argonne National Laboratory (Chicago, IL, USA), Abdus Salaam International Centre for Theoretical Physics (Trieste, Italy), Malaysia, Vietnam and Jordan to support experts from developing countries. He is author of numerous publications as well. He represents Slovakia in the Working Group for Risk Assessment (WGRISK) in OECD, Paris.

Country: SVK
Company: RELKO Ltd, Engineering & Consulting Company
Job Title: Director, Senior Scientist

Paper 4 SO123

Lead Author: Shota Soga Co-author(s): Eishiro Higo higo3791@criepi.denken.or.jp Hiromichi Miura h-miura@criepi.denken.or.jp

Theoretical comparison of models for a seismically induced joint failure probability
An earthquake simultaneously challenges multiple structures, systems, components of nuclear power plants. Seismic probabilistic risk assessment evaluates this phenomenon with a failure condition that a component fails when a seismic response exceeds a component capacity. In literature, there are several models for a seismically induced joint failure probability: a model used in the seismic safety margins research program (Model 1), a model in the SECOM2 (Model 2), and the Reed-McCann procedure (Model 3). We also discuss a model that applies the separation of independent and common variables method to response and capacity (Model 4). In Model 4, common variables among more than two components are explicitly considered. These four methods are analytically compared to clarify their relation. First, it is shown that the first two methods are equivalent by showing their derivations. Next, Model 4 is shown as a limited case of Model 1 by showing that Model 4 results in a multivariate normal distribution with nonnegative correlation coefficients. Finally, Model 3 is shown as a limited case of Model 4 by deriving the failure criterion used in Model 3 from Model 4 by neglecting common variables among more than two components. Thus, we summarize the relation: Model 1 = Model 2 ⊇ Model 4 ⊇ Model 3. Therefore, we recommend Model 1 for a joint failure probability because of its computational efficiency and better applicability.

Theoretical comparison of models for a seismically induced joint failure probability

An earthquake simultaneously challenges multiple structures, systems, components of nuclear power plants. Seismic probabilistic risk assessment evaluates this phenomenon with a failure condition that a component fails when a seismic response exceeds a component capacity. In literature, there are several models for a seismically induced joint failure probability: a model used in the seismic safety margins research program (Model 1), a model in the SECOM2 (Model 2), and the Reed-McCann procedure (Model 3). We also discuss a model that applies the separation of independent and common variables method to response and capacity (Model 4). In Model 4, common variables among more than two components are explicitly considered. These four methods are analytically compared to clarify their relation. First, it is shown that the first two methods are equivalent by showing their derivations. Next, Model 4 is shown as a limited case of Model 1 by showing that Model 4 results in a multivariate normal distribution with nonnegative correlation coefficients. Finally, Model 3 is shown as a limited case of Model 4 by deriving the failure criterion used in Model 3 from Model 4 by neglecting common variables among more than two components. Thus, we summarize the relation: Model 1 = Model 2 ⊇ Model 4 ⊇ Model 3. Therefore, we recommend Model 1 for a joint failure probability because of its computational efficiency and better applicability.

Paper SO123 | Download the paper file. |

Name: Shota Soga (soga@criepi.denken.or.jp)

Bio: Mr. Shota Soga is a Research Scientist in the Nuclear Risk Research Center (NRRC), the Central Research Institute of Electric Power Industry (CRIEPI), Japan. He is working on studies for multi-unit seismic probabilistic risk assessment (PRA) and advanced modeling aproach for PRA. His most recent focus is on seismically inducedd joint failure probability of multiple components and its uncertainty propagation. His other research interests include advanced numerical algorithm for computing PRA model and advanced common-cause failure analysis. He received the B.S. degree in nuclear engineering from University of Tennessee, Knoxville and the M.S degree in nuclear engineering from North Carolina State University, in 2011 and 2012, respectively.

Country: JPN
Company: Central Research Institute of Electric Power Industry
Job Title: Research Scientist