PSAM 16 - Session W14 Overview

PSAM 16 Conference Session W14 Overview

Session Chair: Antonios Zoulis (antonios.zoulis@nrc.gov)

Paper 1 SE279

Lead Author: Sergey Galushin Co-author(s): Anders Riber Marklund anders.ribermarklund@vysusgroup.com Anders Olsson anders.olsson@vysusgroup.com Ola Bäckström ola.backstrom@lr.org Dmitry Grishchenko dmitrygr@kth.se Pavel Kudinov pkudinov@kth.se

Presenter of this paper: Anders Olsson (anders.olsson@vysusgroup.com)

Treatment of Phenomenological Uncertainties in Level 2 PSA for Nordic BWR Using Risk Oriented Accident Analysis Methodology
A comprehensive and robust assessment of phenomenological uncertainties is a challenge for the current real-life PSA L2 applications, since such uncertainty is majorly driven by physical phenomena and timing of events. Typically, the static PSA models are built on a pre-determined set of scenario parameters to describe the accident progression sequence and use a limited number of simulations in the underlying deterministic analysis to evaluate the consequences. The Risk Oriented Accident Analysis Methodology (ROAAM+) has been developed to enable consistent and comprehensive treatment of both epistemic and aleatory sources of uncertainty in risk quantification. The framework is comprised of a set of deterministic models that simulate different stages of the accident progression, and a probabilistic platform that performs quantification of the uncertainty in conditional containment failure probability. This information is used for enhanced modeling in the PSA-L2 for improved definition of sequences, where information from the ROAAM is used to refine PSA model resolution regarding risk important accident scenario parameters, that can be modelled within the PSA. This work presents an example of application of the dynamic approach in a large-scale PSA model and demonstrate the integration of the ROAAM+ results in the PSA model.

Treatment of Phenomenological Uncertainties in Level 2 PSA for Nordic BWR Using Risk Oriented Accident Analysis Methodology

A comprehensive and robust assessment of phenomenological uncertainties is a challenge for the current real-life PSA L2 applications, since such uncertainty is majorly driven by physical phenomena and timing of events. Typically, the static PSA models are built on a pre-determined set of scenario parameters to describe the accident progression sequence and use a limited number of simulations in the underlying deterministic analysis to evaluate the consequences. The Risk Oriented Accident Analysis Methodology (ROAAM+) has been developed to enable consistent and comprehensive treatment of both epistemic and aleatory sources of uncertainty in risk quantification. The framework is comprised of a set of deterministic models that simulate different stages of the accident progression, and a probabilistic platform that performs quantification of the uncertainty in conditional containment failure probability. This information is used for enhanced modeling in the PSA-L2 for improved definition of sequences, where information from the ROAAM is used to refine PSA model resolution regarding risk important accident scenario parameters, that can be modelled within the PSA. This work presents an example of application of the dynamic approach in a large-scale PSA model and demonstrate the integration of the ROAAM+ results in the PSA model.

Paper SE279 | Download the paper file. | Download the presentation pdf file.

Name: Sergey Galushin (sergey.galushin@vysusgroup.com)

Bio: PhD in Nuclear Engineering (Royal Institute of Technology, Sweden). Main areas of interest: Severe accident, severe accident phenomena, PSA L1 and L2, sensitivity and uncertainty analysis.

Country: SWE
Company: Vysus Sweden AB
Job Title: Senior consultant

Presenter Name: Anders Olsson (anders.olsson@vysusgroup.com)

Bio: Anders Olsson holds a master’s degree in Mechanical Engineering and has since 1995 been working in the nuclear industry. He started at ABB Atom where he mainly performed various thermal hydraulic analysis and worked with structural verification. Since 1999 his main focus has been Probabilistic Risk Assessment where he now has extensive experience in PSA Level 1 and 2 for all operating modes including Human Reliability Analysis. He also holds a position as Vice President in Vysus Group with responsibility for the operation of the nuclear consultancy in Sweden.

Country: Sweden
Company: Vysus Group
Job Title: Vice President Nuclear

Paper 2 JL66

Lead Author: James Lin

Thermal-Hydraulic Analyses in Spent Fuel Pool PSA
To support the human reliability analysis (HRA) and the development of event sequence models in the Spent Fuel Pool (SFP) Probabilistic Safety Assessment (PSA), thermal-hydraulic analyses of selected, representative event scenarios must be performed. To evaluate such model parameters as the break flow rate of a SFP loss of inventory initiating event, the time available for specific operator actions or the number of equipment trains required to perform a safety-related function, these analyses can be performed based only on first principle energy and mass balance considerations, which are adequate to determine the broad event characteristics required to perform the HRA and develop the event sequence models with the needed accuracy. The results are considered reasonable approximations intended to reveal overall SFP response behavior and insights. This paper will describe in detail the plant and SFP design input information used in the analyses, the derivation of the SFP heat load and the various heat load cases considered, the estimate of the break flow rate of the loss of SFP inventory initiating event, the analysis of the thermal-hydraulic time windows for the loss of SFP cooling events, and the analysis of the thermal-hydraulic time windows for the loss of SFP inventory events. For the analysis of the thermal-hydraulic time windows for the loss of SFP cooling events, there are two time periods of event progression; i.e., SFP temperature increase to boiling and subsequent SFP level decrease after boiling. Following a complete loss of the SFP cooling event, the decay heat generated from the spent fuel assemblies stored in the SFP will initially cause the SFP water temperature to increase. Once the SFP temperature has reached the boiling temperature, the SFP water level will start decreasing due to boiloff. A number of time windows are determined in the analysis to support the HRA and the development and understanding of the event sequence models. They may include time to SFP high temperature alarm, time to SFP boiling temperature, time to SFP low level alarm, time to SFP cooling system suction strainer level, time to minimum SFP level, time to 3’ above top of active spent fuel, time to top of active spent fuel level, etc. Rupture of a pipe section in the SFP cooling system with no anti-siphoning device could potentially lead to a large SFP coolant loss scenario due to the siphoning effect. If the piping component friction losses are conservatively neglected, the maximum, initial break flow rate that could possibly attain due to the siphoning effect could be substantially greater than the realistic coolant loss flow rate with the consideration of the piping component friction losses. Following a loss of SFP inventory event, the SFP water level will continue to decrease. When the SFP water level drops to below the SFP cooling system suction strainer, the SFP heat removal function will be lost, at which point the SFP water temperature will start increasing. Depending on the size of the SFP leak, the order of the timing for the subsequent events, including boiling, reaching top of active spent fuel, and termination of leak, may be different. For a small leak, the rate of SFP level decrease is smaller and as such there is more time for water temperature increase before the water level dropping to the top of active spent fuel and termination of the leak when the SFP level drops to below the opening of the SFP cooling return line distribution header. For a large leak, the SFP water level could drop rapidly to the top of active spent fuel or leak termination level before the SFP water boils.

Thermal-Hydraulic Analyses in Spent Fuel Pool PSA

To support the human reliability analysis (HRA) and the development of event sequence models in the Spent Fuel Pool (SFP) Probabilistic Safety Assessment (PSA), thermal-hydraulic analyses of selected, representative event scenarios must be performed. To evaluate such model parameters as the break flow rate of a SFP loss of inventory initiating event, the time available for specific operator actions or the number of equipment trains required to perform a safety-related function, these analyses can be performed based only on first principle energy and mass balance considerations, which are adequate to determine the broad event characteristics required to perform the HRA and develop the event sequence models with the needed accuracy. The results are considered reasonable approximations intended to reveal overall SFP response behavior and insights. This paper will describe in detail the plant and SFP design input information used in the analyses, the derivation of the SFP heat load and the various heat load cases considered, the estimate of the break flow rate of the loss of SFP inventory initiating event, the analysis of the thermal-hydraulic time windows for the loss of SFP cooling events, and the analysis of the thermal-hydraulic time windows for the loss of SFP inventory events. For the analysis of the thermal-hydraulic time windows for the loss of SFP cooling events, there are two time periods of event progression; i.e., SFP temperature increase to boiling and subsequent SFP level decrease after boiling. Following a complete loss of the SFP cooling event, the decay heat generated from the spent fuel assemblies stored in the SFP will initially cause the SFP water temperature to increase. Once the SFP temperature has reached the boiling temperature, the SFP water level will start decreasing due to boiloff. A number of time windows are determined in the analysis to support the HRA and the development and understanding of the event sequence models. They may include time to SFP high temperature alarm, time to SFP boiling temperature, time to SFP low level alarm, time to SFP cooling system suction strainer level, time to minimum SFP level, time to 3’ above top of active spent fuel, time to top of active spent fuel level, etc. Rupture of a pipe section in the SFP cooling system with no anti-siphoning device could potentially lead to a large SFP coolant loss scenario due to the siphoning effect. If the piping component friction losses are conservatively neglected, the maximum, initial break flow rate that could possibly attain due to the siphoning effect could be substantially greater than the realistic coolant loss flow rate with the consideration of the piping component friction losses. Following a loss of SFP inventory event, the SFP water level will continue to decrease. When the SFP water level drops to below the SFP cooling system suction strainer, the SFP heat removal function will be lost, at which point the SFP water temperature will start increasing. Depending on the size of the SFP leak, the order of the timing for the subsequent events, including boiling, reaching top of active spent fuel, and termination of leak, may be different. For a small leak, the rate of SFP level decrease is smaller and as such there is more time for water temperature increase before the water level dropping to the top of active spent fuel and termination of the leak when the SFP level drops to below the opening of the SFP cooling return line distribution header. For a large leak, the SFP water level could drop rapidly to the top of active spent fuel or leak termination level before the SFP water boils.

Paper JL66 | Download the paper file. | Download the presentation PowerPoint file.

Name: James Lin (jlin@absconsulting.com)

Bio: Mr. Lin is a senior consultant in the Irvine, California office of ABSG Consulting Inc. He has worked in the area of quantitative risk assessment and risk management for 42 years. Mr. Lin received his master degree in Nuclear Science and Engineering from University of California in Los Angeles. He has authored or co-authored more than 200 technical papers and reports.

Country: USA
Company: ABSG Consulting Inc.
Job Title: Director

Paper 3 NA197

Lead Author: Nathalia Nunes Araujo Co-author(s): Maritza Rodriguez Gual, maritza.gual@labrisco.usp.br; Ulisses Alves Maciel Neto, ulisses.neto@labrisco.usp.br Marcos Coelho Maturana, marcos@labrisco.usp.br Marcelo Ramos Martins, marcelo@labrisco.usp.br

DEVELOPMENT AND PRIMARY APPLICATION OF A LEVEL 2 PSA METHODOLOGY IN A SMALL NUCLEAR PLANT
In the nuclear plant licensing process, a qualitative and quantitative analysis of the probability, progression and consequences of transients and accident conditions must be performed to estimate the risk to public health. Probabilistic Safety Assessment (PSA) is a method widely used in the nuclear industry that numerically quantifies risk and is performed at three different levels. The PSA Level 2 addresses the phenomenological and physical events that can occur during core meltdown to containment failure. The methodology of a Level 2 PSA must contain a clear definition of the steps, procedures and reviews to be carried out in accordance with the standards and guidelines recommended by the International Atomic Energy Agency. This article describes the development of a Level 2 PSA methodology for a small nuclear power plant. Deterministic modeling of the accident progression is also considered, being essential for the construction of the sequence of events and subsequent management measures. Along with the development, a primary application of the methodology is being carried out to identify improvements.

DEVELOPMENT AND PRIMARY APPLICATION OF A LEVEL 2 PSA METHODOLOGY IN A SMALL NUCLEAR PLANT

In the nuclear plant licensing process, a qualitative and quantitative analysis of the probability, progression and consequences of transients and accident conditions must be performed to estimate the risk to public health. Probabilistic Safety Assessment (PSA) is a method widely used in the nuclear industry that numerically quantifies risk and is performed at three different levels. The PSA Level 2 addresses the phenomenological and physical events that can occur during core meltdown to containment failure. The methodology of a Level 2 PSA must contain a clear definition of the steps, procedures and reviews to be carried out in accordance with the standards and guidelines recommended by the International Atomic Energy Agency. This article describes the development of a Level 2 PSA methodology for a small nuclear power plant. Deterministic modeling of the accident progression is also considered, being essential for the construction of the sequence of events and subsequent management measures. Along with the development, a primary application of the methodology is being carried out to identify improvements.

Paper NA197 | Download the paper file. | Download the presentation PowerPoint file.

Name: Nathalia Nunes Araujo (nathalia.nunes@labrisco.usp.br)

Bio:

Country: BRA
Company: LabRisco/USP/Brazil
Job Title: Researcher