Finite element-based probabilistic physics-of-failure analysis to estimate pipe failure rates for risk assessment of nuclear power plants

Cheng, Wen-Chi

Permalink

https://hdl.handle.net/2142/120481 Copy

Description

Title

Finite element-based probabilistic physics-of-failure analysis to estimate pipe failure rates for risk assessment of nuclear power plants

Author(s)

Cheng, Wen-Chi

Issue Date

2023-01-10

Director of Research (if dissertation) or Advisor (if thesis)

Mohaghegh, Zahra

Doctoral Committee Chair(s)

Mohaghegh, Zahra

Committee Member(s)

• Elbanna, Ahmad
• Reihani, Seyed
• Sakurahara, Tatsuya
• Sankaran, Mohan

Department of Study

Nuclear, Plasma, & Rad Engr

Discipline

Nuclear, Plasma, Radiolgc Engr

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• probabilistic physics-of-failure
• finite element
• failure rate
• piping
• thermal fatigue
• probabilistic validation

Abstract

Probabilistic failure metrics of Reactor Coolant Pressure Boundary (RCPB) components are the key inputs to the Probabilistic Safety/Risk Assessment (PSA/PRA) and risk management of Nuclear Power Plants (NPPs. The estimation of these probabilistic failure metrics is challenging, however, especially when operational experience data directly relevant to as-built, as-operated plant conditions are limited or unavailable; for example, for advanced reactors or when maintenance policy is changed, and historical data is not applicable. This research makes three key contributions toward estimating the probabilistic failure metrics of RCPB piping components in NPPs: 1. Reviewed and categorized the existing studies on the estimation of probabilistic failure metrics for RCPB piping and steam generator tubes in NPPs. The literature review and categorization scheme are utilized to justify the selection of methods and the value of methodological advancements in this thesis. 2. Developed a Finite Element (FE)-based Probabilistic-Physics-of-Failure (PPoF) analysis for estimating pipe failure rates by (i) incorporating Finite Element (FE) analysis into Physics-of-Failure (PoF) analysis to enhance spatiotemporal resolution in loading conditions, (ii) developing an uncertainty quantification procedure to make the FE-based PoF model probabilistic or to create a FE-based Probabilistic PoF (PPoF) analysis. The probabilistic failure metrics are then generated by integrating the outputs of FE-based PPoF analysis and maintenance performance using a renewal process model. This research is the first of its kind to incorporate FE analysis into the PPoF analysis to create an integrated framework for estimating probabilistic failure metrics of NPP piping components. Through the incorporation of FE Analysis, the resolution of the PPoF analysis is enhanced as spatiotemporal conditions such as stress and temperature can be considered explicitly instead of relying on simplified assumptions or analytical models with reduced spatiotemporal dimensions. FE-based PPoF analysis is very valuable when operational data do not exist (e.g., new designs). When having limited data, the decision to use an FE-based vs. lower resolution PPoF model (e.g., correlation-based model) depends on several factors such as (i) the acceptable degree of epistemic uncertainty in the risk-informed analysis of the application of interest (ii) the available data at diverse phases of failure phenomena, (iii) whether risk management and the creation of prevention strategies significantly benefit from analyzing the spatiotemporal effects of underlying factors on pipe failures, and (iv) the available computational resources. In this thesis, a case study is conducted for an excess letdown elbow piping in the chemical and volume control system of a Pressurized Water Reactor (PWR) under thermal fatigue degradation mechanisms, which is a dominant failure mechanism for passive components, such as RCPB components. 3. Evaluated the acceptability (or validity) of the Finite Element (FE)-based PPoF model. Because common empirical validations become challenging when validation data are limited, this research uses Probabilistic Validation (PV) as a methodological base for the evaluation of the acceptability of the FE-based PPoF model. The PV methodology enhances the scientific usage of epistemic uncertainty and risk-informed acceptability criteria to support the validity evaluation for simulation predictions. The current PV methodology, however, does not have the capability of handling a situation where there is a correlation among the uncertain input parameters. This research advances the computational algorithm for the PV methodology to account for the correlated physical input parameters because the results of this research show that some of the inputs to FE analysis are correlated, and they have significant impacts on the probabilistic failure metrics estimation. A case study is used to illustrate the extended PV methodology.

Graduation Semester

2023-05

Type of Resource

Thesis

Copyright and License Information

Copyright 2023 Wen-Chi Cheng

Owning Collections