Bayesian Network Modeling of Corrosion

✍ Scribed by Narasi Sridhar

• Publisher: Springer
• Year: 2024
• Tongue: English
• Leaves: 343
• Category: Library

✦ Synopsis

This book represents a compilation of experience from a slate of experts involved in developing and deploying Bayesian Networks (BN) for corrosion management. The contributors describe how probability distributions can be developed for corroding systems and BN can be applied as an ideal framework to deal with corrosion risk. Corrosion can develop suddenly and grow rapidly after a long incubation period and take many non-uniform aspects, including pitting and stress corrosion cracking, that cannot be mitigated by simply bulking up the system. They also describe how complex engineering structures and systems are influenced by many natural and engineering factors that come together in myriad ways. It provides a broad perspective to the reader on the potential of BN as an artificial intelligence tool for corrosion risk management and the challenges for implementing it.

✦ Table of Contents

Preface
Acknowledgements
Contents
Chapter 1: Introduction: Risk Assessment
Risk and Probability
Corrodible Systems
Risk Assessment Approaches
Qualitative Approaches
Semi-quantitative Methods
Quantitative Methods
Component Level Statistical Analysis Methods
Reliability and Performance Assessment Methods
Bayesian Networks
Consequence Analyses
Risk Governance, Acceptability, and Decision Making
Acceptability of Risk
Summary
References
Chapter 2: Bayesian Network Basics
Introduction
Laws of Probability
Bayesian Updating Versus Bayesian Network
Procedure for BN Development
Conditional Probability Tables
CPT from Data
CPT from Models
CPT from Expert Opinion
Fault Trees and BN
Structure of BN and Diagnostic Methods
Guidelines for Structuring BN
Modeling Time-Dependent Phenomena
Validation and Sensitivity Analyses
Strength of Influence
Sensitivity Analysis
Other BN Approaches
Summary
References
Chapter 3: Corrosion Models
Fundamentals of Corrosion
Electrolyte Thermodynamics
Deliquescence and Droplet Chemistry
Equilibrium Considerations of Deliquescence and Efflorescence
Kinetic Considerations
Thermodynamics of Corrosion
Corrosion Kinetics
Models of Corrosion Kinetics
Depassivation pH
General Corrosion of Carbon Steel
Corrosion of Carbon Steel in Seawater
Corrosion of Carbon and Low-Alloy Steel in Atmospheric Conditions
Corrosion of Carbon Steel in Soil Environments
Corrosion of Carbon Steel in CO2-H2S Aqueous Environments
Localized Corrosion
Environmentally Assisted Cracking (EAC) Models
Model Abstraction
Summary
References
Chapter 4: Statistical Models: Propagation of Uncertainty and Monte Carlo Modeling
Monte Carlo Modeling
Constructing a CDF from a Set of Values
Generating a Sample Following Any Arbitrary Distribution
Latin Hypercube Sampling
General Monte Carlo Approach
Monte Carlo Modeling Example, Part I
Sampling Correlated Variates
Monte Carlo Modeling Example, Part II
Additional Notes: Importance Sampling, LHS, and Correlated Samples
Statistical Models: Extreme Value Statistics
Statistical Models: Markov Chains
Summary and Commentary
References
Chapter 5: Corrosion Risk Assessment in Pipelines
Introduction
What Is Special About Pipelines?
How Is Pipeline Corrosion Risk Assessment Performed?
Requirements for a Good Corrosion Risk Assessment
List of Pipeline Corrosion Threats
Internal Pipeline Threat-Uniform Corrosion and Localized Corrosion
Internal Pipeline Threat-Erosion
Internal and External Pipeline Threat-Microbiologically Influenced Corrosion
External Pipeline Threat-External Corrosion
External Pipeline Threat-Stress Corrosion Cracking (SCC)
Probability of Failure-Remaining Strength
Note on Threat Interactions
Consequence of Failure
Qualitative and Semi Quantitative Methods for Pipeline Risk Assessment
Rules of Thumb
Indexing Methods
Barrier-Based Methods (Bowtie)
Quantitative Risk Assessments
Corrosion Rate Assessment
Statistically Active Corrosion (SAC) Analysis Using ILI
Pipeline Corrosion Modelling
Risk Assessment
Pipeline Quantitative Risk Assessment
Pipeline Probabilistic Risk Assessment
Decision Making
How to Prioritize Data Collection? When to Stop Collecting Data? How Many Direct Inspections Are Required to Be Sure a Pipelin...
Step 1: Pre-Assessment Step and Data Collection (How to Prioritize Data Collection?)
Step 2: Indirect Inspection Step (When to Stop Collecting Data?)
Step 3: Direct Examination Step (Quantification of the Decrease of the Corrosion Risk)
Step 4: Post Assessment Step (Verification)
What Level of Risk Is Acceptable? Risk Criteria
How to Communicate Risks Effectively to Decision Makers?
Visualization of Dependencies Between Parameters of Interest (Example SCC in Fuel Grade Ethanol Pipe)
Visualization of Data Uncertainty (Example: Pipeline Erosion)
Comparing Threats (Example: Multi-Analytic Risk Visualization Software)
Final Note: Worst-Case Scenarios Hinder the Decision-Making Process
Future of Pipeline Corrosion Risk Assessment
New Threats Must Be Expected (Hydrogen)
New Risk Assessment Methods (Model Pipeline Lifecycle)
Conclusion
References
Chapter 6: Oil and Gas Production Systems
Introduction
Corrosion Failure Modes
Environmentally Assisted Cracking (EAC)
HSC
HSC Mode
HSC Degree
Environmental Factors
Bulk pH
H2S Content
Chloride Concentration
Ferrous Ion Concentration
Scale Formation
Metallurgical Factors
Alloying Elements
Stresses and Strain Rate
Deformation Mode
Design and Operational Factors
Mobile Hydrogen
HSC Mode and Degree
Case Study 1
Case Study 2
Discussion
Hydrogen Embrittlement of Bolting
Stress Corrosion Cracking (SCC)
Metallurgical Factors
Stacking Fault Energy
Coherent Phases
Cold Work
Environmental Factors
Halide Concentration
Interactions
Summary
References
Chapter 7: Nuclear Energy
Introduction
Probabilistic Approaches
Corrosion and SCC in Light Water Reactors
SCC of Ni-Base Alloys in High Temperature Water
Corrosion Potential (Ecorr)
Temperature
Strain Rate
Deformation Mode
Grain Size
Grain Boundary Carbides
Oxide Stability
Atmospheric SCC of Dry Cask Storage System
The Environment Sub-model
Moisture Concentration
Relative Humidity and Deliquescence Relative Humidity
Aerosol Concentration
Condensed Droplet Chemistry
Materials and Design Factors
Pitting and Environmentally Assisted Cracking
Stable Pitting
Pit to Crack Transition
Loading Factors
Assessment Using the Model
Summary
References
Chapter 8: Localized Corrosion in Saline Environments
Marine Corrosion Systems
Simplified BN Model
Building a Detailed BN and Populating the CPT
Factors Affecting Localized Corrosion
Effect of Alloying Elements-Alloy Equivalence Number
Fabrication Effects
Effect of Seawater Composition
Crevice Tightness/Gap
Temperature
Factors Affecting OCP
Validation of the BN Model
Corrosion Under Insulation
Above Ground Carbon Steel
Buried Carbon Steel
Stainless Steel
BN Approach
Summary
References
Chapter 9: BN for Reinforced Concrete Structures
Introduction
Risk Assessment of RC Structures
Deterioration of RC Structures
Uncertainties of the Deterioration
Bayesian Network Applied to the Risk Assessment of RC Structures
Bayesian Updating
Static Bayesian Network Model
Dynamic Bayesian Network Model
Non-parametric Bayesian Network Model
Advantages and Challenges in Constructing BN Model for RC Structures
Example BN for an Aging RC Structure
Develop a Future Risk Assessment System
Summary
References
Chapter 10: Coatings
Coating Systems
Organic Coatings Exposed to Atmosphere
Design Factors
Mechanical Damage
Substrate Surface Quality
Coating Quality
Design Factors
Environmental Severity
Sensitivity Analysis
Dynamic BN
Metallic Coatings/Linings for Corrosion Resistance
Galvanic Effect
Coating Durability
Coating Quality
Summary
References
Chapter 11: Summing Up
Knowledge Analytics
Accelerated Testing and Service Performance
Challenges in BN Modeling
Aging Systems and Unknown Unknowns
References
Index