Modern Approaches to Mathematical Modeling of Pipeline System Lifetime Considering Corrosion-Erosion Wear: A Review

This article presents a review of modern mathematical models used to predict the service life of pipeline systems, taking into account corrosion-erosion wear. The main approaches to modeling pipeline degradation processes are analyzed, including analytical and numerical methods based on both deterministic and statistical frameworks. Special attention is paid to the influence of hydrodynamic and chemical factors on wear rate and defect formation. The review covers models of localized (pitting) corrosion, uniform thinning, and erosion caused by fluid flow. Analytical models allow for rapid estimation of residual wall thickness or critical pressure but have limited accuracy under complex operating conditions. Numerical modeling, particularly finite element methods (FEM) and computational fluid dynamics (CFD), provides more accurate assessments of the stress–strain state and wear processes in pipelines. Probabilistic approaches, including those based on various distributions (Weibull, normal, log-normal) and Monte Carlo simulations, make it possible to incorporate variability in parameters such as defect depth, length, and growth rate, which is critical for failure risk assessment. The article also examines modern approaches to estimating time to failure and residual service life. Special emphasis is placed on modeling random fields of defects, parameterization of the Weibull distribution, and its adaptation to inspection data. Based on the review, promising research directions are outlined, including the integration of monitoring data with mathematical models, the development of digital twins for pipeline systems, and the application of hybrid stochastic approaches to improve predictive accuracy.