Markov Chain Model Helps Predict Pitting Corrosion Depth and Rate in Underground Pipelines

A continuous-time, non-homogenous pure birth Markov chain serves to model external pitting corrosion in buried pipelines. The analytical solution of Kolmogorov’s forward equations for this type of Markov process gives the transition probability function in a discrete space of pit depths. The transition probability function can be completely identified by making a correlation between the stochastic pit depth mean and the deterministic mean obtained experimentally. Previously reported Monte Carlo simulations have been used for the prediction of the evolution of the pit depth distribution mean value with time for different soil types. The simulated pit depth distributions are used to develop a stochastic model based on Markov chains to predict the progression of pitting corrosion depth and rate distributions from the observed soil properties and pipeline coating characteristics. The proposed model can also be applied to pitting corrosion data from repeated in-line pipeline inspections. Real-life case studies presented in this work show how pipeline inspection and maintenance planning can be improved through the use of the proposed Markovian model for pitting corrosion.