Predicting the rate of underground corrosion of steel pipelines: A review

• M.A. Gavryushina, A.I. Marshakov and V.E. Ignatenko

A.N. Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Leninsky pr. 31, 119071 Moscow, Russian Federation

Abstract: Estimation of the probable rate of corrosion of underground steel pipelines has long been a challenging problem for engineers and researchers and is still of importance. The complexity of solving this problem is due to a large number of influencing factors (the composition of the ground electrolyte, gas and solid phases of the soil), their constant daily and seasonal changes, the use of cathodic protection and protective polymer coatings. Another feature of the corrosion process is its probabilistic nature. Due to the complex nature of the phenomenon, several different approaches have been developed to predict its rate. This review deals with the factors that affect the formation and development of corrosion defects in underground pipelines and the various methods used to predict the corrosion of pipelines. The basis for predicting the growth rate of corrosion defects in the outer wall of underground pipelines are methods for predicting the corrosion rate of pipe steel in soils, which can be divided into qualitative and quantitative ones. Qualitative methods are mainly used to determine the degree of soil corrosion activity (scoring methods). Scoring methods create prerequisites for quantifying the corrosion rate of steels in soils. However, the currently existing quantitative models of underground corrosion take into account no more than two corrosion factors. Due to the imperfection of quantitative simulation of steel corrosion in soils, the methods for predicting the corrosion of the outer wall of pipelines are based on statistical processing of data obtained either in full-scale corrosion tests of pipe steel samples or during pipeline inspection. Models of various types (deterministic, probabilistic, and those created using machine learning) are presented and the criteria of their applicability are analyzed.

Keywords: low-carbon steel, models, soil composition, probable corrosion rate, machine learning methods

Int. J. Corros. Scale Inhib., 2024, 13, no. 1, 254-287
doi: 10.17675/2305-6894-2024-13-1-13

Download PDF (Total downloads: 215)

Back to this issue content: 2024, Vol. 13, Issue 1 (pp. 1-629)