Abstract
This paper proposes a Finite-Element-Analysis-based mechanistic model to predict internal localized pitting corrosion rates of petroleum pipelines in sweet (CO2) production environments. In this model, the computational domain consisted of a hemispherical pit and a thin boundary layer of an electrolyte solution. The mesh was generated using quadratic triangular elements in the Cartesian coordinate system whereas a moving mesh method was utilized to track the dynamic pitting propagation. The flux rate of each participating chemical ionic species was computed by solving the Nernst-Planck equation. Specifically, the convection was obtained by solving the Navier-Stokes equations. The electric field in the electrolyte solution was computed based on the Poisson equation with electroneutrality whereas a Debye-Hückel approximation was applied to describe the variation of potential at the metal-solution interface by reason of the existence of the electrical double layer. The ionic concentration distribution was solved using Fick’s Second Law. Consequently, the growth rate of a pre-existing pit was predicted. Meanwhile, laboratory tests were conducted to validate the proposed model, demonstrating that the developed model agrees well with experimental data. Furthermore, numerical studies were performed to characterize the effects of convection and chloride ion concentration on pitting corrosion rates. Hence, the model presented herein is able to predict localized pitting corrosion rates and incubation times for its onset in a given sweet system set of operating conditions as well as the onset of pit passivation incubation time. The technical benefits to be gained by the corrosion engineering community and pipeline operators include a better understanding of when to batch chemically treat a pipeline before pitting becomes autocatalytic and when it may be impossible to “turn-off” the pitting excursions due to operationally delaying proper corrosion inhibition practices.
