Integration of Localized Internal Pitting Corrosion and Flow Models

Abstract A model was previously developed to predict internal pitting corrosion of oil and gas pipelines. The model considers the influence of temperature, total pressure, H2S partial pressure, CO2 partial pressure, the concentrations of sulphide, bicarbonate and chloride ions and the production rates of oil, gas, water, and solids. The model was based on experiments carried out in the laboratory at high pressure and high temperature under the operating conditions of the oil and gas pipelines and was validated by field trials and using field data. The model accounts for the statistical nature of pitting corrosion. It predicts the growth of internal pits based on readily available operational parameters from the field. The model also considers the variation of the pitting corrosion rate as a function of time and determines the precision of its prediction. Recently a microbiologically influenced corrosion (MIC) risk model has been developed and merged with the internal pitting corrosion model so that the integrated model predicts the optimized pitting corrosion rate due to both non-MIC and MIC activities. In this paper an approach was developed to integrate the internal pitting corrosion model with a flow model. The pressure drops that occur during single phase gas flow, single phase oil flow, two-phase flow and three-phase flow were calculated. Based on the pressure drop and elevation change, the pipeline locations where water and solids accumulate were estimated. The pitting corrosion rates in those locations were then predicted by superimposing the internal pitting corrosion model on the flow model. The pitting corrosion rates were further corrected to account for the variation in the field operating conditions. Though the validity of both the internal pitting corrosion model and MIC models have been individually validated using field data and the flow models have been extensively used for many years in pipeline operation, the validity of the integrated model should be evaluated before it can be used reliably.