Diagnostic and prognostic development of a mechanistic model for multiphase flow in oil-gas pipelines

Abstract

A mechanistic model has been applied to predict corrosion rate in multiphase (water–oil-gas) flow regimes. It is applied due to the inadequacies of existing models to adequately address a wide variety of corrosion-dependent parameters with special emphasis on multiphase flow parameters. This model considers water cut, volumetric phase fraction of oil and gas, temperature, pressure, and other data collected relating to the pipe age, sand particle deposit, CO2 partial pressure, flow velocity, pH, chloride contents, mass transfer coefficients, and hydrogen sulfide. The applied model results show well above 91 percent prediction accuracy, while the prediction error ranges between 0.022 and 0.045 percent. When predicted results compared actual value (experimental data) with DeWaard and Milliams’ model; used in the oil and gas industries, it showed good agreement. It is seen that the resulting model produced an R2 value of 0.91 in comparison to 0.53, the value for Dewaard and Milliams’. It is obvious from the numerical simulation that the multiphase flow parameters have major effects on the entire rate of corrosion of oil and gas multiphase flow pipelines. Finally, this research concluded by showing the developed model performance, and a comparison was made with numerical results. It is expected that the outcome of this research will be of significant help in strengthening the application of mechanistic models for predicting corrosion rates. This model will be useful for oil and gas industries worldwide for accurate corrosion rate prediction.