The study of corrosion product and inhibitor films in multiphase flow

This study used electrochemical impedance spectroscopy and surface analysis techniques to determine the influence of corrosion product and inhibitor films on the CO₂ corrosion process in gas/oil/water multiphase flows. The flow loop experiments were carried out in a 101.6 mm I.D., 15 m long Plexiglas pipeline at 40°C and 0.136 MPa. The flow regimes of slug flow of Froude numbers 6, 9 and 12 and full pipe flow of 0.55 and 1.5 m/s superficial liquid velocity were examined for the inhibitor performance and inhibition mechanism.; The results show that the double layer capacitance, film capacitances and film resistance decrease greatly with an increase of inhibitor concentration. These decreases indicate a formation of inhibitor film on the steel surface to change the structure and chemistry of the electrolyte.; One simple semicircle in Nyquist plots is observed for blank tests and low inhibitor concentration tests. Only in slug flow at Fr 6 with 100 ppm and full pipe flow with 25 ppm inhibitor concentrations, could two clear time constants be seen in the Nyquist diagrams. The impedance responses in full pipe flow and slug flow are quite different.; From the scanning electron microscopy and focused ion beam data, corrosion product film porosity and thickness are determined. The porosity of the corrosion product film produced under full pipe flows is about 6% with a very small pore size (0.013μm2), while it is 24% and 30% for slug flows of Fr 6 and 9, respectively. The thickness of the corrosion product film increases considerably at full pipe flows compared to those at slug flow by using FIB techniques.; The clear images of the corrosion product and inhibitor films have been proposed in this study. For the first time, equations used to calculate the porosity and thickness for the inhibitor film and inner corrosion product film are derived. The calculated inhibitor film thickness is of nanometer scales and the porosity is about 35% for slug flows and 10% for full pipe flows. These reasonable results verify the validation of the proposed model.