Effect of SO2 and NO2 on Corrosion and Solid Formation in Dense Phase CO2 Pipelines

Abstract CO2 has been transported and used for EOR (Enhanced Oil Recovery) for more than 30 years with a good track record. Based on the good experiences it seems to be generally accepted that corrosion will be insignificant in CO2 transport pipelines as long as the water content is well below the water solubility. The solubility in pure CO2 at 100 bar is about 1900-3200 ppmv in the temperature range 4-25 °C. There is no consensus on what the actual target concentration for the maximum water content should be in the CO2 to be transported, but it is often referred to 500 ppmv in the CCS (Carbon Capture and Storage) literature. The question is whether this apparently safe water level also applies when glycols, amines and flue gas contaminants like SOx, NOx and O2 are present in moderate amounts. These impurities dissolve readily in water and induce an aqueous phase at a much lower water concentration than the solubility limits reported for pure CO2 and CO2 contaminated with hydrocarbons. When SO2, water and O2 are present sulphurous and/or sulphuric acid (H2SO3 and H2SO4) might form. The minimum water concentration required for acid formation is not known, but the presence of FeSO3 and/or FeSO4 seen on steel surfaces exposed in dense phase CO2 at low water content (less than 500 ppmv) shows that corrosion takes place when 100-344 ppmv SO2 is present. The corrosion rate was further increased when NO2 was added to the system. Rotating autoclave experiments run with steel exposed to 100 bar CO2, 488 and 1220 ppmv water and 96-478 ppmv NO2 resulted in weight loss corrosion rates of 0.05-1.6 mm/year.Typically less than 5% of the added impurities were apparently consumed by corrosion in the NOx and SOx experiments before the corrosion rate slowed down. The difference in the impurity concentration at start up and when the experiment was terminated was much larger than the consumption estimated from corrosion. The impurities became apparently “non-active” during the exposure and it can be questioned if the measured corrosion rates in the present experiments and in other reported lab experiments reflect the worst case conditions in the pipeline. The experimental observations so far support the need for a dynamic test system with replenishment of impurities and instant (continuous) analyses of the dissolved impurity concentrations. Such dynamic tests are required in order to define acceptable CO2 specifications with confidence.