Corrosion-Fatigue Performance of High-Strength Riser Steels in Seawater and Sour Brine Environments

Although new high-strength steels have recently been developed to meet the demands of increased reservoir pressures, and sour production fluids, the corrosion-fatigue performance of these new higher-strength materials is largely unknown. The goal of this study was to fill this knowledge gap by generating corrosion-fatigue data in two aggressive environments: 1) a sour production brine, and 2) seawater with cathodic protection. The focus of the current paper is on stress-life (S-N) corrosion-fatigue results in these environments, as well as a baseline air environment. Experiments were performed on five different steels with yield strengths ranging from 848 MPa to 1080 MPa. Prior frequency-scan results based on corrosion-fatigue crack growth rate data demonstrated that not all of these material-environment combinations exhibit a saturation frequency where the detrimental environmental effect approached a constant value as the cyclic loading frequency is decreased. Consequently, S-N tests were performed at different frequencies (0.01 Hz, 0.17 Hz, and 1 Hz), depending on the fatigue life regime, in attempting to match the loading frequencies experienced in service. Corrosion-fatigue occurred at stresses well below the fatigue endurance limit in laboratory air, and cyclic lives in the seawater with cathodic protection environment were found to be 2X to 10X less than those in the baseline air environment, while cyclic lives in the sour brine environment were found to be 30X to 100X less than those in the baseline air environment. In both environments, degradation was greatest at lower stresses in the high cycle fatigue regime. The effect of material strength level had little or no measurable effect on the S-N corrosion-fatigue performance, and the effect of cyclic frequency on the corrosion-fatigue performance was mixed. The S-N response to these two variables differed significantly from recently measured fatigue crack growth kinetics in these same materials that were performed in a companion study. Possible reasons for these differences are discussed.