Fracture Toughness Behavior of Carbon Steels in Mildly Sour Waters

Abstract

Abstract Corrosion and cracking of typical materials used in refinery equipment exposed to sour water has long been a concern for the industry [1]. The failure of an amine absorber tower in 1984 heightened the need to understand the performance of these materials in sour water encountered by the industry [2]. This phenomenon is often generically referred to as “wet H2S cracking”, though it includes various forms of damage. In the years following the absorber tower failure numerous studies were undertaken to understand the performance of typical pressure vessel steels exposed to sour water including fracture toughness testing in NACE A or NACE B solutions [3]. NACE A and B solutions are significantly more aggressive solutions (lower pH and higher Hydrogen Sulfide [H2S] concentration) than most common refinery sour waters. Due to the severity of the testing solution, the results are likely overly conservative and may lead to unrealistic integrity assessments of refinery vessels by applying API 579 Fitness for Service (FFS) assessments. Following the 1984 incident, most of the research carried out into sour water environments was targeted to material susceptibility and effects of sour water on new steel. Limited baseline data are available that quantify the effects of sour water on the mechanical properties on service aged or vintage materials. In fact, most data generated on the fracture toughness of steels in sour water has been undertaken for upstream projects, where the conditions are often more aggressive than in downstream applications. An ASME-API 579 Part 9 flaw analysis is required to be performed on wet H2S crack-like defects that fail an ASME-API 579 Part 7 Level 2 analysis [4]. To account for the loss in toughness associated with the presence of hydrogen embrittlement, lower bound crack arrest fracture toughness (KIR) is normally used as a lower bound estimate. Some wet H2S defects, such as Hydrogen Induced Cracking (HIC), require an API 579 Part 9 crack like flaw assessments even though they typically pass the Level 2 analysis. However, the FFS assessment is limited by the small tolerable crack lengths from API 579 Part 9 analysis due to the application of the lower bound KIR values. The use of lower bound KIR values is potentially overly conservative because it is a key input parameter into a crack like flaw assessment. Furthermore, the presence of H2S typically will cause subcritical crack growth due to the presence of hydrogen in the fracture process zone ahead of the crack tip. This can take place under constant load, or under fatigue. The latter issue is often evaluated for oil production facilities, particularly for offshore applications. In the downstream sector of the refining industry, the application of crack growth modelling is not normally applied due to the scarcity of applicable experimental data. As a result, either weld repair or protective coatings are typically used to prevent known tolerable flaw sizes from growing during future operation. To address such issues, a comprehensive test program was conducted to evaluate both the fracture toughness and subcritical crack growth properties of ex-service steels in mildly sour waters, which are more applicable to the downstream refining industry. The environmental severity was also explored by varying combinations of pH and H2S partial pressure. The main objective was to characterize the steel specimen performance in near neutral pH with high as well as low H2S concentrations.