Carburization is a high-temperature corrosion problem experienced in industrial processes such as ethylene production, natural gas reforming and coal gasification. The phenomenon takes place mainly in the petrochemical industry, where ethylene is produced in pyrolysis furnaces by thermal cracking of hydrocarbons in a steam hydrocarbon mixture at temperatures up to 1100oC. In this cracking process, coke deposition occurs at the inner walls of the cracking tubes. Carbon is transferred from the gas atmosphere through the porous coke at the alloy surface, where it diffuses in the interior and forms alloy carbides. Despite the theoretical and experimental work as well as the failure cases reported in the literature, carburization is not taken into account quantitatively in the design codes. For example, the API 530 standard used for the calculation of heater-tube thickness in petroleum refineries provides guidelines for selection of tube materials based solely on criteria for creep resistance. Carburization is only mentioned as a potential mechanism that could limit the service life of the tubes. In this example the ranking of carburization resistance of the steels listed in the API 530 standard was made possible through a simulation of the carburization process. The simulation for the 316 austenitic stainless steel is shown in Fig.1a. The volume fraction of carbides is plotted as a function of distance from the surface. The time required for the carbides front to reach the mid-thickness was used as a criterion of carburization resistance. The calculated carburization mid-thickness time for the API 530 steels is shown in Fig.1b. The austenitic grades exhibit a higher carburization resistance than the ferritic grades. Among the austenitic grades the stabilized steels 321 and 347 exhibit the highest carburization resistance.

Contributors: G.N. Haidemenopoulos, G. Samaras

Reference: G.F. Samaras and G.N. Haidemenopoulos, Carburization of high-temperature steels: A simulation-based ranking of carburization resistance, Engineering Failure Analysis, 51, pp.29-36, 2015.

Fig 1a

Fig 1b