Introduction
High-temperature sulfidic corrosion of carbon and lowalloy steel components has long been a recognized phenomenon in petroleum refineries. Crude oils often contain from 0.5 to 5 wt% sulfur in a variety of different sulfur compounds. Sulfidic corrosion was first encountered in refineries in crude distillation units, thermal and catalytic cracking plants, thermal reforming, and coking units where the crude oil and its fractions were processed at temperatures exceeding 500°F (260°C).¹ Steel alloys containing 5% chromium (Cr) or greater (i.e., 7% Cr, 9% Cr, and 12% Cr, in this order), were found to have increasing resistance to sulfidic corrosion. Over time, empirically based corrosion prediction curves were generated and improved based on refinery experiences. These curves are still useful to this day for refining processes containing significant quantities of sulfur compounds.²
In the 1940s and 1950s, the advent of refining processes that utilized hydrogen, such as catalytic reforming and hydroprocessing, introduced another facet to sulfidic corrosion. It was observed that for sulfidic services containing hydrogen, steel alloys containing up to 9% Cr were, at best, only slightly more corrosion resistant than carbon steel (CS).¹ Sulfidic corrosion in the presence of H₂ is often referred to as H₂-H₂S corrosion. Much research was done and some of this work was published. A separate set of corrosion prediction curves for H2-H2S conditions were compiled and published and are still generally useful.³ Several licensors and refining companies have their own methodology, or set of proprietary corrosion curves, that are used for material selection and corrosion rate prediction.
By the 1990s, several refiners began to report sulfidic corrosion in equipment such as piping and reboiler furnace tubes in fractionation and distillation facilities downstream from hydrotreaters and hydrocrackers.⁴ The corrosion was considered unusual in these instances because these facilities are considered to be H₂-free, and the total sulfur content of the hydrocarbon stock was very low. In some cases, chromium-molybdenum (Cr-Mo) steels corroded at the same rates as CS. It was recognized that the existing corrosion prediction curves were inadequate for these specific circumstances, and efforts were made to better understand the problem. This led to the formation of NACE Task Group 176, Prediction Tools for Sulfidic Corrosion.