Development of predictive model for sulfur species thermal dissociation and related corrosivity in condensate feedstock

摘要

High temperature sulfidic corrosion in refinery process streams depends on the presence and concentration of specific sulfur species. These species thermally dissociate (break down) when heated, producing hydrogen sulfide (H2S). Each sulfur species decomposes at a particular temperature which is thought to be its boiling point. The evolved H_2S reacts with a metal surface at high temperatures (250 to 350°C), forming metal sulfide, and therefore compromising pressure containment wall thickness and ultimately leading to release of hydrocarbons with serious consequences. Several incidents of high corrosion rates and premature failures of refinery process equipment and pipes have been reported in the past [1,3]. Material selection for these facilities was based on prediction tools using total sulfur and not considering the presence of specific sulfur species known as active sulfur. Some of these species dissociate at temperatures as low as 60°C, producing quantities of H_2S responsible for the metal wastage at high temperature. Concentrations of these sulfur species vary from one feedstock to another and have different dissociation temperature ranges. In the previous works by the authors [6,7], dissociation temperature profiles for several condensate and crude feedstock were generated. These profiles are important indicators for estimating the extent of corrosion, and predicting the affected process loop. However, to enable fast and accurate prediction of corrosivity of a particular feedstock, it is necessary to develop a model based on the kinetics and thermodynamic behaviour of sulfur species thermal decomposition. A pilot plant depicting the actual process units was used for data collection. The tested variables were temperature, pressure, liquid hourly space velocity (LHSV) and sweep gas flow rate. For each test condition, a feed and products mass balance was carried out. The extent of dissociation was evaluated using sulfur species material balance across the unit. Monitoring of H_2S in the produced gas was achieved using three complementary methods, namely on line detector, on line GC, and intermittent GC-SCD techniques. This paper will discuss the experimental setup to facilitate kinetic data collection, the significance of the selected test parameters, results obtained towards the development of the predictive tool, and will relate the data to previous work on the dissociation characterization of feedstock carried out by the authors.