Abstract

Refinery operators rely on total sulfur (TS) content (wt.% S) and total acid number (TAN), reported in crude oil assays, to predict high temperature corrosion rates by organosulfur species and naphthenic acids. The sulfur exists in a variety of forms in crude oil, associated with particular molecular moieties; from the standpoint of corrosion, these are grouped into reactive (sulfide and mercaptan) and non-reactive (thiophenic) species. According to an industrial rule-of-thumb, only 1/3 of the TS is considered as reactive sulfur (RS) and, hence, this value, together with TAN, is typically used as input in corrosion models for prediction of corrosion rates. It was hypothesized in this research work that the prediction of a corrosion model should improve if experimentally measured reactive sulfur values were used as an input in modeling instead of employing the 1/3 rule-of-thumb. To measure the percentage of reactive sulfur in a given crude oil, sulfur species were separated into reactive and non-reactive fractions, for 10 vacuum gas oils (VGOs), by an Ag+-ligand exchange chromatography method. The concentrations of separated reactive and nonreactive sulfur fractions were measured by X-ray fluorescence spectroscopy (XRF). The sulfur separation method was validated beforehand, using model oil solutions of known concentration of corrosive and non-corrosive model sulfur compounds. Model corrosion rate simulations for the VGOs were performed using as input the experimentally determined reactive sulfur values and the 1/3 rule of thumb for the reported total sulfur content. The comparison of experimental data with model simulations showed the predictions improved for 5 out of the 10 VGOs when experimental values of RS were used as input.

Introduction

Naphthenic acids and sulfur species in crude oil cause severe corrosion of the steel equipment of crude distillation units in oil refineries.1–3 Because of rapidly changing oil economics, the refineries have inclined towards cheaper "opportunity crudes", but the high levels of corrosive species, mainly naphthenic acids and organosulfur compounds, in these crudes would reduce the life of the equipment, and also increase the risk of catastrophic failure.3 So the opportunity crudes are often blended with the crudes containing lower levels of corrosive species; this decreases overall concentration of corrosive species and the corrosion rates.4,5 However, corrosion rates are not simply proportional to the concentrations of naphthenic acids and sulfur species that are present in the crude oil.4,5 Without accurate estimation of corrosion rates by crude oils or their "blends", carbon steel equipment needs to be constructed with higher wall thickness for safety; if still insufficient, high alloy steels are required. In either case, refinery upgrades significantly increase the capital expenditure.6 Currently, refinery operators rely on survey data and empirical models for the estimation of corrosion rates,1,7–11 but due to the limited accuracy of available corrosion models, confidence on estimation of corrosion rates has not been achieved yet.

This content is only available via PDF.
You can access this article if you purchase or spend a download.