Measurement of Crude Oil Corrosivity Using Radioactive
 Tracer Technology, 03-9263

Printer Friendly Version

Principal Investigators
Martin B. Treuhaft
Douglas C. Eberle

Inclusive Dates: 07/01/01- 09/30/02

Background - Technology is needed to determine accurately and quickly the corrosivity of crude oils and many of their derivative materials. Current methods lack timeliness, reliability, and accuracy, leading to inconsistencies between projected and actual corrosion experiences in refining and processing operations. Currently, there are no robust models for predicting crude oil corrosivity. Traditional estimators, based on sulfur and naphthenic acid concentration measurements, are inadequate for assessing the true corrosivity potential of specific crude oils. This inability to assess the corrosivity has proven problematic for crude oil buyers, sellers, and refiners. As it becomes necessary to process poorer crudes, corrosion problems and economic issues will place a premium on obtaining accurate and timely corrosion data.

Approach - Collectively, the above mentioned factors provided an opportunity to extend SwRI's radioactive tracer technology (RATT®) capabilities to the area of crude oil corrosivity measurement and to develop opportunities for providing corrosivity measurement and corrosion/erosion research and development services. Thus, the research team's focus was twofold: 1) to adapt SwRI's RATT and develop test protocols for making accurate corrosivity measurements and predictions, and 2) to establish a capability for providing standardized measurement and enhanced corrosion/erosion research and development services using this technology. The immediate objective was to apply RATT to the detection and measurement of crude oil corrosivity in a manner conducive to standardized testing and commercial application. Bulk neutron RATT was pursued because its high sensitivity would likely allow for the development of a meaningful laboratory test that could quickly characterize the affects of crude oil corrosiveness on selected fluid-handling materials. Such a laboratory setup could be used as a screening tool to rapidly determine which oils are more corrosive than others with respect to specific pipeline materials, and as a research tool for further investigation of corrosion and other material related issues.

Project efforts focused on three areas: 1) determining the time-resolved detection limits and corrosion-prediction capabilities of RATT when applied to the measurement of crude oil corrosivity, 2) developing bench-top hardware and laboratory testing procedures for applying RATT to crude oil corrosivity measurement, and 3) demonstrating corrosivity measurement capability on neutrally, mildly, and highly corrosive crude oils. In addition, the project served as a stepping stone for indicating potential for developing laboratory hardware and procedures for measuring real-time corrosion and erosion in piping components under simulated refining and processing operations, with possible extension to real-time, on-site, corrosion measurement and monitoring.

The technical approach was rooted in SwRI's experience in using radioactive tracers to measure real-time wear in operating engines by monitoring lubricating oil radioactivity caused from the accumulation of wear particles abrading from irradiated rings and bearings. Because of RATT's high sensitivity, extremely small changes in wear have been measured. Corrosion detection and measurement are characteristically similar. SwRI experiments investigated the ability for obtaining meaningful measurements of material degradation by monitoring the radioactive buildup of corrosion debris in the fluid. Testing was conducted at elevated temperature (600 °F) under flows simulating refinery conditions. The real-time development of corrosion was quantitatively monitored through on-line radiometric interrogation of the test fluid in a continuous flow loop. These on-line measurements are important to the evaluation because they provide a true historical record of the corrosion process as a function of test time and temperature.

Accomplishments - The test loop design and testing protocols were developed to simulate certain refinery conditions. Prospective designs and protocols were guided by CFD simulations and by an extensive literature search, which was initiated to gain a better understanding of corrosion detection and measurement. In addition to simulating specific refining conditions, high shear stresses during experimentation were intended to avoid corrosive film build-up (which could interfere with the corrosion measurements) and to add relevance to the results. Experimentation was accomplished using activated coupons inserted into a specially designed housing, providing confined tangential flows under controlled shear stress conditions.

Results exceeded expectations. Real-time corrosion measurements were made using neat mineral oil doped with cyclopentyl acetic acid (to simulate crude oil naphthenic acid content), crude oils from various locations worldwide, atmospheric gas oils, and vacuum gas oils, which are of particular interest to many refiners. In most cases, the onset of corrosion during warm-up was observed within 10 to 20 minutes, and well-defined corrosion curves were established within four hours of testing.

This method improves upon other methods of corrosion measurement by providing real-time corrosion data with much greater sensitivity. This unique capability, which allows one to view corrosion as a function of time and temperature with 10-minute resolution, has already produced some surprising results. For example, in some cases, high corrosion rates were observed at low temperature with corrosion tapering off at higher temperatures (likely due to highly corrosive constituents boiling off at low temperature and because of the presence of salt water in nondesalted crude oils). This type of information is not available using other methods.

Applying certain assumptions, such as the even distribution of corrosion on the coupon and insignificant amounts of scale build-up, corrosion rates in mils per year can be predicted. In addition, relative corrosion indices can be developed comparing corrosion trends for a given crude with similarly measured trends for "benchmark crudes" whose characteristics are known to producers, buyers, and refiners.

Representative data from crude oil corrosivity measurement experiments
Experimental setup for measuring crude oil corrosivity

2002 Program SwRI Home