The control of internal corrosion of carbon steel structures is a critical element of asset integrity management program in oil and gas production facilities. The selection of a corrosion inhibitor for field deployment is typically based on laboratory screening tests, followed by field trials. This paper describes some of the experience gained in using new localized corrosion monitoring (LCM) technique in laboratory beaker and flow loop tests to differentiate the capability of corrosion inhibitors in controlling localized corrosion. A new data analysis program was employed to identify the occurrence of localized events in terms of magnitude, duration and distribution. This information is used to assess severity and time and spatial distribution of localized corrosion activities. Good correlation with post exposure metallographic examination of the test electrodes was obtained. Examples of monitoring localized corrosion processes including mild steel and stainless steel in various environments are also presented.
Corrosion can manifest in the form of general or localized corrosion. The direct cost of corrosion in the US was reported to be $276 billion per annum in a recent study published in July 20021. An earlier study also suggested that localized corrosion accounted for 80% of process plant failures2. However, it is recognized that corrosion can be controlled if appropriate measures are implemented. For example, it is common industrial practice to deploy corrosion inhibitors to control internal corrosion of carbon steel structures in oil and gas production facilities. One critical component of any corrosion management program is the measurement and verification of the effectiveness of the corrosion control strategy. Such measurements may be by means of direct or indirect methods, e.g. weight loss coupons, electrical resistance measurements, electrochemical monitoring, non-destructive inspection and potential measurements, etc. Selection of the most appropriate techniques is dependent upon the service environment as well as the type of information required.
From a practical perspective, general corrosion is relatively easier to monitor and to predict; whereas due to the random nature of localized corrosion, it is more difficult to monitor. The conventional approach in corrosion monitoring is mainly based on retrospective analyses, typically weight loss measurements and inspection, etc. Although the information may appear to be reliable and the data may be used to trend the corrosion behavior over time in the case of general corrosion, such information may not be relied upon to provide longer term representative behavior of localized corrosion. This is because localized corrosion events, such as pitting, do not corrode at a constant rate. The localized corrosion activity (e.g. pitting) can occur in a recurring process of initiation, propagation and repassivation.
In recent years, electrochemical corrosion monitoring techniques are gaining wider applications in evaluating localized corrosion. The techniques employed include potentiodynamic polarization scans3 and, in particular, electrochemical noise4 (EN) measurement, which is based on the monitoring of corrosion potential and current fluctuations. Analysis of the data can provide information on pit initiation, propagation and repassivation behavior5. A number of analytical methods have been developed to assist the identification of the likelihood of occurrence of the localized corrosion events, which are either based on statistical analysis of the EN data6-10, or digital signal processing/transformation techniques to provide corrosion information11-15. The statistical parameters obtained from the digital s
