Flow Induced Top Of Line Corrosion In A Wet Gas Pipeline Corrosion Model Use To Improve Operational Integrity

An in-line integrity inspection revealed unusual, yet substantial top of line corrosion in a low flow, moderately corrosive wet gas pipeline, despite apparently successful bottom-of-line corrosion control. Following the in-line inspection a root-cause corrosion investigation was conducted, including further detailed analysis of the inline inspection data, a failure analysis of pipeline cut-out samples, and a corrosion modeling assessment using an in-house corrosion prediction tool. The corrosion modeling prediction, which provided a reasonable interpretation of where and why corrosion occurred, proved to be a major element in identifying the cause of the corrosion. Also, the corrosion modeling was key to identifying needed performance improvements for the pipeline corrosion control. Introduction In South Texas, wet gas is transported to sales through a slightly uphill, 13.4 kilometer (8.3 mile) long, 32.4 centimeter (12.75 inch) O.D. pipeline constructed of 7.3 millimeter (0.288 inch) thick API X-52 steel. It is operated at 73.8 bar (1070 psig), at an inlet temperature 49°C (120°F), and at a gas rate of 15 MMSCFD. To control corrosion from the 1.4% CO2 rich gas and potential bacteria contamination, a chemical corrosion control program treated the line monthly with 50-gallons of corrosion inhibitor and one gallon of biocide. In March 2007, the pipeline was pigged using a high-resolution Magnetic Flux Leakage (MFL) tool to assess wall thickness integrity. Three corroded sections (12.2 meter/40 feet each) were replaced based upon analysis of the MFL inspection data. Visual examination confirmed the presence of a large number of broad shallow pits in the top (10-2 o'clock). The observed corrosion attack did not conform to the conventional top of line (ToL) corrosion mechanism, wherein acidic condensing water attacks the top pipe and effective corrosion control is challenged by how to carry over the corrosion inhibitor to the top surfaces. Further, the observed ToL attack was not located early in the pipeline where the condensing conditions would normally be the worst. Therefore, to understand what was occurring and thus define an appropriate corrosion mitigation approach, a corrosion root cause analysis was initiated. Failure Analysis In the root cause assessment, the failure analysis portion included visual examination of retrieved samples, corrosion product analysis, and bacteria analysis. Visual examination of cleaned pipe samples confirmed the presence of broad and shallow ToL pits and minimal bottom-of line damage. Corrosion product analysis revealed FeS2, FeCO3, FeSO4.7H2O, and Fe3O4. Bacteria assessments found low indications of aerobic bacteria, organic acid producing bacteria, Sulfate Reducing Bacteria (SRB), and low nutrient bacteria. Unfortunately, the presence of the aerobic bacteria also suggested that contamination during the sample recovery was probable. Inspection Data Analysis To provide further details on the corrosion, the smart pig data was reviewed in detail, as shown in Figure 3. In general, along the slightly uphill profile, the wall loss was distributed non-uniformly. Little attack was observed Bottom-of-Line (BoL), between 5 to 7 o'clock. Top of the line attack (10-2 o'clock) was observed at 1.3 - 3 miles and sidewall attack (2-5 and 7-10 o'clock) at 2 - 8 miles.