Predictive Modeling for Corrosion Rate on Pipeline Corrosion in Seawater

ABSTRACT Marine oil pipelines are critical for the transportation of oil and gas from offshore production facilities to onshore processing plants. However, exposure to seawater, salt, and other environmental factors can cause corrosion in these pipelines, which can lead to costly and dangerous leaks and spills. Therefore, predicting and managing corrosion in marine oil pipelines is essential for ensuring safe and efficient operation. Uniform/general corrosion, pitting corrosion, and microbiologically influenced corrosion (MIC) are the most common forms of corrosion found in marine oil pipelines. These corrosion mechanisms can lead to thinning of the pipe walls and the formation of pits, which can significantly weaken the structure of the pipeline. To detect and monitor corrosion in marine oil pipelines, various in-line inspection (ILI) tools are available. Some of the most commonly used ILI tools for marine pipelines include magnetic flux leakage (MFL)[1], ultrasonic thickness measurement (UTM), and eddy current inspection (ECI). However, the existing tools are not efficient because of low accuracy. Therefore, a corrosion rate model was developed for the future rate of corrosion in marine oil pipelines. The developed model accounts for various factors such as line diameter, line temperature, line pressure, CO2 concentration, H2S concentration, Volatile fatty acid concentration, Bacteria count(SRBs), Material of construction, service life, bicarbonate ion concentration, chloride ion concentration, sulphate ion concentration, pH, clamp/repair history and details, oil, water and gas flow rate, flow velocity and regime, Inhibitor/biocide frequency, Oil characteristics and Kinetics of reaction to estimate the expected corrosion rate over a given time period [2]. Also the developed model combines the both linear corrosion growth rate, and non-linear corrosion growth rate. The model was trained on historical data of corrosion rates for different conditions and validated on new data to ensure accuracy. Additionally, the model could be updated in real-time with sensor data from the pipeline, allowing for continuous monitoring and prediction of corrosion rates. This could help operators proactively manage and maintain the pipeline to prevent corrosion-related failures and minimize downtime. A validation study was conducted on the developed model using a dataset of real-world pipeline corrosion data. The model was trained on a subset of the data and tested on a separate subset. The results showed a high level of accuracy, with an overall accuracy of 96%. This level of accuracy suggests that the model is reliable and can be used to inform pipeline integrity management and planning with a high degree of confidence.