The prediction of corrosion rate and its migration is of paramount importance to the oil and gas industry. In fact, the total annual cost of corrosion to the industry is estimated to be $1.372 billion. This cost includes $589 million attributed to surface pipelines and facilities, $463 million in downhole tubing expenses, and another $320 million in capital expenditures related to corrosion.
One of the more complex problems in predicting corrosion rates caused by carbon dioxide, is the effect of multiphase flow since many of the pipelines and flow lines carrying oil and gas are operating under two or three-phase flow conditions.
Different flow patterns lead to a variety of steel surface wetting mechanisms which greatly affect corrosion. In the absence of protective scales, multiphase flow can lead to very high fluctuating mass transfer rates (particularly in slug flow) which, in turn, can greatly affect the corrosion rate.
The objective of this article is to showcase how open literature corrosion data can be translated into knowledge using data-driven modeling techniques, which can lead to a digital twin for descriptive, predictive, diagnostic, and prescriptive purposes for pipeline and flowline corrosion.