Effect of flow pattern at pipe bends on corrosion behaviour of low carbon steek and its challenges

• Main Authors: Muhammadu, Muhamadu Masin, Md. Sheriff, Jamaluddin, Hamzah, Esah
• Format: Article
• Language: English
• Published: Penerbit UTM Press 2013
• Subjects: TJ Mechanical engineering and machinery
• Online Access: http://eprints.utm.my/id/eprint/33545/1/JT63SK_July2013_08.pdf; http://eprints.utm.my/id/eprint/33545/; http://dx.doi.org/10.11113/jt.v63.1123

Recent design work regarding seawater flow lines has emphasized the need to identify, develop, and verify critical relationships between corrosion prediction and flow regime mechanisms at pipe bend. In practice this often reduces to an pragmatic interpretation of the effects of corrosion mechanisms at pipe bends. Most importantly the identification of positions or sites, within the internal surface contact areas where the maximum corrosion stimulus may be expected to occur, thereby allowing better understanding, mitigation, monitoring and corrosion control over the life cycle. Some case histories have been reviewed in this context, and the interaction between corrosion mechanisms and flow patterns closely determined, and in some cases correlated. Since the actual relationships are complex, it was determined that a risk based decision making process using selected ‘what’ if corrosion analyses linked to ‘what if’ flow assurance analyses was the best way forward. Using this in methodology, and pertinent field data exchange, it is postulated that significant improvements in corrosion prediction can be made. This paper outlines the approach used and shows how related corrosion modelling software data such as that available from corrosion models Norsok M5006, and Cassandra to parallel computational flow modelling in a targeted manner can generate very noteworthy results, and considerably more viable trends for corrosion control guidance. It is postulated that the normally associated lack of agreement between corrosion modelling and field experience, is more likely due to inadequate consideration of corrosion stimulating flow regime data, rather than limitations of the corrosion modelling. Applications of flow visualization studies as well as computations with the k-ε model of turbulence have identified flow features and regions where metal loss is a maximum