Modeling of Corrosion Rate Inside a Pressure Vessel
Author
Iversen, Nikolaj
Term
4. semester
Education
Publication year
2019
Submitted on
2019-05-31
Pages
36
Abstract
Dette speciale undersøger, om en beregningsmodel kan understøtte en digital tvilling, der forudsiger tilstanden af indersiden i trykbeholdere i olie- og gasindustrien og dermed kan forbedre planlægningen af inspektioner. Udgangspunktet er en MATLAB-implementering af de Waard-Milliams’ CO2-korrosionsmodel, som oprindeligt er udviklet til rørledninger og brøndhoveder. Modellen tilpasses anvendelse i trykbeholdere gennem yderligere antagelser og afprøves mod inspektionsdata for tre forskellige trykbeholdere med varierende indhold. Formålet er at forudsige korrosionsdybde ud fra CO2-drevet korrosion. Resultaterne viser, at modellen i nogen grad kan bestemme korrosionshastigheden for CO2, men at den ikke fuldt ud kan forklare de målte korrosionsdybder. For at opnå mere præcise forudsigelser bør yderligere mekanismer indgå, herunder korrosion forårsaget af H2S og effekter af salte i vand. Projektet fokuserer på korrosionshastighed og udgør et første skridt mod en mere komplet digital tvilling til trykbeholdere.
This thesis examines whether a computational model can support a digital twin that predicts the condition of the inner wall of pressure vessels in the oil and gas industry and helps optimize inspection scheduling. The work implements the de Waard-Milliams CO2 corrosion model in MATLAB, originally developed for pipelines and wellheads, and adapts it for pressure vessels using additional assumptions. The model is evaluated against inspection data from three pressure vessels with different contents, aiming to predict corrosion depth driven by CO2. Findings indicate that the model can estimate CO2-related corrosion rates to some extent but does not fully reproduce observed corrosion depths. To improve predictive accuracy, other factors—such as H2S-induced corrosion and the presence of salts in water—should be incorporated. The study focuses on corrosion rate modeling and provides an initial step toward a more complete digital twin for pressure vessels.
[This summary has been generated with the help of AI directly from the project (PDF)]
Keywords
Corrosions rate ; CO2 ; Tryk beholder ; Rust
Documents