Abstract
Corrosion models and the laboratory testing they are based on, give a measure of corrosion rate but not the spatial distribution of corrosion. Corrosion risk assessments can include information on areas that are more susceptible to corrosion and can give a measure of the probability of corrosion occurring, but when the corrosion is localized rather than general, they do not indicate whether the corrosion is evenly spaced or clustered or how the corrosion will develop in extent over time. To deliver inspection, which is both effective and efficient, requires an understanding of the spatial distribution of corrosion along with the fraction of wall area affected by corrosion. Historically low level sampling using Manual Ultrasonic Techniques (MUT) has formed the primary basis for inspection of pipework on offshore oil and gas assets. For extensive corrosion, low level sampling can be effective and cost efficient but for less extensive corrosion, higher coverage is required to reliably reflect the nature of the corrosion.
Simulation of different spatial and depth distributions of corrosion, to reflect a range of base case scenarios, have been carried out in conjunction with simulations of different inspection strategies. Metrics have been developed to compare the effectiveness of different inspection strategies. This paper discusses the aims of inspection and how efficient and cost effective the different inspection strategies, such as sampling or screening, are in meeting these aims. A range of different corrosion base cases including general uniform, random localized and preferential locations will be examined in relation to the most appropriate inspection strategy, defined primarily in terms of coverage. Recognizing these different corrosion cases from previous inspection results and the optimum time to change inspection strategy will also be covered. This paper quantifies the relationship between spatial distribution of corrosion in pipework and the effectiveness of different inspection strategies.
