Risk based inspection and probabilistic assessment of corrosion in floaters

The project set out to improve the way widespread rust in pontoons of ageing semi-submersibles are managed. This is done by improving the planning of thickness measurements, the checking of the remaining corroded strength, as well as input to decision-makers. The project will contribute methods for sustainable and safe life extension of the drilling units with lowest CO2 emissions on the Norwegian sector. For the planning part a database design is proposed as alternative to the traditional file-based system. This makes it easier for engineers to use the data for inspection planning and predictions. A new corrosion model is suggested which considers observations as well as uncertainties. The model suits well with the risk-based corrosion inspection methodology that is developed. In this method the pontoon structure of the semi-submersible is divided in a way that each group have similar risk, so the places with higher risks are inspected first. The remaining strength with the corrosion should have same safety as new designs. Often the old design methods were conservative, so when re-analyzed with modern rules and software the structure can be found to be on the safe side, even with a certain amount of corrosion. But; the safety factors in the used design codes are made with uncertainties of new structures in mind. That?s why the behavior of corroded steel from an old ship is tested in tension in a lab for comparison. This showed that some material properties seemed to be reduced, which would have a large impact on industry practice, so this should be investigated at full scale. How the stress picture of the hull changes globally and locally is also compared to original thicknesses since rust is not uniform as assumed for thickness in design. Finally, decision input to managers is proposed through indicating future cost of maintenance actions which can help make sustainable and safe decisions regarding scrapping or life extension despite the fundamental uncertainty.

The project has led to improvements of our services and has partly been implemented to our existing clients. A small project with a new customer has also been executed. Dialogue is established to professionals on other fields such as fatigue through a mentor in another company, as well as technical authorities on ageing in PSA to discuss implications of the results and gaps in the current system. For society, the work plays a role of contributing to the safety of petroleum activities for ageing assets. It might initiate further research on parts of the topic Some parts showed results worth looking further into. The work partly criticizes the regulations and industry practice by identifying gaps introducing uncertainty, which may involve a change in industry practice. Recommendations involve more communication between structural and risk disciplines which may introduce new ways of managing ageing assets.

The company has clients with aging floaters approaching their design life. Amongst other issues, assessment of the remaining structural integrity of corroded areas is one of the demanding tasks to demonstrate the fitness for coming years to the class society. With this project we wish to improve this assessment by several measures. Firstly a database is to be developed based on existing thickness measurements of corroded floaters, tagged with relevant variables for corrosion, such as age, coating, humidity, heat flux, etc. This database can be used to investigate the impact of the different variables, and eventually to make predictions based on the developed statistics. Furthermore, the nature of corrosion is such that it produces an uneven surface. In large structures single thickness measurement samples are to represent larger areas. To understand the average corrosion and single defects that might occur over that area, an experiment is planned. Steel plates of various degree of corrosion is to be densely mapped with ultrasonic thickness measurements. The statistical parameters of the thickness distribution can then be applied to give a larger overview of an area based on few samples with an estimated uncertainty. Also, it is a goal to exchange dirty paper work at location, with a rugged tablet including a software for direct insertion of thickness measurement to database. This should include a user friendly graphical interface and readily understandable result presentation according to rules. Eventually, the ultimate goal is to develop a methodology to link measured thicknesses and their uncertainties with the local and global integrity of the corroded structure. Finally, a case scenario is to be applied to demonstrate the tablet with the software, applying the found statistic parameters from the database to predict future corrosion, present the results, and link it with the structural integrity to conclude on the fitness of the rig with aspect to corrosion.