A methodology for cost optimal reliability based inspection and replacement planning of pipelines being in the design-phase is described. The degradation mechanism considered is CO2 corrosion. The methodology is based on application of methods for structural reliability analysis within the framework of Bayesian decision theory. The planning problem is formulated as an optimization problem minimizing the expected lifetime costs with minimum acceptable reliability level as a constraint. The optimization parameters are the nominal design wall thickness, the number of inspections in the expected lifetime, the time intervals between inspections and the inspection method to be used.
Submarine pipelines are among the critical components of offshore installations. Pipeline failures commonly have significant economic implications for the operator in terms of lost production and replacement/repair. Failures may also constitute serious hazards in case of leakage of the containment. Due to the serious consequences of pipeline failures, inspection and maintenance activities are regularly carried out in order to secure the continued integrity and to meet regulatory requirements. The costs associated with these activities constitute a significant portion of the total operation costs of offshore installations. The present paper presents a methodology for cost optimal reliability based inspection and replacement planning of pipelines being in the design phase. The methodology is based on application of methods for structural reliability analysis within the framework of Bayesian preposterior decision theory. Through application of the theory for Bayesian preposterior decision analysis a consistent framework for evaluation of the (uncertain) consequences of different inspection and replacement plans is provided. Thus, expressing the consequences in monetary values, the inspection and replacement plan which provides the lowest expected costs in the expected lifetime can be selected. The potential of structural reliability methods as decision support tools in engineering application, is due to the consistent modeling of uncertainties.
