Assessment of Corrosion Rates for Developing RBIs and IMPs for Production Pipelines

Corrosion rates (CRs) for defect parameters are playing a crucial role in creating an optimal integrity management plan (IMP) for production pipelines (well piping, inter-field pipes, cross country flow lines, and facility piping) with a thinning web and/or growing defects. CRs are indispensible when assessing the remaining strength, probability of failure (POF) and reliability of a production piping/pipeline with defect(s), and permit assessing the time to reaching an ultimate permissible POF, a limit state, or time to actual failure of the leak/rupture type. The CRs are also needed when creating a risk based pipeline inspection (RBI) plan, which is at the core of a sound IMP. The paper briefly describes the state of the art and current problems in consistent assessing CR parameters using different mathematical models of corrosion growth currently introduced to pipeline industry. The possible sources and the needed quality of direct assessment (DA) and in-line inspection (ILI) to construct comprehensive CR models are listed and formulated. Since corrosion or, in general, deterioration of pipelines is a stochastic time dependent process, the best way to assess pipeline state is to monitor the growth of its defects and/or thinning of its web. Currently the pipeline industry is using such methods as electric resistivity probes (ERP), corrosion samples (CS) and weight loss coupons (WLC) to define the CR for pipelines which transport extremely corrosive substances, or are located in a corrosive environment. Additionally, inhibitors are used to bring the CR to an acceptable level. In this setting the most reliable methods which permit assessment of CRs with needed accuracy and consistency, are probabilistic methods. The paper describes a practical method of predicting the probabilistic growth of the defect parameters using the readings of separated in time different DA or ILI measurements, using the two-level control policy [1]. The procedure of constructing the probability density functions (PDFs) of the defect parameters as functions of time, linear/nonlinear CR growth, and the initial size of the defects is presented. Their use when creating an RBI plan and IMP based on time dependent reliability of pipelines with defects is demonstrated in two illustrative cases — a production pipeline carrying crude oil, and a pipeline subject to internal CO₂ corrosion.