Evaluation of predictive assessment reliability on corroded transmission pipelines

Oil and Gas operators have to deal with the ageing process of their transmission pipeline grid. Some of these pipelines can be inspected using In Line Inspection (ILI) tools. In order to maintain an acceptable integrity level, re-inspection operations have to be performed. This process needs to be optimized in terms of resources and cost. CRIGEN, GDF SUEZ Research and Development Center, has developed a methodology which prioritizes rehabilitation operations on a pipeline after in-line inspections, and determines the optimal interval for re-inspection. A reliable help decision software tool, GADPro, which applies the methodology has also been developed. Dealing with defects assimilated to external electrochemical corrosion, the developed methodology is based on: ILI features reported in depth dimensions in order to assess a probable corrosion growth rate, Probabilistic distribution of input parameters (defect geometrical characteristics, pipeline characteristics and corrosion growth rate), Failure probability calculation. The calculation results take the form of three probabilities of failure: A punctual probability of failure for each defect, An annual probability of failure for each defect, An annual probability of failure per kilometer of pipe. To interpret the results, the annual probability of failure per kilometer of pipe is then compared with threshold values from safety studies that can be associated with location places and failure modes. In year 2009, the re-inspection of two 24 inches coil tar enamel gas transmission pipeline formerly inspected in year 1999 gave the opportunity to evaluate the consistency of the predictive assessment from GADPRO tool. The first pipeline was 125 km long. The second one was 93 km long. To reach that goal, a first step consisted to match the metal losses from the two run for each pipe section. Then, corrosion growth rates were calculated from the sizing evolution of the matched defect from the two ILI runs. Results were compared with the corrosion growth rate determined with the assumptions made in GADPro tool. Finally, the preventive intervention program which could have been determined after the 1999 ILI was compared to the one established using the results of the inspection performed on the same line in year 2009. This article presents the procedures employed to carry out a probabilistic assessment of corroded pipelines inspected with ILI tools as well as the methodology validation study based on pipeline re-inspection data analysis. (C) 2011 Elsevier B.V. All rights reserved.