Session: 07-04-01 Crack Management

Paper Number: 87136

87136 - An Integrated Model for Predicting Stress Corrosion Cracking of Buried Pipelines 

Stress corrosion cracking (SCC) is a type of crack that grows in corrosive environment, which threatens the integrity of pipeline and may lead to sudden pipeline failures. SCC predictions are difficult because SCC processes involve complex interactions between cathodic protection, electrolyte chemistry beneath the disbonded coating, metallurgy and stresses. Additionally, data uncertainty in the pipeline operation made a precise SCC prediction even more complicated. To address this issue, and integrated SCC model that combine knowledge of the various mechanisms is developed for pipeline operators. A Bayesian approach was used to consider and interconnect six mechanisms: coating disbondement, surface potential, bicarbonate equilibrium, dents and stresses, crack growth rate, and pipe resistance. The benefits of the Bayesian approach are the threefold: (1) ability to use uncertain data and identify what data would be most useful, (2) consider the effects of uncertain relationships between mechanisms and (3) the graphical nature of the approach allows for easy communication of the risks to decision makers. The model predicts SCC crack growth rate and crack depth as a function of location and time.

The Bayesian SCC model presented in this paper combines literature data, models, and subject matter expert’s knowledge. The inputs to the different sub-models, and their outputs are discussed as well as the approach adopted in integrating them to predict SCC. These four sub-models include: coating disbondment model evaluating the forces experienced by the coating from soil forces and cathodic protection current; surface potential model evaluating a range of possible surface potentials under the disbonded coating; carbonate/bicarbonate concentrations model using soil properties, weather condition, and pipeline temperature to predict the bicarbonate concentration; and SCC type model to predict the likelihood of SCC type (i.e. high pH SCC or near neutral pH SCC). The results of the work also highlight some of the existing gaps and challenges that exist.

Presenting Author: Guanlan Liu DNV

Presenting Author Biography: Dr. Liu is a Senior Engineer working for DNV Pipeline Services. He has over 12 years of combined experience in chemical engineering research and oil/gas industry about material performance and novel mathematical methods to quantify risk. His main research interest involves pipeline risk assessment, material durability, etc. He has published more than 20 papers on diverse subjects such as risk and reliability modeling, corrosion modeling, reaction kinetics and materials degradation, etc.