UCL Department of Chemical Engineering


Songsong Liu

Post-Doctoral Researcher


Research Project

Title: Ductile and brittle fractures in CO2 pipelines

Nora's research is part of the CO2Quest project: “Impact of the quality of CO2 on transport and storage” and focuses on ductile and brittle fractures in CO2 pipelines for transport of CO2 with impurities. Pollution with CO2 emission from firms is one of the main factors responsible for impacting on the climate and causing the ‘global warming’. There is a crucial need to store CO2 in order to minimise its emission in the atmosphere. Carbon capture and storage (CCS) has been deployed to reduce CO2 emission. As part of the CCS chain, 150 bar pressurised pipelines made of steel are used for transporting dense phase of CO2 captured from plants before been injected into a geological formation. Running fractures, considered as catastrophic pipeline failures, can initiate from defects in the pipe, due to mechanical damage, soil movement, corrosion, material defects, etc… It is highly desirable to design the pipeline in such a way that the defect leads to a leak rather than a long running fracture.

How evolve the crack into a ductile or brittle fracture depends on the fracture toughness and the temperature of the material. It is important to take into account the temperature drop in the pipeline during discharge. This temperature can reach -70°C and result to a global pipe/wall heat transfer in the case of a ductile fracture, and localised heat transfer effects in the case of brittle fractures. No localised heat transfer of the crack tip will be observed in the case of a ductile fracture because of the very rapid opening of the crack happening before the local cooling.

CO2 pipelines may be more subject to fracture propagation in comparison with hydrocarbon pipelines (Mahgerefteh et al., 2010). CO2 is rarely pure but contains impurities. The impact of impurities is one major factor that may modify the CO2 depressurisation trajectory. Recent studies using equations of states show that the presence of small amounts of impurities will cause a significant increase in the saturation pressure (Oosterkamp and Ramsen, 2008). The capture technologies (pre-combustion, post-combustion and oxy-fuel) and the post-capture processing are responsible for the type of impurities associated with the CO2.

In order to investigate the behaviour of fracture propagation, a CFD based fluid/structure fracture model is implemented. This involves the development of a transient fluid flow model, a fracture model, their coupling and the validation of the model using experimentation. The transient fluid flow model contains an implicit subroutine. The parameters of interest such as the transient pressure in the pipeline are calculated and this is passed to the Abaqus model fracture through this subroutine, the crack path is then determined, an output file containing the information of the crack path is then used as input for the transient fluid flow model and this is repeated until the crack reaches its final location. The Finite Element Analysis Commercial Software Abaqus 6.14 is used to study the crack propagation in CO2 pipeline. Extended Finite Element Method (XFEM) is employed to model the propagation of discontinuities such as cracks. XFEM is an extension to the Finite Element Method: the approximation space is extended with enriched finite element method so that it is able to naturally reproduce the challenging feature associated with the problem of interest. The path of the crack front as illustrated at several instant in Table 1 is determined using a level set method. The scalar value ‘Philsm’ obtained is used in the post processing step. The implicit dynamic fracture model is then coupled with a transient fluid flow model through the subroutine mentioned above which provides the information about the transient pressure in the pipe. A python program is implemented in order to extract the path of the dynamic crack through the circular 604 millimetre diameter CO2 pipeline and these data are post processed so that the crack speed is determined. Experimental data are used in the purpose to test the fluid flow/structure model and validate the model. 

Research interests

  • Computer aided process engineering
  • Supply chain management
  • Production planning and scheduling
  • Optimisation