1

Assessment of fracture propagation in pipelines transporting impure CO2 streams

S. Martynov1, R. H. Talemi2, S. Brown1,3,H. Mahgerefteh (CO2QUEST project coordinator)1*

1Department of Chemical Engineering, University College London, London, WC1E7JE2ArcelorMittal Global R&D Gent OCAS N.V., Pres. J.F. Kennedylaan 3, 9060 Zelzate, Belgium

3Present address: Department of Chemical and Biological Engineering, The University of Sheffield,S1 3JD, UK

Corresponding author e-mail address: h.mahgerefteh@ucl.ac.uk

Abstract

It is widely accepted that economically viable long-distance transportation of carbon dioxide (CO2)in Carbon Capture and Sequestration projects can be achieved using pipelines transmitting dense-phase CO2 fluid at pressures in excess of ca. 90-150 bar [1, 2]. Given that such pipelines are likelyto pass close to residential areas, their accidental failure may result in release of significantamount of toxic inventory posing potential risks of asphyxiation by toxic CO2 cloud. Of particularconcern for CO2 pipelines are long running fractures, damaging long pipeline sections and resultingwith rapid releases of large amount of CO2 fluid. In order to minimise the chances of CO2 pipelinefailure, the pipeline material, diameter and wall thickness, as well as measures minimising theconsequences of the failure (e.g. placement of the pipeline crack arrestors) are carefully selectedin the design relying on predictions using mathematical models of fracture propagation and arrest.

In particular, ductile fracture control is considered in the design of dense-phase CO2 pipelines [3,4]. Although several methods have been applied to assess the CO2 pipelines propensity to ductilefractures [5, 6], most of them are based on simplified empirical models, such as Battelle theory [7],which are not capable of resolving accurately the outflow and the crack propagation as coupledphenomena, and may not give conservative estimates for CO2 pipeline fracture toughness [8, 9].As such, more rigorous theoretically substantiated computational models of fracture propagationare urgently needed.

While brittle fractures are not of concern for modern gas transmission pipelines, it has recentlybeen suggested that unusually high Joule-Thomson coefficient of CO2 may induce lowtemperatures in the pipe upon the CO2 fluid expansion to atmospheric pressure [10, 11]. In theturn this may lead to loss of ductility by the pipe wall material, increasing the risks of brittlefracture propagation. Although models have been developed to predict both the mechanics ofbrittle crack in steels and leaks from pipelines, they have not been applied yet to assess thepropensity of CO2 pipelines to brittle fractures.

In the present study, performed in the course of CO2QUEST project (http://www.co2quest.eu/), acomputational fluid-structure interaction methodology is developed to resolve accurately the fluid

2

flow in the pipeline with a moving crack. In this methodology, a computational fluid dynamicsmodel is applied to simulate transient one-dimensional flow of a homogeneous vapour-liquidmixture in a pipeline, where an advection equation is applied to track the position of the crack tip.In the turn, the crack tip velocity is predicted using material damage models implemented in athree-dimensional finite element code ABAQUS. In particular, to simulate ductile fractures amodified Bai-Wierzbicki model is applied [12], while to model dynamic brittle fracture behaviour anovel approach based on an eXtended Finite Element Method technique is used [13]. Moreover,to evaluate the temperature distribution in the pipe wall in proximity to the crack, a transientthree-dimensional heat-transfer model accounting for the secondary cooling of the pipe wall dueto the expansion of escaping CO2.

The developed methodology is validated against measurements of fracture propagation in large-scale pipelines transporting natural gas (Figure 1) and compared with predictions using the ductilefracture model using Battelle theory [7]. The brittle fracture has been calibrated using the CharpyV-notch and Drop Weight Tear tests data obtained for X70 pipeline steel. The methodologydeveloped is applied to predict the ductile and brittle fracture propagation in pipelinestransporting impure CO2 streams captured typical for post-combustion, pre-combustion and oxy-fuel capture technologies. Based on the results of simulation of pipeline rupture scenarios forvarious CO2 transportation conditions, the propensity of CO2 pipelines to ductile and brittlefracture is discussed.

(a) (b)

Figure 1. Fracture appearance in a natural gas burst test [14] (a) and predicted using the coupledfracture propagation model developed in the present study (b).

References

1. Doctor, R. et al. (2005). Transport of CO2. IPCC Special Report on Carbon Dioxide Captureand Storage.

2. Mohitpour, M., Jenkins, A., Nahas, G. (2008). CO2 and greenhouse gas emissions, 237–251.3. Bilio, M., Brown, S., Fairweather, M., & Mahgerefteh, H. (2009). CO2 pipelines material and

safety consideration. IChemE Symposium. In HAZARDS XXI Symposium Series (pp. 423–429).

4. DNV (2010). Design and Operation of CO2 Pipelines . DNV-RP-J202 (pp. 1–42).5. Cosham, A., & Eiber, R. (2008). Fracture propagation in CO2 pipelies. Journal of Pipeline

Engineering, 281–292.

3

6. Mahgerefteh, H., Brown, S., & Denton, G. (2012). Modelling the impact of streamimpurities on ductile fractures in CO 2 pipelines. Chemical Engineering Science, 74, 200–210.

7. Inoue, T., Makino, H., Endo, S., Kubo, T., & Matsumoto, T. (2003). Simulation method forshear fracture propagation in natural gas, 5, 121–128.

8. Aursand, E., Dørum, C., Hammer, M., Morin, A., & Munkejord, S. (2014). CO2 pipelineintegrity: Comparison of a coupled fluid-structure model and uncoupled two-curvemethods. Energy Procedia, 00, 1–10.

9. Cosham, A., Jones, D. G., Armstrong, K., Allason, D., & Barnett, J. (2014). Analysis of TwoDense Phase Carbon Dioxide Full-Scale Fracture. In 10th International Pipeline Conference(pp. 1–15). Calgary, Canada: ASME.

10. Mahgerefteh, H., & Atti, O. (2006). Modeling low-temperature-induced failure ofpressurized pipelines. AIChE Journal, 52(3), 1248–1256.

11. Cosham, A., Koers, R., Andrews, R., & Schmidt, T. (2015). Progress towards the new eprgrecommendation for crack arrest toughness for high strength line pipe steels. In 20th JTM,3-8 May 2015, Paris, France.

12. Bai, Y., & Wierzbicki, T. (2009). Application of extended Mohr–Coulomb criterion to ductilefracture. International Journal of Fracture, 161(1), 1–20.

13. Hojjati-Talemi, R., Cooreman, S., & Van Hoecke, D. (2016). Finite element simulation ofdynamic brittle fracture in pipeline steel: A XFEM-based cohesive zone approach.Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Designand Applications, 1–14.

14. Makino, H., Takeuchi, I., Tsukamoto, M., & Kawaguchi, Y. (2001). Study on the PropagatingShear Fracture in High Strength Line Pipes by Partial-gas Burst Test. ISIJ International, 41(7),788–794.