Numerical Modelling of the Rapid Depressurisation and Outflow of High Pressure Containments in the Framework of Carbon Capture and Sequestration

The internationally agreed global climate deal reached at the Paris Climate Conference (COP21) in December 2015 is intended to limit the increase in global average temperature to less than 2°C above pre-industrial levels by 2050. Achieving this goal requires a 50 – 80% reduction in CO2 emissions. Alongside renewable energy sources, CO2 Capture and Sequestration (CCS) is widely considered as a key technology for meeting this target, potentially reducing the cost of inaction by some $2 trillion over the next 40 years. It is estimated that transporting the predicted 2.3 - 9.2 Gt of captured CO2 to its point of storage will require the use of a global network of between 95000 - 550000 km pipelines by 2050. The economic pipeline transportation of such large amounts of CO2 will require operation in dense or supercritical phase. In Europe, this will likely mean pipelines at line pressures above 100 bar, some passing through or near populated areas. Given that CO2 is increasingly toxic at concentrations higher than 7%, the safe operation of CO2 pipelines is of great importance and indeed pivotal to the public acceptability of CCS as a viable means for tackling the impact of global warming. The accurate prediction of the discharge rate of the escaping inventory in the event of accidental pipeline rupture is central to the safety assessment of such pipelines. This information forms the basis for determining the minimum safe distances to populated areas, emergency response planning and the optimum spacing of isolation valves. In addition, in an emergency situation, the controlled depressurisation of CO2 pipeline is critically important given the unusually high Joule-Thomson cooling of CO2. Too rapid depressurisation poses the risk of embrittlement of the pipe wall causing pipeline running fracture, solid CO2 formation leading to blockage of pressure relief valves in the event of crossing the triple point temperature (216.7 K) or a Boiling Liquid Expanding Vapour Explosion (BLEVE) due to the superheating of liquid phase CO2. This thesis presents the development, testing and validation of various transient flow models taking account of the above phenomena. These include a Homogeneous Equilibrium Mixture (HEM) pipe flow model, a Homogeneous Relaxation Mixture (HRM) pipe flow model, a Two-Fluid Mixture (TFM) pipe flow model, and an integral jet expansion model. The HEM model employing Computational Fluid Dynamics (CFD) techniques is developed for predicting solid CO2 formation during pipeline decompression. The pertinent vapour-liquid or vapour-liquid-solid multi-phase flow is modelled by assuming homogeneous equilibrium. The flow model is validated against pressure and temperature data recorded during the Full Bore Rupture (FBR) decompression of an extensively instrumented 144 m long, 150 mm i.d. CO2 pipe initially at 5.25 °C and 153.3 bar. For the conditions tested, the simulated results indicate CO2 solid mass fractions as high as 35% at the release end, whose magnitude gradually decreases with distance towards the pipe intact end. Turning to the HRM model, in its development, thermodynamic non-equilibrium between the constituent fluid phases during pipeline decompression is considered for both pure fluids and multi-component mixtures. The validation of the HRM model is carried out by comparing its predictions of a number of CO2-rich mixtures pipeline FBR decompression experiments against the corresponding measurements. For reference, the HEM model predictions (where thermodynamic non-equilibrium is ignored) for the same tests are also included in the comparison. The results show that improved agreement with the measured data can be obtained by the present model as compared to the HEM model. The last pipeline decompression model presented in this thesis is the TFM model, where the conservation equations are solved separately for each constituent fluid phases during decompression, unlike in the case of the HEM and HRM models. Fluid/fluid interface interactions are accounted for and modelled using appropriate closure relations. Furthermore, a new puncture outflow boundary condition is presented. For the numerical solution of the conservation equations, modifications towards previous schemes are introduced for improved accuracy and numerical stability. Model validation is carried out by comparing its predictions of two CO2 pipeline puncture decompression tests against the corresponding measurements, showing excellent agreement. The experimentally observed heterogeneous flow behaviour, that is, the significant temperature difference between the vapour and liquid phases, is captured by the present model. The final part of this thesis deals with the accurate prediction of the conditions of a pressurised jet upon its expansion to atmospheric pressure, where the simulated outflow data from the decompression flow models is used as the input conditions. Such prediction is of fundamental importance in assessing the consequences associated with accidental releases of hazardous fluids from pressurised vessels and pipes. An integral jet expansion model which for the first accounts for turbulence generation is presented. By the use of accidental release of two-phase CO2 from a pressurised vessel as an example, the proposed model is shown to provide far better predictions of the fully expanded jet momentum flux as compared to the existing integral model where the impact of turbulence generation is ignored