Skip to main content

Discrete Multi-physics modeling of hydrate blockage in pipelines

Periodic Reporting for period 1 - HYDROBLOCK (Discrete Multi-physics modeling of hydrate blockage in pipelines)

Reporting period: 2020-03-02 to 2021-09-01

The project aimed to model and explore the phenomenon of clogging and blockage of pipelines caused by the formation of gas hydrates. Gas hydrates are crystalline compounds where a cage-like water structure surrounds an organic molecule. Under the right conditions, gas hydrates undergo a segregated solidification process that eventually builds up into large regions of solid agglomerate deposits. This phenomenon presents a challenge in the energy sector concerning flow assurance. In fact, this is a serious problem as when this occurs, the blocked section of pipe must be cut out and replaced, resulting in long production stoppages, and high costs associated with repairs and shutdowns of production. In addition, gas hydrates can form in many offshore energy processes, and the unwanted formation of hydrates clogging can produce large negative environmental consequences.

Several software packages have been developed and used by the industry for predicting hydrate formation. These packages, however, are only partially effective since they only focus on the thermodynamics, and omit complex multiphysical phenomena. Traditional multiphysics, however, has difficulties handling phenomena that involve phase transitions and, in general, attempts to account for the formation of solid aggregates within fluids are based on ‘numerical tricks’ aiming at mimicking phase change. At the University of Birmingham a new multiphysics technique, called Discrete Multiphysics (DMP) has been developed to overcome this or equivalent limitations. By employing this new technique, HYDROBLOCK have developed an initial approach to address some of the unresolved challenges in modelling of hydrate blockage and contributed towards new predictive tools that allow to properly explore the gas hydrate formation, growth and blockage.

The preliminary computational tools developed in the project were applied to modelling viscoelastic fluids, an analogous condition to the gas hydrates clogging, and to mimicking the behaviour of substances that might be considered as gels. The computational implementation has already proved to be a practical modelling tool that allows to emulate the flow of substances that exhibit some level of elasticity. Additionally, a computational model focusing on the behaviour of turbulent particle-laden flows was implemented and employed to explore situations where solid particles were transported in similar conditions to those found in the gas pipelines. This set of preliminary computational tools are providing the framework for implementation of a more complete and advanced hydrates clogging analysis tool incorporating the complex and multiphysics character of the phenomenon, and that can be used as a predictive tool by the industry, and as an in-deep analysis methodological groundwork in academia and research.
Three main work packages were proposed to perform and accomplish the objectives of the action. In the first work package an implementation of the Discrete Multiphysics approach (DMP) using linear additive composition of energy potentials was set-up as an initial model for laminar flows of Newtonian and Viscoelastic substances. This involved performing a detailed literature review and gaining knowledge on the use and capabilities of the open-source software selected (LAMMPS) and employed for the main computational implementation.The proposed approach was employed as an alternative to modelling viscoelasticity directly into Smooth Particle Hydrodynamics (SPH). The initial model was focused on exploring complex viscoelastic stress-strain relationships that might appear in flows carrying solid agglomerates, like those dominating some aspects of the gas hydrates formation.

In the second work package a preliminary computational model was developed to explore the simulation process of turbulent particle-laden flows employing a Lagrangian-Eulerian methodology. The preliminary mathematical/numerical model was based on a conventional finite volume code (OpenFOAM), together with a point-particle approximation where the fluid continuity and momentum equations were solved on a Eulerian grid, whereas the particles’ motion was modelled using Newton’s law thus following a conventional Lagrangian approach. The preliminary model of a turbulent internal flow was based on the small-scale dynamics associated with the turbulent fluctuations and modelled using a conventional SGS-LES approach. The consideration of turbulence modelling helped to establish the basis for the adaptation of the DMP model, developed in the first work package, to a version that incorporates the hydrodynamic response of the solid hydrate particles and the collision/agglomeration of hydrate crystals with themselves and with the walls.

Finally, in the third stage of the project (Work Package 3), the preliminary computational prediction tool implemented in LAMMPS, and based on the integrated DMP model, was tested for a range of different conditions of flow and stress/strain responses through modulation of the energy potentials used to represent elastic attraction with viscous dissipation. Additionally, validation tests focusing on the consistency and performance of the numerical model for simulating particles-laden turbulent flow, were conducted at different Reynolds numbers, varying the particle mass fraction, and maintaining the volumetric fraction constant. The tests were carried out using a two-way coupling model as first approach to the particle-fluid interaction.

The results of the project include the development of a set of preliminary computational tools that will provide the framework for implementation of a more complete and advanced hydrates clogging analysis tool which will be able to deal with the complex and multiphysics character of the gas hydrates formation, transport and agglomeration phenomenon. The numerical tools and the set of experiments obtained in the present project can be used as a preliminary predictive tool by the industry, and as an in-deep analysis methodological groundwork in academia and research. For instance, the application of the DMP model based on additive composition of energy potentials for modelling viscoelastic behaviour was presented in a manuscript published in ChemEngineering (an Open Access MDPI Journal). Equally, the model focusing on the particles dynamics in turbulent conditions was employed to explore the effect of varying the particle mass fraction in a generalized channel flow, a phenomenon linked with the hydrates transport, agglomeration and clogging.
The model developed in the project, consisting of a series of modular methodological formulations, will allow to explore in further detail the complex and rich dynamics dominating the gas hydrates formation, transport, and agglomeration. A fully integrated version of the model is expected to be achieved soon, as the main numerical tools have been developed, tested, and validated. The computational tools developed in the project are already providing a fully functional framework for a more complete and advanced hydrates clogging analysis tool that considers the multi-physics character of the phenomenon. It is expected that, once future additional dissemination and outreaching activities take place, the modelling framework will have an important positive impact in the area of prevention of gas hydrates clogging.
Flow response of a Bingham-type flow modelled using DMP
General setup for testing the initial model focusing on stress-strain response in pipe flow
General setup for testing the initial model focusing on stress-strain response of non-newtonian flow