Degree

Doctor of Philosophy (PhD)

Department

Engineering Science

Document Type

Dissertation

Abstract

The study of gas hydrates has increased significantly over the last two decades because of their potential application in energy storage, production and transportation, water desalination, refrigeration, gas separation, etc. There is a significant interest in commercially developing natural gas hydrates because the energy content of methane hydrate reservoirs (MHRs) is at least twice that of all conventional fossil fuels combined. Additionally, due to the higher storage capacity of hydrates, it is proposed to sequester CO2 in the form of CO2 hydrate, thereby reducing the amount of greenhouse gases. Hence, hydrates offer a dual solution to two of the most critical challenges of the future: escalating energy demands and global warming. Depending on the specific application of gas hydrate technologies, it is important to understand the kinetics of their formation or dissociation under certain thermodynamic driving forces. This work will focus on the study of the transient thermal dissociation of methane hydrates and the formation of carbon dioxide (CO2) hydrates.

Various researchers have performed laboratory experiments to study the kinetics of gas hydrate formation and dissociation. Although these experiments typically allow the macroscopic observation of the growth or melting of gas hydrates, they are often fraught with problems, inaccuracies, and complications that border on the control and monitoring of hydrate formation/dissociation at the high pressures and low temperatures of these gas hydrates. Additionally, experimental studies of CO2 or methane hydrate formation/dissociation at different concentrations of these gases and other impurities are tedious. Thus, only a few compositions are usually studied. To avoid these limitations, several authors have performed molecular dynamics (MD) simulations to obtain valuable insights into the thermodynamic stability and kinetics of CO2/methane hydrate formation/dissociation. However, the small scale used in previous simulations (on the order of a few nanometers and a few thousand molecules) has limited the understanding of the formation/dissociation processes.

In this study, large-scale molecular simulations were performed leveraging the coarse-grained (CG) Stillinger-Weber (SW) potential. This, combined with the deployment of the GPU-enabled LAMMPS version on GPU nodes of high-performance computers (HPC), enabled the molecular simulation of hydrate structures that are multiple orders of magnitude larger than those in most previous studies.

At the outset of this study, the transient thermal dissociation of methane hydrates was studied using non-equilibrium MD simulations. To curtail the potential superheating of the solid hydrates, the direct-coexistence method was used to generate the equilibrium configuration of the hydrate/liquid/gas mixture. Subsequently, transient thermal dissociation was simulated using the isobaric-isenthalpic ensemble while imposing temperature gradients at the boundaries of the simulation domain. This resulted in the transient melting of the hydrates in the middle of the simulation domain, effectively emulating the real-life transient heating. The kinetics of dissociation was monitored leveraging an image-processing algorithm. This allowed us to observe how the solid methane hydrate dissociated over time while maintaining a thermal gradient across the dissociation front. For the first time, we observed the formation of an unstable secondary dissociation path, which subsequently led to the formation of nanobubbles of methane gas within the solid hydrate. The kinetics of thermal dissociation appeared to occur in three stages. In the first stage, the system’s energy increased until it exceeded the activation energy, and the solid hydrate started to dissociate. The remaining solid methane hydrate continued to dissociate at a fairly consistent rate during the second stage, whereas the third stage involved the dissociation of the remaining hydrates across a non-planar and heterogeneous interface.

In the secondary phase of this research, a coarse-grained (CG) force field for CO2 hydrates was developed, honoring both experimental data and structural properties obtained from all-atom (AA) simulations. The SW parameters were tuned for CO2-CO2 and H2O-CO2 interaction terms, whereas the monoatomic water (mW) model was employed for H2O-H2O interactions. The force field domain for CO2-CO2 and H2O-CO2 interactions was thoroughly investigated through a rigorous experimental design and an extensive number of molecular simulations. The CG force field accurately reproduced the experimental data on CO2 physicochemical properties, such as enthalpy of vaporization and density along the phase transition curve. The developed force field also successfully replicated experimental data on CO2 solubility in water, as well as the enthalpy of dissociation and dissociation temperature of CO2 hydrates. The force field was then employed to conduct large-scale molecular studies of CO2 hydrate growth.

The models, methodology, and tools developed for studying the transient thermal dissociation of methane hydrates were utilized to conduct large-scale molecular dynamics (MD) simulations of CO2 hydrate formation. This approach is crucial because it leverages the substantial storage potential of gas hydrates for sequestering CO2 in the subsurface, contributing to efforts toward achieving net-zero carbon emissions. Starting from the hydrate/water/CO2 equilibrium state, isothermal–isobaric simulations were conducted at various temperatures to monitor hydrate growth within the simulation domain. The results revealed two distinct growth mechanisms depending on the imposed temperature. At temperatures near the hydrate equilibrium temperature, hydrate growth initiates and progresses at the crystal/liquid interface. Conversely, at sufficiently low temperatures, hydrate formation initiates not only at the crystal/liquid interface but also heterogeneously within the liquid phase. Under these conditions, the rate of hydrate growth surpasses the rate of CO2 diffusion into the hydrate structures. This leads to the novel observation of nanobubbles trapped within the solid hydrates, which enhances the storage capacity of the hydrates. Nanobubble entrapment was consistently observed in simulation replicates initiated from various initial conditions and in a simulation domain eight times larger than the base simulation domain. This indicates the observed phenomena are not limited to a specific realization of the initial conditions or size of the system studied and should be expected in nature. The tools, framework, and force field developed in this research can be applied to large-scale simulations of various hydrate structures and compositions governed by homogeneous or heterogeneous formation. So, the results from this work could pave the way for a comprehensive understanding of gas hydrate dissociation and formation kinetics.

Date

8-20-2024

Committee Chair

Olufemi, Olorode

Download

Share

COinS