Comparison of Experimental, Numerical, and Field Data for the Dynamics of Natural Gas Bubbles in the Hydrate Stability Zone of the Ocean Water Column

Abstract

Mass transfer between bubbles and the surrounding ambient water is an important process in subsea blowouts and for predicting the fate of seepage from natural gas sources at the seabed. To understand the fundamental behavior of natural gas bubbles in the ocean environment, the National Energy Technology Laboratory (NETL) conducted experiments for methane bubbles and bubbles of gas mixtures of methane, ethane, and propane in their high-pressure water tunnel (HPWT) un-der different thermodynamic conditions. For each experiment, individual bubble behavior was recorded to measure the bubble dissolution rate and observe the behavior of non-hydrate and hydrate armored bubbles during their release and transport under simulated ocean conditions. In this dissertation, we analyze the complete dataset of the NETL HPWT experiments for methane and natural gas and conduct new experiments in the deep ocean to observe natural seeps on the Cascadia Margin. To both analyze the experimental data and extend the NETL results to the field scale, a numerical model, the Texas A&M Oil spill / outfall Calculator (TAMOC), was also adjusted to simulate the mass transfer and bubble dynamics. Analysis of the NETL HPWT data was conducted in two phases. First, we analyzed experiments for bubbles at high pressure and low temperature, but that did not have hydrate shells due to low dissolved gas concentration in the recirculating water. We calibrated empirical models for mass transfer coefficient by comparing the theoretical mass transfer rates for clean (Johnson et al.[1969]) and dirty (Clift et al. [1978]) bubbles in quiescent ambient water with those for pure bubbles in the HPWT. With an amplification factor (α = 1.9), mass transfer rates for dirty bubble exhibit the best agreement with the HPWT experiment results: βnetl = αβemp, where βnetl is the apparent mass transfer coefficient in the HPWT and βemp is the empirical coefficient for dirty bubbles from the literature. The amplified mass transfer is probably due to the background turbulence generated by the counter-flow in the water tunnel. Utilizing the dirty bubble mass transfer rate with α = 1.9, TAMOC accurately predicts the time-evolving size of non-hydrate bubbles in the HPWT experiments with −2.0 % average relative error (µ) for the bubble size at the end of each experiment with a standard deviation (σ) of 12.6 %. Second, we analyzed the experiments with higher dissolved gas concentration and in which the bubbles became armored with hydrate. For these hydrated bubble experiments, we analyzed the shrinkage rates of each bubble and demonstrated that either the gas vapor within the bubble or the hydrate skin itself may be the principal dissolving phase under various conditions. The vapor phase is the primary phase that dissolves when the subcooling (∆T ) is less than 11 K or when the pressure is continuously decreased to simulate the bubble rising through the ocean water water column. Subcooling is defined as the difference be-tween the hydrate-vapor-water equilibrium temperature (Thyd) and the ambient temperature (Tamb), such that ∆T = Thyd − Tamb. Hydrate may be the primary dissolving phase when the pressure is maintained constant at ∆T > 11 K. In either case, the dirty bubble mass transfer rates with α = 1.9 was found to yield the appropriate mass transfer rate. TAMOC accurately predicts the time-evolving size of hydrate bubbles in the experiments with (µ = 0.4 %, σ = 5.6 %). Experiments conducted during the R/V Falkor cruise to the Cascadia Margin utilized a stereoscopic, high-speed camera system (TAMU-CAM) developed previously at Texas A&M University (Wang and Socolofsky [2015b]). The TAMU-CAM allows the time-evolving observation of bubble size and trajectory along 3-D paths. Data from these observations is used in this dissertation to extend observation and conclusions from the NETL HPWT experiments to the field scale. This comparison encompasses the hydrate’s morphology, the dynamics of the bubble interface, and the slip velocity; bubble shrinkage rate could not be observed in the field as the system was not designed in these experiments to track the bubbles over long distances (a minimum of 300 s or approximately 10 m of rise is required to obtain stable shrinkage rates). Nonetheless, these other parameters help validate the field applicability of the NETL HPWT experiments. The analysis of the three-dimensional bubble trajectory data provided by the stereoscopic TAMU-CAM also al-lows comparison of the bubble wake dynamics with the HPWT and other laboratory observations. These field observations revealed rapid hydrate formation close to the seep source, as well as similar bubble morphology and rise rates for the hydrate-coated bubbles in the field as observed in the HPWT. With these data, we contribute to the extension of HPWT result to the field scale.