AmericanChemicalSociety.com

In the past year our group (the PI, graduate student Isolde Belien, and undergraduate student Kristen Fauria) has made significant progress on two main fronts. 1. We have developed a preliminary model for the conditions under which hydrate nodules form by displacing sediment particles. The model calculations are being tested and refined and we anticipate publication of these results in the coming year. 2. We have performed a series of analogue experiments to examine free gas transport through liquid infiltrated porous media. Our experimental constraints on bubble population dynamics and related modeling efforts will be the subject of an upcoming presentation at the fall meeting of the American Geophysical Union. These new developments are described briefly below.

1. Hydrate nodules are observed to form preferentially in the coarse-grained fraction of marine sediments. These centimeter to decimeter scale lumps of solid hydrate are orders of magnitude larger than adjacent sediment pores. This morphology suggests a similar formation process to that dictating the development of segregated ice lenses in subaerial environments.

Previous treatments of the formation of hydrates in porous media have focussed on the importance of interfacial curvature in adjusting the phase equilibria through the Gibbs-Thomson effect. Our new model is based upon a more complete treatment of the phase interactions and highlights the important role of interfacial liquid films in both transporting the dissolved guest species that are necessary for hydrate growth and facilitating the sediment displacement necessary to accommodate nodule development. As in conventional models of disseminated hydrate growth, nodule growth rates are primarily limited by the rate that gas is supplied through diffusive transport. This contrasts with the case of ice lens growth where thermal diffusion and sediment deformation are the rate-limiting steps. A further important difference from segregated ice growth in sediments is that hydrate nodules are generally expected to be unconnected whereas ice lenses are understood to often nucleate within a zone that is already partially frozen. The Gibbs-Thomson effect favors hydrate nucleation in the coarse-grained sediment fraction and further growth tends to occur on previously existing nodules, with a coarsening process also favoring those nodules that are bounded by larger pore aperatures. We find that the development of complete hydrate lenses is only expected in regions where the effective stress is low - such as occurs near the seafloor and nearby regions of focussed fluid flow.

Model refinements are still being made and we are currently re-examining published field results to identify supporting or contradictory data sets and better constrain model parameters. We hope to present our results at the upcoming Gordon Research Conference on Natural Gas Hydrates in June, 2010 and anticipate submitting our first model publication beforehand.

2. The transport of free gas through marine sediments can both enable rapid hydrate formation and control the rate at which dissociation supplies greenhouse gases to the water column. Early models for the development of natural hydrate deposits assumed that gas transport was limited by the concentrations that could be dissolved in the pore liquid. This results in predicted hydrate accumulation rates that are slow in comparison to Milankovitch cycles. Observed seafloor gas seeps and seismic evidence for through-going free gas concentrations within the hydrate stability zone suggest the intriguing possibility that bubble transport may be significant.

Using analogue experiments in a hele-shaw cell in our lab, we have examined the manner in which bubble populations transit sediment systems. We identify two regimes of bubble behavior, one that accommodates steady rates of gas transport, and one in which episodic transport is favored.

In the first regime, interactions with sediment particles cause bubble populations to evolve with break-up and coalescence both occurring and leading to size distributions that are skewed towards those of the matrix sediment particles. Bubble rise speeds are slow in comparison to the Stoke's settling velocity. Our results suggest that in this regime, bubble flux through the hydrate stability field can be limited by hydrate nucleation and growth. Under special circumstances, this bubble transport may facilitate rapid hydrate accumulation within marine sediments.

In the second regime, gas is inhibited from infiltrating a stationary sediment pack until the bubble size achieves some critical level. The gas then erupts through the sediment layer in a rapid series of catastrophic `burps' that cause significant particle displacements at high Reynolds numbers. We are currently running further experiments to examine what controls the onset of burping behavior and how this might influence the predicted response of hydrated sediments to the dissociation of nodules and lenses.