Hydrate and sediment mixtures, thermal property results: sI methane hydrate + methane pore gas + quartz sand

Motivation:

In gas-rich, hydrate-bearing sediment, like that found on Cascadia's convergent margin [Bohrmann et al., 1998], it is of interest to understand the role of methane hydrate in determining thermal properties of hydrate-bearing sediment. Unlike water-saturated, hydrate-bearing sediment, in which thermal conductivity is nearly identical in the water and hydrate, measurements from Vargaftik et al. [1993] show methane gas has a much lower thermal conductivity than does methane hydrate. This leads to an interesting dependence of thermal conductivity on the pore space hydrate saturation that is not observed in gas free systems such as the THF hydrate + sediment system.

Sample Preparation:

Thermal conductivity dependence on methane pore gas pressure at -10°C. Blue circles represent λ in each of the four hydrate + quartz mixtures. Red diamonds are for λ of methane gas [Vargaftik et al., 1993]. The strong dependence of λ on pressure for methane gas dominates our measured pressure dependence of λ for hydrate + quartz sand + methane pore gas mixtures (Click for larger picture).

We formed methane hydrate directly in the thermal conductivity measurement chamber by slowly heating granular (180-250 µm) water ice in a pressurized methane atmosphere, as described by Stern et al. [1996; 1998]. Careful testing of gas volumes liberated during sample dissociation showed this highly reproducible procedure produced methane, CH₄, hydrate with a bulk composition of CH₄·nH₂0, where n = 5.89 ± 0.01 [Stern, 2000]. Within individual synthesis runs, we checked for full reaction of water ice to methane hydrate by lowering the bath temperature below the freezing point of water. If more than 1% of the initial seed ice converted to and remained as water in our sample throughout the synthesis process, the gas pressure transducer would detect the pore gas pressure increase due to unreacted water expanding while freezing. Heat liberated during freezing provided an additional indicator of unreacted water in the neighborhood of the thermistor bead. The absence of volumetric and thermal freezing anomalies in both the pressure and temperature records was therefore required before thermal conductivity measurements were made.

Measurement Results:

These measurements were made in the early stages of our thermal property investigations, and did not utilize a data acquisition system fast enough to acquire meaningful thermal diffusivity or specific heat data. The uncertainties in thermal conductivity are also larger than in subsequent datasets, in part because of the data acquisition rate, and in part because of temperature fluctuations in the bath surrounding the sample. In our early experiments, the bath was mixed via convection alone rather than by active mixing with a pump. As a result, the bath temperature fluctuated enough to cause uncertainties of ±10% in the thermal conductivity results.

Thermal conductivity dependence on the hydrate content in the solid fraction of the sample. The most efficient heat transport in our sample set occurs when the solid portion of the sample is 33% hydrate, for which increased inter-granular heat flow provided by hydrate cementation between quartz grains outweighs hydrate's relatively low thermal conductivity. Error bars give the range of results obtained as a function of gas pressure shown previously.(Click for larger picture)

Dependence of thermal conductivity, λ, on methane pore pressure: We measured the dependence of λ on methane pore pressure over the range 3.5 to 27.6 MPa at a constant temperature of ‑10°C. Thermal conductivity increased with increasing gas pressure for all four sample compositions. Because λ in the solid components of recovered sediment cores is only expected to rise 0.25% per 10 MPa [Ratcliffe, 1960], and the pressure dependence of λ in methane hydrate is weak [Waite et al, 2007], we attributed the observed pressure dependence to the conductivity of CH₄ gas, which increased by nearly a factor of three over our measured pressure range [Vargaftik, et al., 1993].

Dependence of thermal conductivity, λ, on hydrate concentration: The thermal conductivity's peaked shape as a function of hydrate concentration likely represents a competition between inter- and intra-granular heat flow. Despite the high conductivity of quartz, heat transport may have been relatively inefficient between sand grains because small, rough intergranular contact areas and the insulating effect of low conductivity methane pore gas hindered intergranular heat transfer. Methane hydrate can raise λ by bridging intergranular contacts, creating efficient heat flow paths between high λ quartz grains. For samples with a larger hydrate fraction, where quartz grains have effectively been replaced by methane hydrate grains, λ was reduced because the increased efficiency of intergranular heat transport was offset by less efficient heat transport within each grain due to the low λ in methane hydrate relative to quartz.

Publication:

These results for gas-rich mixtures of methane hydrate and quartz sand are published in Waite, W.F., deMartin, B.J., Kirby, S.H., Pinkston, J., and Ruppel, C.D., 2002, Thermal conductivity measurements in porous mixtures of methane hydrate and quartz sand: Geophysical Research Letters, v. 29, p. 2229, doi:10.1029/2002GL015988.

Measurement Data:

Our thermal conductivity measurements in gas-rich, hydrate-bearing quartz sand can be downloaded as a tab-delimited text file.

Top of Page