Hydrate and sediment mixtures, thermal property results: sII Tetrahydrofuran hydrate + quartz sediment

Motivation:

Results from the methane hydrate and quartz sand mixtures indicate hydrate bridging intergranular contacts can improve heat transport in systems containing free gas [Waite et al., 2002]. Most naturally-occurring hydrate reservoirs contain water and hydrate rather than gas and hydrate in the pore space, however, so it is of interest to examine the thermal properties of a hydrate-bearing, gas free sediment. Of particular interest is determining what the dominant controls on thermal properties are: is heat flow controlled primarily by porosity, or by the number of intergranular contacts between high-conductivity grains? Porosity control implies the differing thermal properties of sediment and the pore fill combine in some manner controlled by their relative volumes to determine the bulk thermal properties. If heat flow is limited by poor thermal contact between grains, the number of intergranular contacts may significantly influence the bulk thermal properties. As the grain size decreases, the number of intergranular contacts increases, so samples were made covering four different grain size ranges.

Sample Preparation:

For each sample, the THF + water mixture was prepared as described for the pure THF hydrate. The sample containing silt-sized granular quartz was prepared by adding small increments of the THF + water solution into a re-sealable plastic bag containing silt. When the final mixture ratios were obtained, a corner of the bag was cut, and the material was squeezed into the sample chamber. The sample porosity, Φ, was 51.0 ± 0.5%. The three sand-sized quartz samples contained Ottawa sand, sieved to obtain three different grain size ranges: 125 - 250 µm ("small sand," Φ = 42.4 ± 0.4%), 250 - 500 µm ("medium sand," Φ = 40.5 ± 0.4%), and 500 µm - 1 mm ("large sand," Φ = 40.5 ± 0.4%). Ottawa sand samples were prepared by slowly adding sand to the THF + water after the liquid had been poured into the sample chamber.

Following sample preparation, the pressure vessel was cooled to 0.5°C, approximately 4°C below the 1 atm stability temperature for THF hydrate, in a temperature-controlled bath capable of holding temperature constant within ±0.01°C. The onset of exothermic hydrate formation was noted by an increase in sample temperature to nearly 4.4°C as measured by a thermistor within the thermal conductivity probe. Hydrate formation was completed in under 12 hours, after which the internal sample temperature re-equilibrated with the 0.5°C bath. To check for the presence of unreacted water, the sample was then cooled to -10°C, while the internal temperature was monitored for an increase from heat liberated during the ice freezing process. The confining pressure fluid volume was also monitored for signs of a sample volume increase caused by water freezing to ice. For the samples described here, no indications of ice were observed.

Measurement Results:

As with the pure THF hydrate experiments, nonlinear thermal property behavior was observed in the THF hydrate + sediment samples for temperatures above -7.5°C. We confine our discussion here to results measured between -25 and -7.5°C. The full nonlinear results are shown in Waite et al. [2005].

A dependence on porosity is apparent for thermal conductivity. Measurement uncertainties obscure the effect for thermal diffusivity and specific heat, but the high-porosity silt sample certainly differs from the sand samples. If porosity does indeed control the thermal properties in gas-free, hydrate-bearing sediment, we should be able to model our measured results with a porosity-based mixing law that combines the pure-phase thermal properties of the THF hydrate and sediment in our samples.

Thermal properties in THF hydrate-bearing sediment. Thermal properties appear to be controlled by porosity, φ, rather than by grain size, which varies from silt-sized (~10 μm, black diamonds), to sand sized (125-250 μm, blue squares; 250-500 μm, green circles; 500-1000 μm, pink triangles) (Click for larger image).

Thermal Conductivity, λ: We used the mixing model of Woodside and Messmer [1961]. This is by no means the only porosity-based mixing model for thermal conductivity, and Revil [2000] has published a comprehensive model comparison, but the Woodside and Messmer model has been shown by Huang and Fan [2005] to work well in THF hydrate-saturated sediment. Given our measured porosity and thermal conductivity for pure THF hydrate, we found the Woodside and Messmer model predicts our measured thermal conductivities within our measurement uncertainty. Moreover, the prediction is made using a value for the thermal conductivity of quartz, ~5.9 W/m·K, that falls within the broad range of thermal conductivities measured for quartz sand and crystalline quartz, 5-9 W/m·K [Huang and Fan, 2005; Clauser and Huenges, 1995].

Thermal Diffusivity, κ:  Again using the Woodside and Messmer model, with our measured porosity and pure THF hydrate thermal diffusivity, we were able to predict our measured thermal diffusivity results in all samples by using (4 ± 1) x10^-6 m²/s as the thermal diffusivity of quartz. This is in excellent agreement with the published value of 4.6 x10^-6 m²/s [Clauser and Huenges, 1995].

Specific Heat, c_p: Unlike λ and κ , in which heat is transported through a material, specific heat describes the energy required to raise the temperature of a given mass of material. The specific heat of a mixture is simply an average of the specific heats of each component (THF hydrate and quartz, in the present study), weighted by the fractional mass of each component in the sample. As with λ and κ , a single value for the specific heat of quartz could be used to predict the measured specific heats for the hydrate-bearing sediments measured in this study. The value of 520 ± 100 J/kg·K is lower than the published value of 740 J/kg·K [Clauser and Huenges, 1995], but using the relationship between λ, κ , c_p and density, our model values for λ and c_p yield κ = (4.2±1) x10^-6 m²/s, in agreement with the published value of 4.6 x10^-6 m²/s [Clauser and Huenges, 1995].

Implications:

Model agreement for each sample, obtained by varying only the porosity, suggests porosity does indeed provide the dominant control on thermal properties in gas-free, hydrate-bearing sediment. If grain size, or alternately, the intergranular contacts, control the thermal properties of the bulk material, the control is too weak to be observe in our measurements.

Publication:

The thermal property results for THF + sediment mixtures was published in Waite, W.F., Gilbert L.Y., Winters, W.J. and Mason, D.H., 2005, Thermal property measurements in Tetrahydrofuran (THF) hydrate and hydrate-bearing sediment between -25 and +4°C, and their application to methane hydrate, Paper 5042 in: Fifth International Conference on Gas Hydrates, Trondheim, Norway, Tapir Academic Press, vol. 5, 1724-1733.

Measurement Data:

Our thermal property measurements for THF hydrate can be downloaded as a tab-delimited text file.

Top of Page