Gas Hydrates - Part III: Rock Physics Hydrate Experiments

“In science one tries to tell people, in such a way as to be understood by everyone, something that no one ever knew before. But in poetry, it’s the exact opposite.” - Paul Dirac, Physicist, Nobel Prize in Physics, 1933
This article appeared in Vol. 9, No. 5 - 2013


Methane hydrate crystals – cryogenic SEM images taken at very low temperatures to keep the hydrate stable. Source: L. Stern / USGS In the previous issue of Geo ExPro we discussed whether gas hydrate is a part of the pore fluid or a part of the rock itself. It was shown that this matters if we want to estimate geophysical parameters for a gas hydrate rock. There are two ways (at least!) to investigate this further: either to observe geophysical parameters from wells being drilled through a hydrate-bearing rock, or to inject methane hydrate into a rock sample in the laboratory. We will discuss examples of both and put particular emphasis on a recent lab experiment performed in China.

Seismic Attenuation

Like the presence of free gas, the presence of gas hydrate affects seismic attenuation. Attenuation has the potential to map hydrate concentrations through the effect of local blanking of sediment stratigraphic reflectivity. However, there are few studies related to seismic attenuation in hydrate-bearing sediments and it remains an open topic for future studies. The attenuation of seismic energy by gas hydrates is likely to depend on the concentration of hydrate, the thickness of hydrate, the mechanism of hydrate formation, and the dominant frequency of the seismic measurements, in addition to the lithology changes.

VSP data in the Mackenzie Delta indicate that quite thick hydrate-bearing zones have significant attenuation at seismic frequencies of 10–200 Hz (Q-values, which describe inverse attenuation, of around 10).

Observations from Wells

Methane hydrate concentration in the sediments at ODP site 995 from P-wave sonic and resistivity logs. Column 1: Comparison of P-wave velocity (red line) with model results assuming methane hydrate (solid black lines) or homogeneously distributed methane gas (dashed lines) are part of the pore fluid. Column 2: Comparison with model results assuming methane hydrate is a sediment frame component (solid black lines) or methane gas is patchily distributed in the pore space (dashed lines). Column 3: Comparison of methane hydrate concentration estimates derived from the resistivity log to estimates derived from the compressional wave sonic log using the gas hydrate as sediment frame component model. Source: Helgerud, 2001 In a field example from a well drilled at the Blake Ridge (ODP site 995) offshore South Carolina, USA, Helgerud found that the rock physics model that assumes hydrate is part of the rock frame gave a reasonable fit between hydrate concentrations estimated from P-wave well log measurements and those obtained from the resistivity log. The deviation between the two models is not huge, but based on the well log observation it is clear that the model which assumes hydrate is part of the rock itself and acts as a kind of cement explains the well log data best. A relative good fit between observed saturations of hydrate and those estimated using the Archie equation (discussed in more detail below) is achieved.

Qingdao Experiment

This year, a very interesting experiment on hydrate formation and dissolution was presented by Hu and Ye of the Qingdao Institute of Marine Geology in China. They measured P- and S-wave velocities as the hydrate concentration in two rock samples was gradually increased from zero to 70%. The sediment samples were first immersed by pure water, then loaded into a high-pressure vessel, and injected with methane. The temperature was kept at 2°C for hydrate formation. To simulate the effect of hydrate dissolution, the temperature was gradually increased to room temperature.

P- and S-wave velocities versus hydrate concentration for a consolidated sand sample. Source: Hu and Ye, 2012 P- and S-wave velocities versus hydrate concentration for an unconsolidated sand sample. Source: Hu and Ye, 2012
For an unconsolidated sand sample, they found that the P-wave velocity increased from approximately 1,600 m/s for zero hydrate concentration to approximately 3,600 m/s for 70% hydrate concentration. The corresponding values for the S-wave velocity were 600 and 1,600 m/s, respectively. This means that the Vp/Vs-ratio decreases from 2.7 for no hydrate to 2.25 at 70% hydrate concentration. For comparison, it is interesting to note that Helgerud (2001) measured a Vp/Vs-ratio of approximately 1.9 for pure hydrate. A linear extrapolation of the measured Vp/Vs-ratio from Hu and Ye’s experiment yields a Vp/Vs-ratio of approximately 2.0.

For the consolidated sample, the Qingdao experiment showed, as expected, higher acoustic velocities: a P-wave velocity increase from 4,250 m/s to 4,700 m/s as the hydrate concentration is increased from 0 to 70%. The corresponding numbers for the S-wave velocity are 2,500 m/s and 2,750 m/s. These numbers correspond to a constant Vp/Vs-ratio of approximately 1.7, practically independent on hydrate concentration.

Hu and Ye also noticed a hysteresis effect for the P-wave velocities. For the unconsolidated sand sample they found that for a hydrate concentration of 50% the P-wave velocity was 2,700 m/s as hydrate was formed and only 2,000 m/s as hydrate was dissolved. The S-wave velocity showed a similar, but weaker hysteresis effect. For the consolidated sample they found that this hysteresis effect was opposite: the P-wave velocity during formation of hydrate was lower than the corresponding value for dissolution. Hu and Ye suggest that this hysteresis effect is caused by two very different mechanisms for unconsolidated and consolidated rocks. For unconsolidated rocks they suggest that the high velocity during formation is caused by hydrate cementation, and that this cementation process behaves differently during formation and dissolution. For consolidated rocks they discuss a two-stage formation process of methane hydrate: first a water-hydrate slurry is formed, followed by a slow solidification process. This two-stage process might explain that P-wave velocities are lower during formation compared to dissolution for the same hydrate concentration. These results are interesting, and might be of importance for geophysical analysis of data from hydrate-bearing sediments. A velocity difference of 700 m/s related to whether hydrate is slowly dissolving or being formed should be possible to detect. For instance, if a hydrate-bearing rock is being produced by heating the rock, hydrate will start to dissolve, and a significant change in P-wave velocity should be detectable on conventional time lapse seismic data. This depends of course on the initial concentration of hydrate within the porous rock; for low concentrations the changes in P-wave velocity will be less pronounced, as demonstrated by Hu and Ye’s experiment.

Resistivity Variations

Nigel Edwards is one of the pioneers in investigating the mapping of submarine hydrates using electromagnetic surveying methods. In 1997 he published a paper describing how seafloor transient electric dipole-dipole methods can be used to detect hydrates. He uses a simplified version of the Archie equation (see box above) to assess the effect of methane hydrate concentration (Sh) on the formation resistivity ():

Here, is the resistivity of the sea water (0.3 Ohm-meter) and denotes porosity. We clearly see from this equation that the formation resistivity increases rapidly as the hydrate concentration increases, a fact that makes methane hydrates a candidate for electromagnetic surveying and complementary to seismic surveying. Spangenberg and Kulenkampff (2006) use artificial samples (glass beads) to measure the resistivity versus hydrate concentration. They found that the resistivity increased from 5.1 Ohmm at zero percent hydrate saturation to 265 Ohmm at 95% hydrate saturation. They also investigated the accuracy of the Archie-equation given above, and found that the exponent (equal to 2 in the above equation) deviated significantly from 2, especially for hydrate saturations above 50%, where an exponent above 4 was found.

Lasse Amundsen is Chief Scientist Exploration Technology in Statoil. He is adjunct professor at the Norwegian University of Science and Technology (NTNU) and at the University of Houston, Texas. Martin Landrø is professor in Applied Geophysics at NTNU, Trondheim, Norway.

N. Edwards, 1997, On the resource evaluation of marine gas hydrate deposits using sea-floor transient electric dipole-dipole methods: Geophysics 62 63-74.

M. Helgerud, 2001, Wave speeds in gas hydrate and sediments containing gas hydrate: A laboratory and modelling study, PhD thesis, Stanford University, USA.

G. Hu and Y. Ye, 2012, Ultrasonic waves on gas hydrates experiments, Ch 5, Ultrasonic Waves, InTech.

E. Spangenberg and J. Kulenkampff, 2006, Influence of methane hydrate content on electrical sediment properties: Geophysical Research Letters 33 L24315-19.

J. Dai, H. Xu, F. Snyder, and N. Dutta, 2004, Detection and estimation of gas hydrates using rock physics and seismic inversion: Examples from the northern deepwater Gulf of Mexico: The Leading Edge 60-66.


Related Articles