GAS HYDRATE SATURATIONS

Electrical resistivity and acoustic transit-time downhole logs from Sites 994, 995, and 997 indicate the presence of gas hydrate in the depth interval between 185 and 450 mbsf on the Blake Ridge (logging Unit 2). Electrical resistivity downhole log data can be used to quantify the amount of gas hydrate in a sedimentary section as discussed in Leg 164 ODP Proceedings, Initial Reports volume (Paull, Matsumoto, Wallace, et al., 1996) and in Collett (1998). In the following section, we have used data from the electrical resistivity (DITE) logs in Holes 994D, 995B, and 997B to quantify the amount of gas hydrate in logging Unit 2 (approximate depth interval of 185-450 mbsf) on the Blake Ridge.

For the purpose of discussion we have assumed that the anomalous high resistivities and velocities measured in logging Unit 2 on the Blake Ridge are a result of the presence of in situ natural gas hydrate. However, an alternative hypothesis suggests that interstitial water salinity changes could account for the electrical resistivity log responses observed at Sites 994, 995, and 997. Geochemical analyses of cores from logging Unit 2 at all three sites on the Blake Ridge have revealed the presence of pore water with relatively low chloride concentrations (Shipboard Scientific Party, 1996a, 1996b, 1996c). This may indicate that logging Unit 2 contains waters with relatively low salt concentrations that will contribute to an increase in the measured electrical resistivities. However, because the acoustic log is not affected by changes in pore-water salinities, it appears to refute the hypothesis that salinity changes are contributing to the anomalous acoustic velocity and resistivity properties of logging Unit 2. To further evaluate the effect of pore-water salinity on the measured log values at Sites 994, 995, and 997, we have attempted to quantify the observed changes in electrical resistivity in logging Unit 2 in respect to potential pore-water salinity changes. We have determined that to account for the high resistivities (as high as 1.50 m) observed in the upper part of logging Unit 2 (Fig. 4A-C), would require the pore waters to be diluted, relative to a seawater baseline of 32 ppt, by almost 72% (to ~9 ppt NaCl). A required pore-water salinity change of 72% is much greater than the maximum observed chlorinity changes measured in the recovered cores, which was determined to be about 15% (Shipboard Scientific Party, 1996a, 1996b, 1996c). Therefore, it is unlikely that interstitial salinity differences could account for the observed resistivity log trends. To further evaluate the effect of variations in pore-water salinities on the log-measured formation resistivities, it is possible to compare the formation water resistivities (Rw) (Fig. 7) calculated from the recovered core water samples at Sites 994, 995, and 997 with the log measured formation resistivities (Rt). The log measured formation resistivities (Rt) in logging Unit 2 on the Blake Ridge is characterized by a maximum resistivity range of 1.5 m (Fig. 4A-C). However, the observed pore-water salinity variations in cores recovered from logging Unit 2 correspond to a formation water resistivity (Rw) range of only 0.05 m, which is less than 4% of the total formation resistivity (Rt) range measured in logging Unit 2. Therefore, the observed formation resistivities (Rt) variations in logging Unit 2 cannot be attributed only to changes in pore-water salinities, and the resistivity log in logging Unit 2 is likely responding to the presence of in situ gas hydrate.

Two forms of the Archie relation (Archie, 1942), discussed in the Leg 164 ODP Proceedings, Initial Reports volume (Paull, Matsumoto, Wallace, et al., 1996) and in Collett (1998), have been used to calculate water saturations (Sw) [gas-hydrate saturation (Sh) is equal to (1.0-Sw)] from the available electrical resistivity log data (DITE) at Sites 994, 995, and 997. In the first computation, the "standard" Archie equation [Sw = (a Rw /m Rt)1/n] has been used with two different sets of sediment porosity data to calculate two comparable water saturations. Both sets of porosity data used in the standard Archie equation were from the core-derived physical property data (Paull, Matsumoto, Wallace, et al., 1996). In the first calculation, the absolute value (not statistically manipulated) of the core-derived porosities were used and the sediment porosities between the core measurements were linearly interpolated. However, in the second standard Archie calculation of water saturation (Sw), the required sediment porosities were obtained from a regression trendline (power function) projected through the core porosity data in each hole. The formation water resistivities (Rw) (Fig. 7), calculated from the recovered core water samples in logging Units 1 and 3, were used in both standard Archie calculations along with the a and m Archie constants discussed in the sediment porosity section of this report (a = 1.05, m = 2.56). The value of the empirical constant n was assumed to be 1.9386 as determined by Pearson et al. (1983). In Figure 9A-C, the results of the two standard Archie calculations are shown as water saturation (Sw) log traces for Holes 994D, 995B, and 997B [gas-hydrate saturation (Sh) is equal to (1.0-Sw)].

In all three holes (Holes 994D, 995B, and 997B), the standard Archie relation yielded water saturations (Sw) ranging from about 100% to a minimum of about 80% (Figure 9A-C). In comparison, the standard Archie calculation, which used the nonstatistically manipulated core porosities, resulted in the calculation of more highly variable water saturations; whereas, the standard Archie calculation, which employed the average core porosities, yielded more consistent water saturations within each hole. The zones in each hole characterized by water saturations exceeding 100%, which is impossible, are likely caused by poor hole conditions that have degraded the resistivity log measurements. In an enlarged borehole, such as in logging Unit 1 of all three holes, the resistivity log will underestimate the true formation resistivity, which will correspond to an apparent increase in water saturations. The low water saturations in logging Unit 3 of all three holes, which is most pronounced in Hole 997B, is likely because of the presence of free gas as discussed earlier in this report.

The next resistivity log approach used to assess gas-hydrate saturations is based on the modified "quick-look" Archie log analysis technique (discussed in Collett, 1998) that compares the resistivity of water-saturated and hydrocarbon-bearing sediments. Electrical resistivity (Rt) log measurements from Holes 994D, 995B, and 997B (Fig. 4A-C) were used to calculate water saturations (Sw) [gas-hydrate saturation (Sh) is equal to (1.0 - Sw)] using the following modified Archie relationship: Sw = (Ro/Rt)1/n, where Ro is the resistivity of the sedimentary section if it contained only water (Sw = 1.0), Rt is the resistivity of the gas hydrate-bearing intervals (log values), and n is an empirically derived constant. This modified Archie relationship is based on the following logic: if the pore space of a sediment is 100% saturated with water, the deep-reading resistivity device will measure the resistivity of the 100% water-saturated sedimentary section (Ro). This measured Ro value is considered to be a relative baseline from which hydrocarbon saturations can be determined within nearby hydrocarbon-bearing intervals. To determine Ro for logging Unit 2 in all three holes, we have used the measured deep resistivity log data from the non gas-bearing portions of logging Units 1 and 3 (Sw = 1.0), to project a Ro trend-line for Unit 2 (Fig. 4A-C). Laboratory experiments on different sediment types have yielded a pooled estimate for n of 1.9386 (reviewed by Pearson et al., 1983). Now knowing Rt, Ro, and n, it is possible to use the modified quick-look Archie relationship to estimate water saturations. Displayed in Figure 9A-C, along with the standard Archie derived water saturations, are the water saturations calculated by the quick-look Archie method. The quick-look calculated water-saturations are very similar to the water saturations calculated by the standard Archie relation that employed average core porosities. However, the quick look-derived water saturations are 2% to 3% higher, which is mostly controlled by the method used to select the Ro baseline.

Gas-Hydrate Saturation Calculations—Summary

In logging Unit 2 (approximate depth of 185 to 450 mbsf) of all three holes (Holes 994D, 995B, and 997B) on the Blake Ridge the standard Archie relation yielded for the most part gas-hydrate saturations (Sh) ranging from 0% to a maximum near 20% (Fig. 9A-C); which are similar to the range of gas-hydrate saturations calculated from interstitial water chloride freshening trends (Fig. 5). In comparison, the standard Archie relation that employed the nonstatistically manipulated core porosities resulted in the calculation of more highly variable gas-hydrate saturations than the saturation calculations that used average core porosities. The use of data from different sources (downhole logs and core data) and noncompatible downhole depths have likely contributed to the more variable nature of the gas-hydrate saturations calculated with the nonstatistically manipulated core porosities. In comparison, however, the use of average porosity trends will mask localized porosity variations in complex geologic systems, which could lead to erroneous gas-hydrate saturation calculations. Because of the uniform nature of the sedimentary section cored on Leg 164, the log analysis methods that use both the nonstatistically manipulated and average core porosities yield similar results. The quick-look Archie method also yielded reasonable gas-hydrate saturations (Fig. 9A-C); however, the quick-look method is very dependent on the selection of an accurate Ro baseline.

In general, the Archie relation appears to yield accurate hydrocarbon (gas-hydrate and free gas) saturations on the Blake Ridge in spite of the fact that the sedimentary section at all three core sites on the Blake Ridge consists of mostly clay (shale), which exhibits unique electrical properties that must be corrected for in "conventional" log analysis studies. In conventional log studies, a clay can be modeled as consisting of two components: electrically inert dry clay and bound water. The electrical conductivity of a clay-rich rock is modeled as being derived solely from the clay-bound and free water. Because the porosity data used in the Archie resistivity log studies of the Blake Ridge gas hydrate accumulation actually represent the total water content of the sediments, which include both the clay-bound and free water, the Archie relation accounts for the electrical properties of both the clay-bound and free water. This assumes the electrical conductivity of the clay-bound and free water are similar, which is likely true in these low salinity pore-water systems. It should be noted that there are several electrical conductivity models that offer improvements over the Archie relationship when considering clay-rich sediments (Serra, 1984). Application of these extended electrical conductivity models would be a step forward to fully understand the electrical logs from the Blake Ridge boreholes. However, the use of these complex extended Archie relations are beyond the scope of this paper.

NEXT