For all the holes logged, we generated two different composite figures. The first is a summary of the STC analysis results of the different modes recorded during each sonic log pass. It provides a quality control assessment of the velocity data and also helps identify some of the loss of coherence that can be associated with free gas–bearing intervals. The second composite for each site combines the different logs associated with the sonic data and the presence of gas hydrate. The VP and VS logs and the sonic waveforms constitute our primary data. When available, we added the interval velocities calculated from the VSP and the comparison of the TvD relationships between the VSP and the synthetic seismograms. High resistivity values in the resistivity log are a qualitative indicator for the presence of gas hydrate, and the deep induction resistivity log is used to derive gas hydrate saturation estimates. The caliper log provides an assessment of the quality of the hole and of the reliability of the logs. The density log is combined with the VP log to generate the synthetic seismogram and consequently can help identify some of the reflectors.
We highlight the panels with the main new results: the synthetic seismograms generated from the density and VP logs are displayed as a function of depth to allow direct correlations between the logs and the seismic data; the gas hydrate and the free gas saturations derived from the cementation and the Gassmann models are compared with the estimates provided by the resistivity logs. These estimates are based on Archie's law and are calculated with the method described by Collett (1998), applied to the deep induction wireline resistivity log for each hole. The resulting water saturations can be interpreted in terms of gas hydrate or free gas saturation depending on the location within or outside the gas hydrate stability zone.
In order to illustrate these correlations and their implications for the Hydrate Ridge system, we superimpose each synthetic seismogram on a short east-west section of the 3-D seismic survey. The results are discussed from the west flank sites (Holes 1245E and 1247B) through the crest (Hole 1250F) to the eastern flank (Hole 1244E) and the eastern basin (Holes 1252A and 1251H) (see Fig. F1 for locations).
Figure F6 shows that, despite the irregular hole conditions (see caliper in Fig. F7), the monopole and lower dipole waveforms displayed strong coherence overall, allowing us to draw reliable velocity logs. However, as a result of the irregular hole and of the low variability in the data, both the sonic and density logs are noisy and combine to produce only an adequate synthetic seismogram. The comparison of the seismogram with the seismic data and the logs in Figure F7 still allows us to identify the main reflectors in Hole 1245E: the BSR at 130 meters below seafloor (mbsf) and Horizon A at 180 mbsf. However, the amplitude of these reflections in the synthetic seismogram is significantly lower than in the seismic data. Both reflections in the synthetic seismogram are generated by slight decreases in VP, attributed to the bottom of the gas hydrate stability zone for the BSR and to the presence of free gas within the volcanic glass–rich sediments and ashes that constitute Horizon A (Tréhu, Bohrmann, Rack, Torres, et al., 2003). The low amplitude of the synthetic reflections might also indicate that some of the free gas responsible for the reflections in the 3-D seismic data escaped during drilling and prior to the recording of the sonic logs. As one of the main conduits for gas into the Hydrate Ridge gas hydrate system, and despite possible escape of free gas, Horizon A is still the location of the highest free gas saturation identified by the Gassmann model. Another strong reflection at ~150 mbsf coincides with the "double BSR" at 1.37 s two-way traveltime (TWT) described by Bangs et al. (2005). In Hole 1245E, this reflection corresponds to a steplike increase in the resistivity log. Although there is no indication of a lithologic change associated with this resistivity increase, it seems to mark the top of the occurrence of free gas indicated by the Gassmann model. The lower monopole waveform amplitudes and the results of the Gassmann model suggest that free gas is pervasively present below this depth.
Gas hydrate saturations predicted by the cementation theory above the BSR are overall similar to the values predicted from the resistivity log by Archie's law but suggest a more uniform distribution (Fig. F7). Below the bottom of the gas hydrate stability zone, the presence of gas hydrate suggested by the cementation model is likely an artifact resulting from the lithological variability and the deformed nature of the sediments, which are not taken into account in the model.
The high coherence of the sonic waveforms (Fig. F8) and the smooth caliper log in Figure F9 show that Hole 1247B was in excellent condition despite the poorly consolidated nature of the Hydrate Ridge sediments. As a result, all logs are of very good quality and the synthetic seismogram clearly identifies the BSR at 128 mbsf, Horizon A at 153 mbsf, and most of the other reflectors. The amplitude of Horizon A is again lower in the synthetic trace than in the seismic data, which could also be attributed to the escape of some of the original free gas present in this interval. The favorable hole conditions also allowed a strong coupling of the geophones at most attempted stations of the VSP. The good agreement between the traveltimes derived from the synthetic seismogram and from the VSP validates these two independent data sets. However, despite the good hole conditions and data quality, the interval velocity values derived from the VSP are noticeably higher than the VP log below the BSR. Considering the overall narrow range of the VP values, this does not affect the good agreement between the TvD relationships. This discrepancy could suggest some local effect due to the heterogeneous distribution of free gas and the differences between the tools. While the DSI waveforms travel in close vicinity to the borehole, the VSP shots travel from the surface to the downhole geophone and are less sensitive to local free gas occurrences over their path. In Figure F8, the wide coherence peaks in the monopole waveforms below Horizon A also suggest some arrivals faster than the ones picked with the highest coherence, indicating the possible influence of an heterogeneous free gas distribution.
The sharp drop in VP that generates the BSR, typical of the traditional representation of gas hydrate reservoirs, is mostly due to the high gas hydrate saturation indicated by the results of the cementation theory. The presence of gas hydrate is surprisingly less apparent in the resistivity log, and although both hydrate saturation curves provide similar values near 20% of the pore space, the hydrate saturation derived from the sonic log is overall higher and more uniform than predicted by Archie's method. Only a few gas hydrate samples were recovered from the core at this site (Tréhu, Bohrmann, Rack, Torres, et al., 2003), but the presence of strong chlorinity anomalies (Tréhu et al., 2004b) supports the sonic log indication of significant amounts of gas hydrate immediately above the BSR. This is also consistent with the location of Hole 1247B near the crest of southern Hydrate Ridge, where the highest hydrate saturations have been observed, and with the presence of a very clear BSR (Fig. F9).
In agreement with the observations of Guerin and Goldberg (2002), all waveform amplitudes are significantly lower above the BSR, and the effect of gas hydrate on waveform amplitude is stronger on the upper dipole waveforms than on the lower dipole because of the higher source frequency of the upper dipole source.
Below the BSR, the Gassmann model indicates consistent amounts of free gas, with the highest saturations (~1.5%) between 165 and 170 mbsf, 10 m below Horizon A. This coincides with significantly lower amplitudes in the logging waveforms. Similar correlation between significant free gas occurrence and low amplitudes is observed in several intervals below the BSR.
As in Hole 1247B, the high waveform coherence in Figure F10 and the caliper log in Figure F11 indicate very good hole conditions and very reliable logging data in Hole 1250F. Accordingly, the synthetic seismogram reproduces all the most significant reflectors, including the BSR at ~115 mbsf and Horizon A at 152 mbsf. The resulting TvD relationship is also in good agreement with the transit times measured by the VSP.
Gas hydrate saturations calculated from the sonic log by the cementation theory indicate values similar to the results of Archie's law but suggest a less uniform distribution and significant saturations all the way to the BSR, whereas the resistivity log does not identify any gas hydrate below ~105 mbsf. The absence of strong chlorinity or infrared anomalies immediately above the BSR (Tréhu, Bohrmann, Rack, Torres, et al., 2003; Tréhu et al., 2004b) also supports the absence of gas hydrate in this interval. This contradiction between the resistivity and the VP log can be due partially to the different tool geometries, orientations, and energy paths. It is also an indication of the highly heterogeneous gas hydrate distribution at the crest of Hydrate Ridge. The VP log and the saturation model results show clearly, as in Hole 1247B, that the BSR marks the shallowest occurrence of free gas. Milkov et al. (2004) and Tréhu et al. (2004a) suggest that, because of the excess input of free gas in the system, free gas and gas hydrate might coexist above the BSR. The consistently high VP values above the BSR seem to contradict this interpretation or to indicate a more complex interplay between the heterogeneous gas hydrate and free gas distributions in this active system.
Despite the consistent presence of free gas identified by the Gassmann model below the BSR, this model fails to identify free gas in Horizon A in Hole 1250F. By comparison, Tréhu et al. (2004a) use the logging-while-drilling (LWD) density logs to estimate very high (>50%) free gas saturations in this horizon, which is supported by a crossover of low density and low neutron porosity in the LWD logs (Tréhu, Bohrmann, Rack, Torres, et al., 2003). The low dipole waveforms amplitude and coherence in this interval also support the presence of free gas and the interpretation of Horizon A as the main free gas conduit to the system (Tréhu, Bohrmann, Rack, Torres, et al., 2003; Tréhu et al., 2004a). The absence of a decrease in VP indicates that most of the free gas present might have escaped by the time of the sonic log, which does not affect the LWD measurements made immediately after the bit penetration. The strong decrease in VS in Horizon A suggests that there might have been an elevated pressure in Horizon A before drilling and that, considering the poorly consolidated nature of the sediments, the pore fluid was carrying most of the burden. As the gas and the pressure were released by drilling, the matrix collapsed and lost part of its cohesion and shear strength, generating a decrease in VS.
The strong waveform coherence for all the passes and modes in Figure F12 and the almost perfect caliper log in Figure F13 indicate that all the wireline logs recorded in Hole 1244E were of excellent quality. Figure F13 also shows that Hole 1244E was the only location on Hydrate Ridge where it was possible to obtain good coupling for the VSP at every attempted depth, every 5 m. Accordingly, the interval velocities derived from the VSP and the VP log are in excellent agreement, as well as the TvD relationships derived from the VSP and the synthetic seismogram. In addition to the very clear direct arrivals, the VSP waveforms in Figure F4 also feature the reflection of the most prominent reflector at this site: Horizon B´ at ~217 mbsf. The other reflectors identified by correlation between the synthetic seismogram and the seismic data are Horizon B at ~180 mbsf and the BSR at ~140 mbsf. The good agreement between the TvD relationships suggests that the BSR is deeper by 11 m than estimated by Tréhu, Bohrmann, Rack, Torres, et al. (2003).
Archie's relationship and the cementation theory applied to the resistivity and sonic logs indicate lower hydrate saturations in Site 1244 than in the crest sites. This observation is supported by the fact that Site 1244 is located to the east of Hydrate Ridge (Fig. F1), away from the main gas hydrate accumulation fed by Horizon A. All estimates suggest a fairly uniform gas hydrate saturation around 15% of the pore space above the BSR. A reflection at 90 mbsf, or 1.32 s TWT, coincides with the bottom of an interval with higher resistivity, where the Archie relationship predicts gas hydrate saturation values significantly higher than the results of the cementation model. Correlation between the synthetic seismogram and the seismic line shows that this interval also displays higher seismic reflectivity, which is an apparent contradiction with high gas hydrate saturations (Guerin and Goldberg, 2002). The slightly higher density in this interval, also contradictory with high hydrate saturations, indicates that the resistivity increase is not due uniquely to the presence of gas hydrate but possibly to some different lithology, although this was not identified in the core observation (Tréhu, Bohrmann, Rack, Torres, et al. 2003).
The lower VP values and the loss of waveform coherence in several short intervals between Horizons B and B´ suggest the presence of free gas, which is identified by the Gassmann model and also indicated by the low monopole waveform amplitudes in discrete intervals. The highest gas saturations are measured within Horizon B, indicating that this turbiditic horizon also contributes to the gas hydrate accumulation at the crest of Hydrate Ridge, as well as the volcanic glass-rich Horizon B´.
As in Hole 1245E, the suggestion by the cementation model of gas hydrate below the bottom of the gas hydrate stability zone is an artifact resulting from the overall low gas hydrate saturations and the lithological variability and deformed nature of the sediments. In particular, between 160 and 180 mbsf, the increase in velocity responsible for the apparent gas hydrate occurrence is due to a decrease in porosity, indicated by the higher resistivity and density, which likely reflects a local deformation immediately above Horizon B.
Figure F14 shows that the higher frequency upper dipole waveform displays only a weak coherence in the upper section of Hole 1252A, as well as the high frequency monopole waveforms in the deeper section (Pass 2). However, all the VP and VS logs agree well between the different passes and sources, suggesting that the sonic logs are of good quality, despite the low amplitudes shown in Figure F15. Drilled on the western flank of a secondary anticline east of Hydrate Ridge, Hole 1252A is located where the BSR within the accretionary core of the anticline intersects the overlying sediments of Hydrate Ridge's eastern flank (see the seismic section in Fig. F15). As a result of this deformation, the sediments are likely to be less cohesive, which is confirmed by the very irregular caliper in Figure F15. However, the hole enlargements never exceed the maximum reach of the tools. As another consequence of the local deformation, the BSR is interrupted on the seismic line near the well location and can only be correlated to a weak reflector in the synthetic seismogram at ~165 mbsf. The strongest reflector in the seismogram is generated by the high velocity and density in a series of glauconitic sand layers between 115 and 120 mbsf, at the contact between the anticline core and the overlaying sediments.
Below the BSR, the VP log shows a slightly decreasing trend with depth, opposite to a normal consolidation profile within these disturbed sediments. Because this trend does not occur in the deformed sediments above the BSR, this suggests that the presence of gas hydrate might contribute to the consolidation of the sediments above the BSR.
Gas hydrate saturations derived from the elastic logs indicate a relatively uniform gas hydrate distribution, increasing slightly downhole to a maximum value of ~10% of the pore space at the BSR. The resistivity log suggests a more heterogeneous gas hydrate distribution, indicating in particular some high-saturation intervals between 80 and 100 mbsf, in the undeformed sediments above the glauconitic sands. Only few indications of gas hydrate were observed in the recovered cores, but several infrared camera anomalies suggested that any gas hydrate present had to be disseminated (Tréhu, Bohrmann, Rack, Torres, et al., 2003). The low dipole waveform amplitudes above the BSR tend to support the prediction of the cementation model of a significant disseminated gas hydrate distribution (Guerin and Goldberg, 2002). The high saturations derived by the cementation theory within the glauconite-rich sands and the apparent hydrate presence below the BSR by the resistivity and the elastic logs are indications that these highly disturbed sediments present a consolidation state distinctly different from the normal pore space configuration assumed in the cementation theory and possibly in Archie's law. Similarly, the apparent presence of gas hydrate suggested by the cementation model below the bottom of the gas hydrate stability zone is also a consequence of the deformed nature of the sediments. By comparison, the Gassmann model, which makes no assumption on the pore scale grain interactions, predicts some possibly significant gas presence in a low-velocity interval between 210 and 220 mbsf. The low sonic waveform amplitudes in this interval support the presence of free gas, which was not identified previously.
The results of the STC analysis of all the waveforms recorded in Hole 1251H in Figure F16 indicate consistently low velocity and low dipole waveform coherence over the entire interval logged, characteristic of poorly consolidated sediments. In particular, VS values are so low that the 20-ms window used to record the dipole waveforms was too short to capture the flexural waves above ~105 mbsf. The irregular caliper log in Figure F17 indicates that the bad hole conditions could be responsible for some loss in waveform amplitude and coherence. In particular, the high-frequency upper dipole waveform in Figure F17 seems to have been particularly affected by the hole conditions. But the low quality of the borehole, and consequently of some of the waveforms, mostly underlines the poorly consolidated nature of the sediments. Although the hole was drilled to 445 mbsf, the unstable formation caused obstructions to form and the hole to collapse in the deeper section, so that no sonic data were recorded below 185 mbsf, ~10 m above the BSR (Tréhu, Bohrmann, Rack, Torres, et al., 2003). The enlarged hole and the lack of cohesion of the sediments also prevented a good mechanical coupling between the formation and the VSP tool so that waveforms were recorded successfully only at four stations (Fig. F5). As another consequence of the enlarged hole, the density log was of bad quality in intervals where the hole was too large for the tool to make contact with the formation. Consequently, we used the density measured on core samples to generate the synthetic seismogram (Fig. F17).
Despite the difficult hole conditions, the other logging data recorded in Hole 1251H were of good quality and the synthetic seismogram reproduces well most of the reflectors penetrated by the well. The TvD relationship derived from the sonic log and the seismic/synthetics correlation is also in good agreement with the check shots provided by the VSP. In the absence of a BSR, the most significant reflector for the log/seismic correlation is generated by the angular unconformity at 1.78 s TWT, which was identified in the cores at ~130 mbsf. The brightest reflector in the seismic line and in the synthetic seismogram corresponds to a bed dipping to the east at 1.81 s TWT that is generated in the synthetic seismogram by an interval with reduced VP between 143 and 148 mbsf. This interval does not correspond to any lithology change in the cores (Tréhu, Bohrmann, Rack, Torres, et al., 2003) but is also characterized by low dipole waveform amplitudes. The low VP and dipole amplitude, as well as the brightness of the reflector, suggest a possible presence of free gas within the gas hydrate stability field, similar to observations by Guerin et al. (1999). However, the simultaneous presence of gas hydrate and free gas changes the sediment frame properties, so the Gassmann model does not detect any free gas.
The cementation theory predicts a steady increase in gas hydrate saturation with depth, from 0% to ~10% of the pore space, whereas the resistivity log suggests somewhat higher values in a few discrete intervals. The discrepancy between the two methods could be partially due to the hole conditions and to the difference between the geometry of investigation of the two tools in an overall heterogeneous hydrate distribution. However, the high gas hydrate saturation at the bottom of the interval indicated by the cementation theory agrees with the strong infrared and chlorinity anomalies observed immediately above the BSR in adjacent Holes 1251A and 1251D (Tréhu, Bohrmann, Rack, Torres, et al., 2003; Tréhu et al., 2004b).