Seismic Processing for Gas Hydrate Research Drill Site in the Mackenzie Delta, Canada

Thirteen seismic profiles totaling 242 km were reprocessed in the Mackenzie Delta area in order to map the shallow gas hydrated deposits and to map detail structure around the proposed drill site for the gas hydrate research program. Seismic data were acquired for conventional oil exploration purposes in this area in the middle of 1980, so the field parameters were not optimum for the shallow structure studies. However, reprocessing results provided adequate information necessary to understand the distribution of hydrate deposits and fault patterns near the drill site. All seismic data were stacked, migrated and converted to depth. Also, all seismic data were depth migrated before stacking. With the exception of one line, pre-stack depth migration provided better subsurface images in this area. Due to the limitation of software, only a single step (iteration) pre-stack, finite difference, depth migration was used in the shot domain. The final velocity model for the pre-stack depth migration and post-stack time migration was derived from the dip-moveout (DMO) corrected common-mid-point (CMP) gathers. Post-stack wavelet deconvolution and zero-crossing predictive deconvolution was applied to all seismic data in order to increase temporal resolution and to remove the differences of the source wavelets. INTRODUCTION The Mackenzie Delta gas hydrate research drilling program was initiated by the Japan National Oil Corporation (JNOC) in order to verify gas hydrate drilling/coring and completion technologies which were developed for a gas hydrate exploration program they have planned for offshore Japan in 1999. The Geological Survey of Canada (GSC) and the United States Geological Survey (USGS) undertook a complete science program in cooperation with fellow North American and Japanese scientists. This program included fundamental gas hydrate studies, permafrost and stratigraphic studies, well logging studies and geophysical studies including vertical seismic profiles (VSP's). After a long review process, a drill site near the Mallik L-38 industry exploration well in the Mackenzie Delta, Canada, was selected and seismic lines near the Mallik L-38 were provided by Imperial Oil Resources Limited and reprocessed on SUN workstations using ProMAX processing software developed by Landmark, Inc. together with in-house developed software. The surficial sediments of the Mackenzie Delta are composed in part of modern deltaic sediments and of older fluvial and glacial deposits. At depth, the area is underlain by deltaic sandstone and shales of Mesozoic and Cenozoic age that thicken to more than 12km over a short distance seaward from the present shoreline. This sedimentary section overlies faulted Paleozoic rocks (Dallimore and Collett, 1997). The proposed well site overlies an anticlinal structure. Interpretation based on well logs indicates that three sequences, each with distinctive natural gamma log response exist in the upper 1400m, the Iperk sequence (0-350m) appears to be composed almost entirely of coarse grained sediments, which are unconsolidated with no grain cementation. The Mackenzie Bay sequence (350-950m) is also sand dominated and the Kugmallit sequence (>935m) consists of interbedded sandstone and siltstone (Dallimore and Collett, 1998). The base of ice bearing permafrost is estimated at around 640m in this region. Regional ground temperature data form nearby wells suggest that near surface temperatures are likely to vary from -6° to -2°C in the upper 100m, but below that depth an extensive isotherm section (close to -1°C) may occur. Ice bonding is expected to be quite variable throughout the sequence with possible taliks or non-ice bonding sections (Collett personal communication, 1998). The wireline logs and mud gas log responses at Mallik L-38 indicate that there are 10 separate hydrated layers ranging in thickness from 3.4 m to 25.4 m and the total cumulative thickness of the hydrated layers are 111.4 m (Collett and Dalimore, 1997). The shallowest hydrate layer occurs at 819 m (Mackenzie Bay sequence) and the deepest hydrate layer occurs at 1091 m (Kugmallitt sequence). The existence of free gas under the hydrated layer was interpreted by Bily and Dick (1974) based on a spontaneous-potential well log response and a rapid pressure response during a production test. But Collett and Dallimore (1997) were not able to confirm the occurrence of free gas because of insufficient data. The primary purpose of reprocessing was to image the continuity of the hydrate deposits near the proposed well site and to map the structural details around the drill site, particularly the fault patterns, so that the seismic profiles could be used to assist in spudding and drilling the gas hydrate research well. This report discusses the details of the processing steps. DATA PROCESSING The processing flowchart is shown in Figure 1 and relevant information pertaining to the available seismic lines is listed in Table 1.