Análisis De Avo En Rocas De Alta Impedancia

Case study: AVO analysis in a high impedance Atoka sandstone north Arkoma basin, Mcintosh County, Oklahoma
Mohamed A. Eissa* and John P. Castagna, Institute for Exploration and Development Geosciences, The University of Oklahoma, Norman, OK

Summary Amplitude-Variation with Offset (AVO) analysis can be used to detect and delineate gas bearing fluvial deltaic Atokan sand stratigraphic traps.These class I high impedance gas sand reservoirs are recognized by a characteristic phase change with increasing offset, a positive AVO intercept, and a negative AVO gradient. Introduction The variation of reflection coefficients with source-toreceiver spacing in multioffset seismic data is known to contain information about lithology and pore fluid content of subsurface rocks (e. g. Ostrander,1984). However, such analysis is particularly challenging for thin, high impedance reservoirs (Rutherford and Williams, 1989). In such instances, a key component of amplitude variation with offset (AVO) analysis is a methodology for response calibration so that fluid properties can be interpreted vertically and laterally away from a borehole with a high degree of confidence (e.g., Hong, et al 1993).The use of AVO as a direct hydrocarbon indicator in clastic rocks is based on differences in the response of the P-wave velocity (Vp) and S-wave velocity (Vs) of a reservoir rock to the introduction of gas into the pore spaces (Ostrander, 1984). The objective of this study is to determine if AVO analysis can be used to identify and delineate thin high-impedance Atokan gas sands in the North Arkomabasin. Location The area of study is located in the shelf area of the northern Arkoma basin, Mcintosh County, Oklahoma (Fig. 1). The target reservoir is the Pennsylvanian Atoka Sand.

Available data Well logs from the Wright 1-15 well (Fig. 2) were available for this study. These included the gamma ray, deep and shallow resistivity, and porosity logs (neutron, density, and sonic). The targetreservoir is fluvial Pennsylvanian Lower Atokan sand at depths 2585` to 2695`. The perforated interval (2610` to 2638`) showed initial production of 88 MCFGPD. The main reservoir package is 110 feet thick (from 2585` to 2695`). It consists of sand, shaly sand and shale. We assume that most production comes from the thin clean sand body (2611` to 2618`) within the package. The pay interval shows anaverage porosity of 10%. There are some other sand bodies at 2350` (45` thick) and 2495` (30` thick) which show gas.

Figure 1 Location map.

Figure 2 Well logs for the target reservoirs within Atoka in the Wright 1-15 well (gas sands indicating by arrows)

SEG Int'l Exposition and Annual Meeting * San Antonio, Texas * September 9-14, 2001

AVO Analysis in Atoka sandstone in Arkoma basinSpecial processing for AVO included parabolic Radon filtering (Humpson-Russell INVEST procedure, 10-80Hz and time shifts from +80 to –80 msecs were passed) of supergathers consisting of 10 summed CDP gathers. Figure 3 shows real data gathers before and after Radon filtering and the subtracted noise model. than the seismic data are responding to. The AVO response of the real data exhibit typical AVObehavior for class I sands with a phase change with offset according to the Rutherford and Williams (1987) classification. Near stack and far stack sections (Fig. 7) exhibit clearly anomalous AVO responses between CDPs 215 to 226 at TWT 480 ms.

Figure 3 Real data before applying Radon filter (A), after applying Radon Filter (B), and the noise model (C). Seismic modeling Primaries onlyray-trace synthetic seismograms using sonic and density logs from the Wright 1-15 well were generated using the extracted wavelet at the well location. Synthetic gathers are NMO corrected and calculated at the same offset as the real data (from 0` to 3800`) with no correction for spreading, transmission loss, attenuation, or geophone/array response as the data were previously amplitude balanced. Shear...