John B. Gustavson, Xin Shi Gong
The Suizhong 36-1 oil field is located in Liaodong Bay, People's Republic of China. Oligocene fluvio-deltaic and lacustrine sandstones contain over 1 billion barrels of oil-in-place in a combination structural-stratigraphic trap. An international effort to delineate the reservoirs by geophysical methods was completed in June 1989. Three hundred and fifty kilometers of seismic data were reprocessed and interpreted.
An attempt was made to reprocess vertical seismic profile (VSP) data to better relate seismic data to subsurface reservoir parameters. Unfortunately, incomplete field data and/or documentation made such an effort infeasible.
Synthetic seismograms were used for correlation and for wavelet processing of the seismic data. Seismic-petrophysical analyses were carried out to relate the measurable seismic parameters to the subsurface rock and fluid parameters.
Seismic data processing comprised conventional reprocessing, special processing primarily for hydrocarbon indicators, and 3-D velocity analysis. The reprocessing of the seismic data improved data quality over that of the original processing. Amplitude versus offset (AVO) processing revealed no obvious hydrocarbon indicators in the main reservoir zones. Velocity analysis produced a 3-D velocity field that was the key to the interpretation and engineering work that followed.
Stratigraphic and structural interpretation were carried out on a GeoQuest workstation and iterated several times to obtain a consistent match with the wells. Seismic attribute generation and analysis were used primarily to reveal patterns associated with the depositional environments.
Two kinds of seismic trace inversion processes were used to obtain quantitative estimates of the subsurface acoustic impedance distribution. One, the approximate or pseudo-inversion method developed by Lindseth, was applied to 300 km of seismic data and was the basis for all the reservoir parameter estimates. The other, the generalized linear inversion approach developed by Backus and Gilbert, was applied to 20 km of seismic data in the center of the field as part of the special processing requirement.
Three principal approaches to the detection of hydrocarbons were tried in an effort to map oil-water contacts in the field. Relative Amplitude Profiles (RAP) sections and normal-incidence P-wave profiles were analyzed for amplitude anomalies. AVO gradient profiles were analyzed for amplitude gradient anomalies. These efforts failed to detect hydrocarbon presence or oil-water contact levels conclusively.
In order to estimate hydrocarbon reserves and reservoir parameters, seismically derived values of acoustic impedance were converted first into estimates of porosity using the relationships derived from seismic-petrophysical analysis. Next, the porosity distribution was transformed into maps of net reservoir thickness, permeability, and oil saturation. At each stage, the results were calibrated back to the wells to ensure consistency of the estimates. The results suggest that accurate predictions of oil-in-place are possible with this methodology.
AAPG Search and Discovery Article #90097©1990 Fifth Circum-Pacific Energy and Mineral Resources Conference, Honolulu, Hawaii, July 29-August 3, 1990