Critical techniques for sweet spot prediction for tight sandstone reservoirs in the Dongsheng gas field and their application effects
The Xinzhao gas zone of the Dongsheng gas field resides at the junction of three first-order tectonic units:Yishan slope,Yi-meng uplift,and Tianhuan depression.Due to the influence of paleogeomorphology and provenance,different channel sediments vary sig-nificantly in this zone.The main target layer is the first member of the Shihezi Formation,which hosts a braided river sedimentary sys-tem,with tight reservoirs characterized by typical low porosities and permeabilities.The effective sandstone reservoir exhibits a small thickness,pronounced heterogeneity in gas content,and low-resolution seismic data,complicating sweet spot prediction.Hence,this stud-y delved into the critical seismic prediction techniques for sweet spots in braided-river tight sandstone reservoirs.Firstly,the stratigraph-ic framework was established based on the three-dimensional seismic relative isochronous surface,characterizing the spatial distributions of channel sand bodies at different stages using the relative-spatial-resolution stratal slicing technique.Then,the seismic data were trans-formed from the time domain to the frequency domain using the wavelet transform time-frequency analysis technique.Based on the anal-ysis of instantaneous spectrum differences in seismic data of different channels,the qualitative gas content prediction was achieved using the low-frequency energy ratio and the ABV absorption attribute,effectively supporting well deployment.Finally,the facies-controlled geostatistical inversion in the depth domain combining logs and seismic data was conducted for quantitative reservoir prediction,guiding the accurate design and optimization of horizontal well trajectories.The application of these techniques has increased the probability of penetration of gas reservoirs in the study area by seven percentage points,demonstrating satisfactory application effects.
tight sandstonereservoir predictiongas content predictiontrajectory design of a horizontal well
国家科技期刊平台
NSTL
万方数据
