多源反演数据优势频段融合薄储层预测技术及应用OA

The development of thin sand bodies(thickness:4～6 m)in H3 reservoir,the Huoshaoshan Oilfield,Junggar Basin,faces verti-cal and horizontal prediction challenges.Given this,this paper proposed a technology that combines spectrum bluing for frequency broadening with dominant frequency band fusion of multi-source inversion data.Based on the blue spectrum of logging-derived reflec-tion coefficient,the proposed technology expanded the pre-stack seismic high-frequency band to 120 Hz through well-seismic modeling.By doing so,it can enhance reservoir responses above 60 Hz and suppress low-frequency noises.The frequency-specific processing was then conducted for the three types of data volumes.Specifically,the low-frequency model(0～8 Hz)retained the structural trend,while the intermediate-frequency pre-stack waveform indicator inversion VP/VS volume(8～60 Hz)depicts the lithological boundary.The high-frequency pre-stack geostatistical inversion VP/VS volume(60～120 Hz)captures details of thin sand bodies.The adaptive weighting allo-cation achieved full-band synergistic integration,overcoming the identification limitations(8‒10 m seismic tuning thickness)of tradition-al inversion methods.The application of this technology shows that the fused data volumes can clearly identify thin sand bodies(4 m),with the lateral inter-well connectivity error reduced by 25%compared to conventional inversion methods.This technology established a multi-scale collaborative model of"trend controlling in low-frequency bands,boundary controlling in intermediate-frequency bands,and detail capturing in high-frequency bands",breaking through the bottleneck of multi-solutions from single-source inversion.The informa-tion fusion of full-frequency bands provides a replicable technical path for the detailed assessment of continental clastic reservoirs with thin interbeds.It also offers methodological inspirations for the development of similar complex reservoirs.