Prestack Simultaneous Inversion Based On Multi-point Geostatistics

The exploration and development of oil and gas fields are mostly directed at reservoirs.Reservoir modeling and stochastic inversion methods have also been hot topics in industry and academia.Geostatistical modeling is based on well data with high vertical resolution,and the simulation results are usually not accurate in the far well area.Conventional seismic inversion is based on seismic data with high resolution on the plane,and the spatial structure of the inversion model lacks geological significance.In recent years,consensus has been reached on the integration of multidisciplinary inversion to improve the value of seismic inversion and the value of geological modeling.Geostatistical inversion is an inversion method that combines geological modeling and seismic data with geological modeling and seismic inversion.Traditional geostatistical inversion effectively integrates traditional geostatistics with conventional stochastic inversion in different workflows,and achieves good inversion results.With the development of geostatistics and random inversion,it is the current development trend to use effective and efficient means to perform geostatistical inversion.Based on geological,logging,seismic and petrophysical data,combined with DS algorithm,McMC method and prestack simultaneous inversion theory,a McMC prestack simultaneous inversion method based on DS algorithm is developed in this paper.Compared with Simpat algorithm,DS algorithm is simple and flexible,which can greatly improve the calculation efficiency.McMC sampling method can obtain reservoir parameter samples from stationary distribution equivalent to a posteriori probability by constructing Markov chain.The prestack simultaneous inversion technology can use the relationship between P-wave impedance,S-wave impedance and density to perform the P-wave impedance,S-wave impedance and density data volume,eliminate the instability of inversion algorithm,and predict lithology and fluid properties more accurately.Starting from geostatistics,sampling method and seismic inversion theory,this paper introduces their basic concepts,and then introduces the current geostatistics inversion method and the new method proposed in this paper for comparative analysis.Finally,two-dimensional channel model and Stanford ?-E reservoir and Pengzhou thin interbed model are selected,The inversion test of shows that single angle inversion extends to multi angle inversion,and the inversion result is slightly better than single angle inversion in the description of sand body shape;MCMC sampling inversion result is close to the result of refusing sampling,while the former takes more samples,about twice more than refusing sampling;Simpat algorithm based inversion result is better than theoretical model It is consistent,but the calculation cost of this method is twice as much as that of DS algorithm;it will have a good effect to apply the relationship between P-wave impedance and density to inversion and fluid prediction;the new method can identify reservoir characteristics more effectively by comparing with the conventional geostatistical inversion method and the geostatistical inversion method based on the constant renewal ratio theory.Therefore,the McMC pre stack simultaneous inversion based on DS algorithm is feasible and efficient.