Application of SVR attribute fusion constrained by geological model in reservoir description of shallow water delta: A case study of X gas field in Xihu Sag

Abstract

X gas field is located in the central inversion structural belt of Xihu Sag, East China Sea Basin.The main target layer H4 is in shallow water delta sedimentary environment.The dominant frequency of seismic data of the gas field is low (25 Hz), while the buried depth of H4 layer is large (3 300-3 400 m).The reservoir has low porosity and low permeability, and the consistency of sand body thickness predicted by conventional seismic inversion is low.Aiming at the characteristics of full coverage of 3D seismic data and less drilling in X gas field, this paper makes up for the lack of sample points in SVR algorithm through forward and inverse modeling under the guidance of geological model and setting up virtual wells.Then, by extracting seismic attributes and optimizing sensitive attributes characterizing sand body thickness, SVR algorithm is used for multi-attribute fusion to complete the quantitative prediction of H4 sand body.Based on the reservoir prediction results, it is proposed that H4 layer is a distributary channel deposit of shallow water delta meandering flow type, and further completed the dissection of sand body deposition model.After the development adjustment wells drilled, the coincidence between the actual drilled sand body thickness and the predicted sand body thickness is more than 84%.This paper explores and obtains the quantitative reservoir prediction scheme of SVR algorithm constrained by geological model under the condition of few wells on the sea, which completes the accurate prediction of the middle and deep distributary channel reservoir in X gas field, and also has certain guiding significance for the reservoir description of the same type of oil and gas fields.

Keywords

• xihu sag
• distributary channel
• svr attribute fusion
• reservoir prediction