A frequency-divided gas prediction technology based on resistivity constraint

Abstract

The gas–water relationship in the Permian and Triassic biohermal beach reservoirs in the Sichuan Basin is complex, which results in a low drilling success rate and leads to sidetracking or even secondary sidetracking in some wells. Therefore, high-precision gas–water identification is the key to the drilling success rate. In this paper, the biohermal gas reservoir in the T work area of the eastern Sichuan Basin was taken as an example. Firstly, the frequency information of seismic data sensitive to gas bearing properties was exploited sufficiently by taking the deep/shallow resistivity information which is sensitive to gas–water discrimination as a constraint and the nonlinear algorithm as a bridge. Then, combined with the information of full-band relative wave impedance, a nonlinear mapping relationship between the logarithmic difference curve of deep/shallow resistivity and the seismic waveform was established, and the gas bearing properties of the reservoir were predicted. In this way, a post-stack gas prediction method was developed. Finally, field application and effect analysis were carried out. And the following research results were obtained. First, gas layer information indicated by deep/shallow resistivity not only has the function of verification well, but can be extracted as a priori information for the constraint of seismic data. Second, frequency division information increases the mapping relationship between multi-frequency data volume and the logarithmic difference of deep/shallow resistivity, and the colored inversion, as the more reasonable full-band relative-wave impedance seismic data, is more conducive to gas and water identification. Third, based on the nonlinear mapping relationship, the data volume of a gas sensitive factor was obtained, and the range of gas layers was quantified, so as to realize the quantitative characterization of gas bearing properties. It is concluded that the gas prediction technology developed in this paper provides an effective and practical post-stack gas prediction method, and its field application effect is good, presenting wide popularization and application prospects.