Seismic facies analysis method based on semi-supervised well-log facies classification and MCC-SOM identification and its application

Seismic facies analysis is an important basic work in lithological reservoir exploration and development. However, geological interpretations are often non-unique based on the same or similar seismic reflection data. Moreover, seismic interpretations require significant amounts of time, experience, and expertise from interpreters. Artificial intelligence (AI) techniques can help interpreters reduce some of these problems associated with seismic facies analyses. The problem is that having a good training model is not always guaranteed to derive good prediction results for an extended area. In order to improve the accuracy and efficiency of seismic facies analysis, this study proposed a new seismic facies analysis method based on semi-supervised well-log facies classification and MCC-SOM. The application in M oilfield in the Middle East shows that this method can improve the accuracy and efficiency of seismic facies analysis. The advantage of this method is that it can integrate core facies, well-log facies and seismic attributes, and optimize the best solution from the non-unique seismic facies interpretations based on the well log facies calibration and Walther’s law of the correlation of facies.