Full text loading...
-
Using Convolutional Neural Networks (CNN) and Seismic Data to Predict Gas Content in Shales Gas Reservoirs
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, 81st EAGE Conference and Exhibition 2019, Jun 2019, Volume 2019, p.1 - 5
Abstract
In this study, in order to quantitatively predict the gas content in the shale reservoir and determine the spatial distribution of sweet spots with high gas contents in the block, prediction of gas content is conducted through a combination of well log and seismic data.
For this purpose, first, a well log interpretation was conducted on a single well basis to compute the gas content log and determine the distribution and variation in gas content in the vertical direction. At the same time, a seismic processing, inversion and attributes extraction were carried out to estimate the distribution and gas content. Then, single attribute and multiple attribute neural network analyses were conducted to determine the optimal seismic attribute combination for gas content prediction and to establish the relationship between the gas content and these attributes. Finally, this workflow was applied to the whole 3D data volume to determine gas content, which can be utilized for predicting the spatial and areal distribution of the gas content.