Reservoir Pressure Prediction Using Time-Lapse Seismic Multi-Attribute Analysis

摘要

Time-lapse seismic is used to reveal subtle signals which are representative of reservoir variation. With the widespread application of 4D technology, many attentions have been paid on the quantitative prediction of reservoir variation via time-lapse seismic data. Artificial Neural Network(ANN) method is usually applied to the unstructured computation, which is hardly expressed with a definite function. It shows great advantages in non-linear relationship building. In this article, the author introduces fuzzy self-organizing neural network, it can achieve unsupervised clustering and classification of seismic attributes automatically. Hereafter reservoir parameters are estimated with another neural network. Based on the synthetic model built from the real data, pore pressure change in a gas reservoir is computed. Results show that this method is practical and efficient.