Gas-Chimney Detection in 3D Seismic by Neural Network

摘要

Recognition of hydrocarbon migration is so vital for petroleum exploration. Developing intelligent systems (artificial neural network) enable experts to achieve more details from seismic data. Although detection of migration direction using seismic data is difficult, Chimney-cube analysis overcomes this problem. The authors used several filters, seismic attributes, neural network (supervised and unsupervised), and interpreters' viewpoints. In supervised method artificial and human intelligence cover their limitations and in unsupervised method the authors eliminate the experts' views. Chimney recognizes the migration direction and locates the spill points, mud volcanoes, gas seepages, sealing, and nonsealing faults and finally the origin of hydrocarbon.