Application of Well Controlled Principal Component Analysis in Reservoir Prediction

Key words: 
• lithology distinguish /
• principle component analysis (PCA) /
• sandstone reservoir /
• seismic attribute analysis

Abstract: The igneous rock of shallow layer developed in NanPu Depression has many types and strong heterogeneity laterally and vertically present, which are difficult to describe the spatial distribution of igneous rocks and sandstone reservoirs. As the identification of the igneous rock distribution is very critical for the confirmation of sandstone gas/oil reservoir development feature, so, multiple technologies integration is urgent need instead of only by single geology and geophysical technology. The prediction of sandstone reservoirs restricts the exploration and development of sandstone reservoirs in shallow layers. In this study, based on the rock physical property analysis, combined P-wave impedance, Gamma, Poisson ratio and Sweet seismic attributes which are sensitive to various lithology and PCA dimension reduction technology, four lithologies which are basalt, altered volcanic rock, sandstone and mudstone have been effectively distinguished. And lithology cube is built by using innovative technology of well constrained interpretation, so that, the identification of sandstone reservoir with high accuracy can be achieved. From the comparison, the final seismic prediction result is consistent with the regional geology distribution feature, and with coefficient of 87.1% matching with well data. The distribution of igneous rocks and sandstones are matched with regional geology.