Geological Statistics Inversion And Its Application In Carbonate Reservoir Prediction

Carbonate reservoir is an important type of oil and gas reservoir. And half of oil and gas reserves come from carbonate reservoirs. During reservoir evaluation, the diverse rock type, multiple causes, strong heterogeneity and carbonate instability lead to complex reservoir distribution. Therefore, accurate carbonate reservoir prediction and interpretation are very necessary for oil and gas production. Seismic impedance inversion technology is an important means and widely used for reservoir prediction. Due to applicable condition, an inversion method isn’t suitable for all the reservoir prediction and need be tested carefully for specific research area. And the inversion method and processing parameters have important influence on the final prediction results. Constrained sparse spike inversion(CSSI) method is employed on post stack seismic data, which provide lateral continuous distribution and poor vertical resolution information of reservoir prediction. Based on geostatistics analysis, stochastic simulation technique and priori geological information, geostatistical inversion(GI) method use both seismic data and logging data, which renders accurate reservoir body with lateral continuous features and vertical high resolution. Carbonate rock reservoirs in the study area are thin interbedding structure. So the GI inversion is employed to recognize thin sand body and discontinuous sand body.Aiming at field data of block A from Turkmenistan, a comprehensive seismic inversion application is studied and related parameters are tested to obtain reservoir information. First of all, seismic inversion methods are introduced, and their applicable conditions are analyzed. Secondly, geostatistical inversion method and process steps are detailed, and some related parameters and modeling method are summarized, such as the variable difference function calculation method, fitting method and analysis of variable difference function parameter values. Thirdly, some important data processing steps are detailed and analyzed, including seismic data preprocessing, standardization of logging data, time-depth calibration, seismic wavelet and statistical analysis of reservoir parameters. Some impact factors are also discussed.Both CSSI and GI methods are used to predict reservoir. And similar reservoir bodies are achieved. The results show that the predicted results were very consistent with logging data and laterally continuous bodies. But the reservoir by GI method can provide higher resolution information than that by CSSI, which can achieve a clear identification of thin sand body and discontinuous small sand body distribution. An accurate reservoir prediction provides a basis for understanding the reservoir body distribution.