Abstract:

Based on method of poststack geostatistical inversion, some key techniques, such as construction of the fine geological model, classification of reservoir and analysis of statistics parameters, equivalent simulation for porosity and geostatistical inversion，have been discussed in details. Aimed at the problem resulted from prediction of carbonate reservoir, the further study procedures which are to optimize systematically the relevant inversion parameters and their reasonable distribution in local reservoirs were carried out in order to get the better prediction precision through well seismic reservoir fine calibration. The prediction results showed this method performed well in spatial distribution of reservoir and nonreservoir; 35 wells were used to test the precision of inversion results; the accumulative thickness of the reservoir when their total porosity is above 1.8% was quantitatively analyzed combined with the local geologic features，it showed high conformity to the real situation of reservoir distribution. The conformity rate is over 80% between the prediction results and practical drilled reservoirs occurrence, so that the prediction results from poststackbased geostatistical inversion have the better precision and approach the actual geologic situation; it has better prediction identification between the reservoir and nonreservoir heterogeneity than conventional wave impedance inversion in practical field operation.

Key words: Halahatang oil field, Xinken area, poststack geostatistics inversion, variation function, reservoir prediction