Application of geostatistical inversion in reservoir prediction in the third member of Shahejie Formation,Baimiao Gas-field,Dongpu Depression
Bian Shu-tao1,2, Di Bang-rang1,2, Dong Yan-lei1, Jiang Yi-wei3, Wang Le-zhi3, Liao Lin4,5
1. College of Geosciences,China University of Petroleum,Beijing City,102249,China;
2. CNPC Key Lab of Geophysical Prospecting,China University of Petroleum,Beijing city,102249,China;
3. Research Institute of Petroleum Exploration and Development,Zhongyuan Oil-field Branch Company of Sinopec,Puyang City,Henan Province,457001,China;
4. Department of Geosciences,Zhejiang University,Hangzhou City,Zhejiang Province,310027,China;
5. Structural Research Centre of Oil & Gas Bearing Basin,Ministry of Education,Hangzhou city,Zhejiang Province,310027,China
Abstract：In this paper based on constrained sparse spike inversion the geostatistical inversion was conducted with the normalized natural potential curve as input and the impedance data volume obtained from constrained sparse spike inversion as correlation control function.At first calculating random variable's histogram distribution and variogram,determining random variable's range in different directions,then in each seismic trace by convoluting extracted reflection coefficient with estimated seismic wavelet to obtain synthetic seismogram,by matching the synthetic seismogram with original seismic trace to calculate residual value for the inversion until matching degree meeting the accuracy requirement,at last by choosing the best node value for the synthetic seismogram as inversion result to conduct inversion for the next randomly selected node until finishing the whole inversion,after that conducting root mean square calculation for each realization whose correlativity is larger than 90% until completing whole inversion of the random process.In this paper the submarine fan in the third member of Shahejie Formation,Baimiao Gas-field was taken as study object,based on the analysis for reservoir's lithology,electrical properties and physical characteristic,the seismic spatial-variable wavelet was chosen and the low frequency model was built up,by selecting reasonable inversion parameterλand taking normalized natural potential curve as input,the inversion was conducted by using constrained sparse spike inversion and geostatistical inversion,the matching rate between inversion result and the drilling result reaches 81.3%,therefore excellent application result was achieved in the paper.
边树涛, 狄帮让, 董艳蕾, 姜贻伟, 王乐之, 廖林. 地质统计反演在东濮凹陷白庙气田沙三段储层预测中的应用[J]. 石油地球物理勘探, 2010, 45(3): 398-405.
Bian Shu-tao, Di Bang-rang, Dong Yan-lei, Jiang Yi-wei, Wang Le-zhi, Liao Lin. Application of geostatistical inversion in reservoir prediction in the third member of Shahejie Formation,Baimiao Gas-field,Dongpu Depression. OGP, 2010, 45(3): 398-405.
