地质统计反演方法在新木油田128区块砂体预测中的应用
|
摘要
以约束稀疏脉冲反演结果为背景约束,进行高分辨率的自然电位地质统计反演。首先,进行直方图和变差函数分析,确定自然电位在不同方向上的变程。然后,在每一道地震道上,将随机提取的反射系数与地震子波进行褶积,生成的合成地震道与原始地震道进行匹配。计算反演残差,直到匹配精度达到满意为止。随机实现4个反演结果,对这4个模拟结果进行求均方根运算,得到最终的反演结果。最后,通过设定砂泥岩门限值、岩性数据提取,砂岩速度场建立,利用数学函数计算,得到深度域的岩性体和5个砂组砂岩平面分布图。实现了砂体的定量预测。其中95%的后验井砂岩厚度的相对误差都在10%以内,能够有效的为钻井及层系开采决策提供依据。
With the constrained sparse pulse inversion results as the background constrained,high resolution of spontaneous potential geo-statistic inversion was launched.First,with histogram and variogram analysis,the spontaneous potential change process in different directions was determined.Second,in every seismic trace,in order to get synthetic seismic trace,reflection coefficient and seismic wavelet were extracted randomly into deconvolution.Then generating synthetic seismic trace was matched with the original seismic trace,to calculate inversion residuals,until meeting precision requirements.Four random inversion results was gained for RMS operations,to get the final inversion results.Finally,by setting the threshold value of sand shale,lithologic data extraction and sandstone velocity field establishment,mathematical function is used to get the rock body in depth domain and sandstone distribution in map of 5 sand group.With these process,the sand body quantitive predicting was realized.The result show that the relative error of 95% post-test well sandstone thickness is less than 10%.Therefore it can provide an effective basis of drilling and strata exploration decision-making.
With the constrained sparse pulse inversion results as the background constrained,high resolution of spontaneous potential geo-statistic inversion was launched.First,with histogram and variogram analysis,the spontaneous potential change process in different directions was determined.Second,in every seismic trace,in order to get synthetic seismic trace,reflection coefficient and seismic wavelet were extracted randomly into deconvolution.Then generating synthetic seismic trace was matched with the original seismic trace,to calculate inversion residuals,until meeting precision requirements.Four random inversion results was gained for RMS operations,to get the final inversion results.Finally,by setting the threshold value of sand shale,lithologic data extraction and sandstone velocity field establishment,mathematical function is used to get the rock body in depth domain and sandstone distribution in map of 5 sand group.With these process,the sand body quantitive predicting was realized.The result show that the relative error of 95% post-test well sandstone thickness is less than 10%.Therefore it can provide an effective basis of drilling and strata exploration decision-making.
引文
1边树涛,狄帮让,董艳蕾,等.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.石油地球物理勘探,2010;45(03):398—405
2王健,谢华锋,董洪奎.随机地震反演在岩性油气藏储层预测中的应用.科学技术与工程,2011;11(12):2640—2648
3李方明,计智锋,赵国良,等.地质统计反演之随机地震反演方法-以苏丹M盆地P油田为例.石油勘探与开发,2007;34(4):451—455
4鲁健.新木采油厂让29-让41井区扶余油层断裂与砂体匹配控油研究.大庆:东北石油大学,2011:40—44
5 Bortoli L J,Alabert F,Hass A,et al.Constraining stochastic imagesto seismic data.Troia.International Geostatistics Congress,1992
6 Hass A,Dubrule O.Geostatistical inversion-A sequential method ofstochastic reservoir modeling constrained by seismic data.First Break,1994;13(12):561—556
7李庆忠.论地震约束反演的策略.石油地球物理勘探,1998;33(4):423—438
2王健,谢华锋,董洪奎.随机地震反演在岩性油气藏储层预测中的应用.科学技术与工程,2011;11(12):2640—2648
3李方明,计智锋,赵国良,等.地质统计反演之随机地震反演方法-以苏丹M盆地P油田为例.石油勘探与开发,2007;34(4):451—455
4鲁健.新木采油厂让29-让41井区扶余油层断裂与砂体匹配控油研究.大庆:东北石油大学,2011:40—44
5 Bortoli L J,Alabert F,Hass A,et al.Constraining stochastic imagesto seismic data.Troia.International Geostatistics Congress,1992
6 Hass A,Dubrule O.Geostatistical inversion-A sequential method ofstochastic reservoir modeling constrained by seismic data.First Break,1994;13(12):561—556
7李庆忠.论地震约束反演的策略.石油地球物理勘探,1998;33(4):423—438
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心