论文详情
我国致密碎屑岩气藏具有低孔、低渗、储层非均质性强等特点。由于致密气藏埋深较大,目的层地震资料通常信噪比低、主频低,利用单一叠前弹性参数预测有效气层的精度有限。为提高致密气藏的预测精度,以建立适用于低孔、低渗气藏的岩石物理模型为基础,优选出对气层识别敏感的弹性参数组合。通过确定性岩石物理模型和统计岩石物理方法实现测井资料中没有的储层及泥岩的岩石物理表征,并和原测井数据结合建立各类岩性概率密度函数。最终基于地震弹性参数反演数据和概率密度函数,利用贝叶斯岩性判别方法预测岩性流体概率。结果表明:Xu-White模型适用于表征工区低孔、低渗砂岩。运用多弹性参数贝叶斯判别方法能提升岩性识别的成功率。多变量Monte Carlo随机模拟方法有效地拓展了泥岩的岩石物理表征。岩性及流体概率预测结果与实际钻井结果一致,证明了该套方法预测致密碎屑岩气藏的有效性。
The tight clastic gas reservoir in China is characterized by low porosity,low permeability and strong heterogeneity.Meanwhile,due to the deep burial depth,the seismic data of tight clastic gas reservoir owns low signal-to-noise ratio and low dominant frequency.The precision of predicting effective gas layer is limited with the aid of a single pre-stack elastic parameter.In order to improve the prediction precision of tight gas reservoir,the rock physical model adaptive to the gas reservoir with low-porosity and low-permeability was established as a fundamental work.Accordingly,the rock physical model based elastic parameter combination sensitive to gas reservoir was selected.Deterministic rock physical model and statistical rock physics method are applied to expand the rock physical characterization of reservoirs and mudstone that are absent in well logging data.Then,the expanded data and original well logging data are integrated to establish probability density function of every type of lithology.Finally,the Bayesian lithological identification method is utilized to predict the lithology and fluid probabilities on the basis of seismic elastic parameter inversion data and probability density functions.The application results show that the Xu-White model is adaptive to characterize the low porosity and low permeability gas sandstone of the target area.Bayesian lithological identification with multiple elastic parameters is able to considerably improve the prediction precision of lithology.The Monte Carlo stochastic simulation of multi-variations can feasibly expand the mudstone rock physical characterization.The final lithology and fluid probability prediction result highly matches the practical production of exploratory wells,which proves that proposed lithology and fluid predicting workflow is helpful in characterizing the tight clastic gas reservoir.
国家自然科学基金项目“陆相坳陷湖盆浅水三角洲地震沉积学模型”(41272133)资助。