密井网区叠前高分辨率反演及其在薄储层预测的应用

Pre-stack high-resolution inversion for dense well pattern area and its application in thin reservoirs prediction

针对大庆长垣油田密井网区SⅡ油层组陆相河流-三角洲相沉积微相类型变化快、井间厚度2 m以下的窄小河道砂体预测误差大、河道走向预测多解性强、砂体间连通关系判别差异性大等问题,以长垣油田A区块为例,开展叠前井控道集处理技术、基于改进的Xu-White岩石物理模拟技术、基于“波形指示马尔科夫链-蒙特卡洛”随机模拟算法的叠前波形指示反演技术研究。结果表明:研究区砂岩储层具有低纵波阻抗、低纵横波速度比值的特征,纵波阻抗为3.8~8.2 Gg/(m^(2)·s),纵横波速度比为1.7~2.1;随时砂岩厚度的增加,其纵波阻抗及纵横波速度比有减小的趋势。通过70口后验井的定量分析及评价,在密井网区叠前波形指示反演对厚度为[1,2)m的薄互层砂体识别精度为80.7%,对厚度大于等于2 m的砂体识别精度为85.1%。研究成果对河流―三角洲相薄储层开发方案的合理制定及措施调整具有指导意义。

In view of the problems of SⅡreservoir member in dense well pattern area of Daqing Placanticline oil-field,including rapid changes in sedimentary microfacies types of continental fluvial-delta facies,much prediction error of narrow channel sand bodies with thickness<2 m between wells,high ambiguity of channel trend prediction and significant differences in distinguishing sand bodies connectivity,Block A of Placanticline oilfield is taken as an example to perform research on pre-stack well-controlled gathers processing technique,rock physics simulation technique based on improved Xu-White model and pre-stack waveform indication inversion technique based on“waveform indication Markov chain-Monte Carlo”random simulation algorithm.The results show that sandstone res-ervoirs in studied area have characteristics of low P-wave impedance and low P-S wave velocity ratio,with P-wave impedance values of 3.8~8.2 Gg/(m^(2)·s),and P-S wave velocity ratio of 1.7~2.1.With the increase of sandstone thickness,its P-wave impedance value and P-S wave velocity ratio tend to decrease.Based on quantitative analysis and evaluation of 70 posteriori wells,identification accuracy of pre-stack waveform indication inversion for[1,2)m thin interbedded sand bodies in dense well pattern area is 80.7%,and identification accuracy of sand bodies with thicknesses≥2 m is 85.1%.The research provides guiding significance for rational design and adjustment of stimula-tion solutions for development plan of thin fluvial-deltaic reservoirs.