论文部分内容阅读
波阻抗反演是储层预测的主要方法之一.然而,当声波曲线和其计算出来的波阻抗不能很好地区分砂泥岩时,尤其是在研究区发育砂泥岩薄互储层的情况下,就会给储层预测带来困难.针对这一问题,本文以尼日尔Fana地区为例,采用声波测井曲线重构的方法进行测井约束反演,使储层的波阻抗特征更加明显.然后,对得到的波阻抗体作为输入进行神经网络岩性反演,结果精细刻画了研究区内砂体的空间分布和几何形态,提高了地震方法识别薄储集层的精度,指导了该区下一步勘探.
Wave impedance inversion is one of the main methods of reservoir prediction.However, when the acoustic wave curve and its calculated wave impedance can not distinguish sandstone well, especially in the development of sandstone and mudstone thin inter-reservoir , This will bring difficulties to the reservoir prediction.In this paper, taking the Fana area in Niger as an example, the logarithmic inversion method is used to reconstruct well logging constants, which makes the impedance characteristics of the reservoir more obvious. Then, the retrieved wave impedance body is used as the input to carry on the neural network lithologic inversion. The results refine the characterization of the spatial distribution and geometry of the sand body in the study area and improve the accuracy of the seismic method in identifying thin reservoirs, guiding the area Next exploration.