基于高分辨率层序约束的薄储层地震预测方法研究

Thin reservoir seismic prediction method based on high resolution sequence constraint

塔里木油田碎屑岩储层埋深大,储层物性好且油气资源含量大,但面临砂岩和泥岩频繁交互发育、岩性横向变化快、单储层厚度较薄的难题.常规地震反演技术由于低频初始模型精度低和横向约束性弱,难以满足深部薄储层预测需求.本文基于贝叶斯反演理论和INPEFA(Integrated Prediction Error Filter Analysis)层序划分理论,提出了一种高分辨率层序约束的叠后贝叶斯反演方法.该方法首先通过INPEFA技术建立井上高分辨率层序格架,结合序贯高斯模拟方法建立精度高低频初始模型,同时在高分辨率层序约束下利用相邻道的相近性,在地震反演中引入横向约束算子,最终获得高分辨率和高稳定性的反演结果.该方法在轮古地区深部碎屑岩储层预测应用,有效提高了薄储层成像精度,纵向结果可识别3 m的薄储层,同时能准确刻画砂体的横向变化情况,解决了该地区深部碎屑岩薄储层预测的难题,为研究区进一步的勘探部署提供了有力支撑.

The Tarim oilfield has a large depth of clastic reservoirs with good physical properties and large hydrocarbon resources,but faces the difficulties of frequent interactive development of sandstone and mudstone,rapid lateral change of lithology,and thin thickness of a single reservoir.Conventional seismic inversion technology is difficult to meet the demand of deep thin reservoir prediction due to the low accuracy of low-frequency initial model and weak lateral constraints.In this paper,based on the Bayesian inversion theory and INPEFA(Integrated Prediction Error Filter Analysis)layer order division theory,a high resolution layer order constrained post-stack Bayesian inversion method is proposed.The method firstly establishes a high-resolution layer sequence grid on the wells by INPEFA technology,and establishes a high-accuracy low-frequency initial model by combining with the sequential Gaussian simulation method,and at the same time introduces lateral constraints operator in the seismic inversion by utilizing the proximity of the neighboring traces under the constraints of the high-resolution layer sequences,so that the inversion results of high-resolution and high-stability can be obtained in the end.The application of this method for deep clastic reservoir prediction in the Lungu area effectively improves the imaging accuracy of thin reservoirs,and the longitudinal results can identify 3 m thin reservoirs,while accurately portraying the transverse changes of the sand body,which solves the difficult problem of predicting the thin reservoirs of deep clastic rocks in the area,and provides a strong support for the further exploration and deployment in the study area.