超深断控缝洞型储层迭代反演方法——以富满油田为例

Iterative inversion method for ultradeep fault-controlled fracture-vug reservoirs:A case study of the Fuman oilfield,Tarim Basin

塔里木盆地塔河南岸富满油田超深层奥陶系灰岩地层缝洞型油藏现如今是油田产能建设的核心靶区。研究区储层是受走滑断裂作用控制地层破裂形成的大量溶蚀孔洞,非均质性强。而常规波阻抗反演的低频模型是由层位、断裂和测井数据建立起来的,不能表征储层的断控非均质性。为此,提出以“断裂破碎带相”为约束的迭代反演方法来预测储层,其技术流程是先对原始地震数据进行解释性处理,得到能反映储层断裂破碎带特征的属性体,将其与常规波阻抗反演得到的优质储层空间轮廓数据和初始低频模型进行比例融合,生成新的非均质性低频模型,进而开展多轮次迭代反演。反演结果可真实地刻画非均质储层断裂破碎带特征,10口井的预测吻合率可达92.86%。应用表明,该方法改善了储层预测效果,提高了断控储层的预测精度,有效地支撑了研究区的工作。

The ultradeep Ordovician limestone fracture-vug reservoirs in the Fuman oilfield on the south bank of the Tahe River in the Tarim Basin is the core target area for the production capacity construction of the oilfield.These reservoirs in the study area contain massive dissolution vugs formed by the formation fracturing due to the strike-slip faulting and thus are highly heterogeneous.The low-frequency models based on conventional wave impedance inversion are built using data on horizons,faults,and logs,and thus they cannot characterize the fault-controlled heterogeneity of the reservoirs.For this reason,this study proposed an iterative inversion method constrained by fault fractured zone facies to predict reservoirs.The technical process of this method is as follows.First,the original seismic data were interpreted to obtain the attribute volumes that can reflect the characteristics of the fault fractured zones of the reservoirs.Then,the attribute volumes were proportionally fused with the initial low-frequency model and the spatial profile data of high-quality reservoirs obtained from the conventional inversion based on wave impedance.As a result,a new heterogeneous low-frequency model was formed.Using this model,multiple rounds of iterative inversions were conducted.The inversion results can truly describe the characteristics of the fault fractured zones in the heterogeneous reservoirs.The prediction results of 10 wells had coincidence rates of up to 92.86%.As indicated by the application,the method proposed in this study can improve the reservoir prediction performance and the prediction precision of fault-controlled reservoirs,thus effectively supporting the work in the study area.