开发前期少井条件下低渗储层“甜点”建模表征技术及其应用

Modeling and characterizing technique of the“sweet spot”and its application in low-permeability reservoirs with limited wells at the early stage of the development

海上A气田发育河流相,储层非均质性强且该气田目前处在开发前期研究阶段,井数少、井距大,地震资料无法直接有效识别低渗储层中的“甜点”,给储层表征带来很大的挑战。针对此问题,提出了在地震反演资料约束下,在地震—地质一体化思路指导下,根据可靠性程度不同的储层预测结果,分别采用地震约束和和多属性融合神经网络技术建模方法,建立河流相“甜点”和物性模型。结果表明,该方法可以较好地反映储层内部非均质性,通过对不同渗透率级别进行储量分类评价,有效地对“甜点”富集区进行了筛选,为井位优化和开发方案设计提供了合理的地质依据,从而达到降低气田开发风险的目的。

Fluvial facies reservoirs are well developed in offshore Gas Field A,which are characterized by strong heterogeneity.At present,the field is at the early stage of development and research,which has limited wells and large well spacing,and moreover the seismic data can not directly and effectively identify the“sweet spot”in the low-permeability reservoirs,resulting in a great challenge for the reservoir characterization.In view of the existing problems,constrained by the seismic inversion data and guided by the idea of seismic-geological integration,according to the predicted reservoir results with different reliabilities,the models for the fluvial-facies“sweet spot”and physical properties were established with the modeling methods of the seismic constraint and multi-attribute-integration neural network.The results show that the method can well reflect the inner heterogeneity of the reservoirs;by means of the reserves classification and evaluation for different permeability levels,the“sweet spot”enriched areas were effectively screened,and the reasonable geological evidences were presented for the well location optimization and development plan design,thus the purpose of reducing the risk of the gas field development can be realized.