多属性反演技术在储层预测研究中的应用——以珠江口盆地陆丰A区中新统珠江组储层为例

珠江口盆地陆丰A区中新统珠江组中-下段2370层是该地区主力油层之一, 但储层厚度差异大、分布不清, 严重制约了区域勘探部署.本文综合地震、测井和地质资料, 分析得出纵波阻抗和V_P/V_S为研究区识别岩性的敏感参数, 并通过叠前同步反演技术刻画出研究区2370层砂岩分布特征, 整体呈北厚南薄的展布.以地震振幅、分频属性及叠前反演属性为基础, 利用多属性反演技术分析统计了各属性与孔隙度之间的关系, 优选出密度、V_P/V_S、地震导数、LR、V_P、25~30Hz地震分频6种属性为储层物性识别参数, 综合预测研究区2370层储层滨岸砂坝砂体孔隙度发育特征.多属性反演结果表明, 在研究区东北部、西部, 厚层的滩坝储集砂体的孔隙度发育, 普遍在18%以上; 而在中部和东部, 受滩间、坝间泥岩与滩坝砂的叠合区影响呈条带状分布, 孔隙度为10%~18%.预测结果与同期钻井的钻后结果基本一致, 预测成功率在75%以上.多属性反演技术在研究区的储层预测中表现出良好的有效性和实用性, 对类似地区的储层预测也具有借鉴意义.

Application of multi-attribute inversion method in reservoir prediction: A case study of Lufeng A area in the Pearl River Mouth Basin

The 2370 layer of Miocene Zhujiang Formation is one of the major reservoir in Lufeng A area, Pearl River Mouth Basin. But the large thickness difference and complicated distribution of sandstone seriously restrict the field exploration and deployment. This paper integrate s the seismic data, logging and geological data, and concludes that the P-wave impedance and V_P/V_S are the sensitive parameters of lithology identification. The distribution of 2370 layer of sandstone is described by using prestack simultaneous inversion method, and the north area is thicker than the south on the whole. On the basis of seismic amplitude and the properties of frequency division and prestack inversion, the relationship between the property and porosity is analyzed by using multi-attribute inversion method. The reservoir identification parameters are six properties, which contain the density, V_P/V_S, seismic derivative, LR, V_P and 25~30 Hz seismic frequency. The results of multi-attribute inversion show porosity generally above 18% in the northeast and the west of the study area, which belongs to the reservoir type of beach sand and dam sand, and it is a thick layer. In the middle and east part, the porosity is 10%~18%, which is the influence of the interaction between the sandstone and mudstone. Porosity prediction results are consistent with the drilling results of the same period, and the predicted coincidence rate is more than 75%. Multi-attribute inversion method shows good effectiveness and practicability in reservoir prediction of the study area, and it also has reference significance to the similar area of reservoir prediction.