利用聚类分析方法研究储层流体类型

On Cluster Analysis for Reservoir Fluid Property

随着川渝油气田天然气勘探开发的不断深化,川中地区某构造雷口坡组储层目前成为勘探开发的重点。该构造裂缝孔洞型白云岩储层在横向上分布较稳定,但裂缝、孔、洞并存,孔隙结构、气水关系复杂,基岩孔隙度低,裂缝影响大。利用常规测井、电成像和阵列声波测井资料来判别储层流体类型效果不佳,解释符合率低。针对该构造独特的地质特点,充分利用测井、岩屑、气测、岩心、试气等各种有用信息,开展用聚类分析方法进行储层分类、判别储层流体类型的研究。经过26口井的测试检验,解释符合率大于90%,效果显著。

The structure LG×× located in the central region of Sichuan becomes the focus of ex- ploration and development with the exploration of natural gas constant deepening in oil ＆ gas field. The dolomite reservoir with fracture-pore types is distributed stably in the structural LG. But, the reservoir space contains multiplex types, e.g. fracture-type, pore-type and erosion pore- type. Further more, pore configuration and air-water relationship are complex going with low basal-rock porosity and fracture-type reservoir space playing a leading role. So, we acquire low log interpretation coincidence rate by using conventional logging, electrical imaging logging and array acoustic logging to distinguish reservoir fluid property. We can use well logging data, drill- ing cuttings describing data, oil testing data, and drill core data, etc. , for establishing multi-pa- rameter cluster analysis system to classify reservoir types and distinguish fluid property against the structure＇s unique geological features. After checking oil testing data in 26 wells, we get re- markable effect with the log interpretation coincidence rate exceeded 90%.