基于交会图决策树的缝洞体类型常规测井识别方法——以塔河油田奥陶系为例

塔河油田奥陶系具有丰富的油气资源，受岩溶作用影响发育洞穴、孔洞和裂缝等多种储集空间，洞穴中充填不同物质。多种储集空间和充填物组合形成多种缝洞体类型，导致常规测井响应特征复杂，识别类型难度大。依据露头、岩心和井壁成像测井等信息，确定了6种缝洞体类型，分别为未充填洞穴型、砂泥充填洞穴型、角砾充填洞穴型、方解石充填洞穴型、缝洞型和裂缝型。结合不同类型缝洞体之间岩性和物性差异，对比分析不同类型缝洞体的双侧向、自然伽马、声波、中子、密度测井响应特征及分布规律。将测井响应两两组合建立交会图，优选出9个识别缝洞体类型敏感的测井响应交会图版，结合决策树方法，形成逐级识别缝洞体类型方法。应用该方法对塔河油田实际井资料进行处理解释，识别结果与岩心和录井对比，符合率均大于85%。该识别方法减少不同类型缝洞体常规测井响应之间相互混淆的影响，有助于发挥每个常规测井响应的作用，提高识别符合率。

Conventional logging identification of fracture-vug complex types data based on crossplots-decision tree:A case study from the Ordovician in Tahe oilfield,Tarim Basin

The Ordovician in Tahe oilfield are rich in oil and gas resources.It develops caves,vugs and fractures filled with different materials due to karstification.Different reservoir spaces and fillers form a variety of fracture-vug complex types.So the conventional logging responses are complex and it is very difficult to identify the fracture-vug complex types.Based on outcrops,cores,borehole image logging data,six fracture-vug complex types are determined.They are unfilled cave type,sand and shale-filled cave type,breccias-filled cave type,calcite-filled cave type,fractured-vuggy type and fractured type,respectively.In combination with the differences of lithologies and physical properties between different fracture-vug complex types,their characteristics of dual lateral,natural gamma,sonic,neutron and density logging and distribution patterns are analyzed.Nine crossplots sensitive to fracture-vug complexes are selected from the pairwise logging response cross plots and are integrated with decision tree to formulate a step by step fracture-vug complex type identification method.This method is applied to real well data processing and interpretation in Tahe oilfield.Compared with of cores and geologic logging,the coincidence rates of the identification results with core and geologic logging data are greater than 85%.The method can reduce the confusion among the conventional logging responses of different fracture-vug complex types,thus increase the accuracy of identification results.