塔里木盆地塔河油田4区和6区奥陶系小尺度缝洞储集体概率识别方法

塔河油田中、下奥陶统小尺度缝洞储集体类型多样，是未来油田开发的重要潜在目标，但储集空间复杂多变，导致常规测井响应特征混杂不清，储集类型识别难度大。根据钻录井特征、岩心、测井等资料对小尺度缝洞储集体类型进行了划分，通过测井信息交会优选了小尺度缝洞储集体的敏感测井信息，利用多元逐步判别原理，建立了小尺度缝洞类型判别函数方程；通过确定各类储集体母体后验概率识别界限，形成了概率约束下的小尺度缝洞储集体识别方法。结果显示，小尺度缝洞储层可划分为5类，分别为未充填溶洞型、砂泥砾充填溶洞型、裂缝-孔洞型、裂缝-孔隙型和溶孔-孔隙型；自然伽马、电阻率和密度是区分各类缝洞储层的最敏感的测井信号；这5类小尺度缝洞储层的临界概率识别界限分别为0.98，0.98，0.60，0.60和0.60。研究区裂缝-孔洞型、裂缝-孔隙型和溶洞型储层大量发育，但单井和平面间的发育程度存在明显差异，4区小尺度缝洞储集体较6区更为发育，垂向上呈现两期岩溶特征。小尺度缝洞识别方法体现了“逐级识别”思路，最大程度上避免了多重测井信息的混淆，能有效提高缝洞识别效果，为深入认识油田小尺度缝洞发育规律及开发潜力提供了有效方法。

Probability-constrained identification of Ordovician small-scale fractured-vuggy reservoirs in Blocks 4-6,Tahe Oilfield,Tarim Basin

Various types of small-scale fractured-vuggy reservoirs occur in the Middle-Lower Ordovician of Blocks 4-6 in the Tahe Oilfield， Tarim Basin， and have been considered as the most potential targets for future oilfield development in the area. However， these reservoirs are difficult to identify due to their complex storage space that induces confusing responses during conventional well logging. To tackle the problem， this study starts with classifying the reservoirs according to drilling and logging data as well as core observation， continues with identifying logging responses typical of the reservoirs based on crossplotting well logging parameters， setting up a discriminant function by using the multivariate stepwise discriminant principle and ends with establishing a method for identifying the reservoirs under probability constraints by determining the parent posterior probability identification limits of all types of reservoirs. The results shows that the reservoirs can be classified into five types： unfilled vug， vug filled with sand， clay and gravel， fracture-vug， fracture-pore and dissolved cave-pore types， which can be recognized respectively with critical probability identification limits of 0.98， 0.98， 0.60， 0.60 and 0.60. These reservoirs are senstive to natural gamma， resistivity and density logging rays. The study suggests that reservoirs of fracture-vug， fracture-pore and cavern types are highly developed in the study area， but their development show great variation vertically in a single well and laterally in the study area. Block 4 develops more fractured-vug reservoirs than Block 6 and shows a vertical two-phase karstification. This identification method embodies well the idea of “step by step identification” and avoids the pitfall of confusing logging signals of these fractured-vuggy reservoirs. Using the method can effectively improve the identification of fractured-vuggy reservoirs， thus providing an effective way for a deeper understanding of the evolution and potential of the reservoirs.