靖西地区马家沟组高阻水层影响因素分析与识别方法

靖西地区马家沟组气水关系复杂，不同流体间的地球物理测井响应差异微弱，给传统测井解释工作带来了困难。本文以地球物理测井、薄片、扫描电镜、压汞等资料为基础，探究了储集空间结构以及地层特征对高阻水层响应特征的影响，深入分析了高阻水层成因，建立了常规地球物理测井与电成像测井资料相结合的随机森林流体识别模型。结果表明，靖西地区马家沟组储层复杂的孔隙结构和广泛发育的薄互层是致使水层地球物理测井响应表现为高阻特征的主要因素；建立的随机森林流体识别模型能有效地解决高阻水层问题，判别精度达到84 %并在验证盲井中表现稳定。研究结果有效地提高了该地区的测井解释符合率，为后期油藏评价与勘探开发奠定了基础。

Analysis and Identification Method of High-resistivity Water Layer in Majiagou Formation of Jingxi Area

The phenomenon of high-resistivity water production is ubiquitous in the Majiagou Formation of Jingxi area, the gas-water relationship in the block is complex, and the difference in logging response is weak, all of which lead to difficulties to traditional logging interpretation. Based on geophysical logging, thin slices, scanning electron microscopy, mercury intrusion and other data, the influence of reservoir space structure and stratigraphic characteristics on the response characteristics of the high-resistivity water layer was explored. The cause of the high-resistivity water layer was deeply analyzed. A Random Forest fluid identification model combining conventional geophysical logging and electrical imaging logging data was established. The results show that the complex pore structure and widely developed thin interbeds of the Majiagou Formation reservoir in Jingxi area are the main factors that caused the water layer geophysical logging response to high resistance characteristics. The established Random Forest fluid identification model can effectively solve the problem of the high-resistivity water layer with a discrimination accuracy of 84 % and perform stably in the verification of blind wells. The research result can effectively improve the coincidence rate of logging interpretation, laying a foundation for later reservoir evaluation and exploration and development.