关键词: 核磁共振录井, 优化实验, 可动水信号比, 雷达网络评价模型, 流体评价

Abstract: NMR logging technology plays an important role in compound logging, providing effective guidance and evaluation basis for reservoir physical property evaluation and fluid property discrimination. However, for light oil reservoirs with strong volatility, factors such as sampling method and analysis time can have a great impact on the evaluation results of NMR logging. The traditional sampling and analysis process limits the application effects of NMR logging technology. By conducting NMR logging sampling method, analysis time window and other experiments, the sample sampling and analysis process has been optimized, ensuring the accuracy of basic data and providing data guarantee for the research of NMR logging fluid evaluation technology. Based on NMR triple scan analysis, the quantitative parameter of movable water signal ratio for water content evaluation was defined according to the fluid property evaluation principle of T₂ spectrum and mathematical derivation. NMR radar network evaluation model was established based on the multivariate data area distribution discriminant method. The optimized technology improves the accuracy for analysis parameters and fluid property discrimination, and the coincidence rate of interpretation and evaluation reaches 90%, which has high practicability, laying a solid foundation for the in-depth promotion of NMR logging technology in Junggar Basin.

Key words: NMR logging, optimization experiment, movable water signal ratio, radar network evaluation model, fluid evaluation