用最大熵谱分解定量预测曲流河薄砂体

Application of maximum entropy spectrum decomposition in quantitative prediction
of thin sand body in meandering river

周宗良 张会卿 曹国明 马跃华 何书梅 周涛

中国石油大港油田公司勘探开发研究院，天津 300280 中国石油东方地球物理公司研究院大港分院，天津 300280 中国石油大港油田公司第一采油厂，天津 30028

Research Institute of Exploration and Development, Dagang Oilfield Company, PetroChina, Tianjin 300280, China Dagang Branch, GRI, BGP Inc., CNPC, Tianjin 300280, China No.1 Oil Production Plant, Dagang Oilfield Company, PetroChina, Tianjin 300280, China

针对大港X开发区曲流河薄砂体定量预测的难题，文中提出了基于井震联合的最大熵谱分解预测方法。通过谱分解算法对比分析选取最大熵谱分解算法进行地震分频扫描，利用响应频率不同分析薄砂体厚度变化，避免了不同频率信息的相互干扰；利用薄层调谐原理制作调谐体，获取调谐频率并计算调谐厚度；通过实钻井数据拟合建立了单砂体厚度￣调谐厚度关系，实现了薄砂体定量预测。新井实施结果证实，该方法能够准确实现曲流河薄砂体定量预测，预测符合率90%以上。

Aiming at the difficult problem of quantitative prediction of meandering river thin sand body in Dagang X development area, we propose a maximum entropy spectrum decomposition prediction method based on well?鄄seismic combination. Firstly, through the comparative analysis of spectral decomposition algorithm, the maximum entropy spectral decomposition algorithm is selected for seismic frequency division scanning, and the thickness variation of thin sand body is analyzed by different response frequencies, thus avoiding the interference of different frequencies information; secondly, the tuning body is made by using the thin layer tuning principle, the tuning frequency is obtained, and the tuning thickness is calculated. Finally, the relationship between thickness of thin sand body and tuned thickness is established by drilling well data fitting, and the quantitative prediction of thin sand body is completed. The results of new wells show that this method can accurately predict meandering river thin sand body, and the prediction coincidence rate is more than 90%.

10.6056/dkyqt201906008