叠后地质统计学反演在H地区储层预测中的应用
Application of post-stack geo-statistical inversion in reservoir prediction in H area
刘灵 1徐乐意 1陈人杰 1朱焕 1谢雨 1邴琦 1杨帆1
作者信息
- 1. 中海石油(中国)有限公司深圳分公司研究院,广东深圳 518054
- 折叠
摘要
H地区A3 层是珠江口盆地惠州凹陷寻找岩性圈闭最有利的地区之一,但A3 层砂岩厚度薄,叠后地震资料和常规叠后确定性反演结果纵向分辨率低,难以准确识别砂岩的尖灭位置和平面展布.采用叠后地质统计学反演方法,利用地质、测井、地震等信息和贝叶斯概率密度函数建立统计学模型,同时结合马尔科夫链-蒙特卡洛算法进行随机模拟,得到了高精度、高分辨率的反演阻抗体和岩性体;相比于常规确定性反演,地质统计学反演结果高频细节更清晰,分辨率提高了一倍,有效识别出了A3 层薄层砂岩的走向和尖灭位置,为H地区岩性油藏的储层预测提供了重要依据.
Abstract
Layer A3 in H area is one of the most favorable areas for looking for lithologic traps in Huizhou sag of Pearl River Mouth Basin.However,the thickness of the A3 is thin,and the longitudinal resolution of post-stack seismic data and conventional post-stack deterministic inversion results is low,making it diffi-cult to identify the pinch-out position and plane distribution.With the post-stack geo-statistical inver-sion method,statistical model is established by using geological,logging,seismic information and Bayesian probability density function.The random simulation is carried out by Markov chain-Monte Carlo algorithm,and then high precision and high resolution inversion impedance and lithology are obtained.Compared with conventional deterministic inversion,geo-statistical inversion results have clearer high-frequency details,doubled resolution,and identified the strike and pinch-out position of the A3 thin layer sandstone,provi-ding a good reference for reservoir prediction of lithological reservoirs in H area.
关键词
地质统计学反演/贝叶斯理论/MCMC随机模拟/薄层砂体识别Key words
geo-statistical inversion/Bayesian theory/MCMC random simulation/thin sand body identifi-cation引用本文复制引用
基金项目
中海石油重大科技项目(十四五)(KJGG2022-0403)
出版年
2024
国家科技期刊平台
NSTL
万方数据