PREDICTION OF SAND BODIES IN DEEP-WATER CHANNEL TURBIDITE RESERVOIR WITH SEISMIC ATTRIBUTES
-
摘要:
深水水道型储层蕴含了丰富的油气资源,但海上油气勘探往往由于井数量不足导致地下储层预测困难。基于对深水水道的地质认识,构建了不同成因类型的水道概念模型,设计了井点处砂体叠置样式,并通过正演模型分析地震响应特征及地震属性敏感性,结合多种敏感地震属性进行聚类分析及概率神经网络预测。将该方法应用于西非X油藏,预测结果符合率达到87.5%,能够有效预测该油藏的砂体展布特征。预测结果结合地震沉积学综合研究认为,目的层内砂体主要分布于水道中部,两端分布较少,其主要受控于沉积环境和弯曲段流态变化。
Abstract:Deep-water channel sand bodies are good reservoir for oil and gas accumulation. However, the prediction of such a reservoir is difficult since usually no sufficient wells are available offshore. Following the geological characteristics of deep-water channels, this paper constructed some models for different types of channels and proposed some sand body superposition patterns at the well point using forward model analysis, the seismic response characteristics and sensitive seismic attributes as tools. Cluster analysis and probabilistic neural network (PNN) prediction are carried out based on sensitive seismic attributes. The X reservoir in West Africa is selected as a case for application, and the coincidence rate of the prediction results reached as high as 87.5%, which is efficient enough to predict the sand body distribution pattern for oil and gas exploration. Combined with the integrated study of seismic sedimentology, it is concluded that the sand bodies in the target layer are mainly distributed in the middle of the channel but few in both ends under the control of the sedimentary environment and the change in flow regime in the curved part of channel.
-
Key words:
- deep-water channel /
- seismic attributes /
- sand body architecture /
- sand body prediction /
- neural network
-
-
表 1 模型参数表
Table 1. Model parameters
主频/
Hz砂岩速度/
(m/s)砂岩密度/
(g/cm3)泥岩速度/
(m/s)泥岩密度 /
(g/cm3)调谐厚度/
m30 2 841 2.37 2 561 2.41 23.70 表 2 模型厚度表
Table 2. Thickness used in modeling
m 纯泥 单期
薄砂单期
厚砂厚薄砂互层 两期
厚砂薄厚
砂体多期
砂体砂厚 1 3 7 7 7 3 7 泥厚 5 - - 3 3 3 3 砂厚 1 - - 3 7 7 7 泥厚 5 - - - - - 3 砂厚 1 - - - - - 7 砂体累计厚度 3 3 7 10 14 14 21 注:-代表无数据。 表 3 井点砂体垂向叠置样式表
Table 3. Sand body superimposing pattern at well point
井名 实际砂体样式 预测砂体样式 是否符合 总符合率 X31 1 训练井 87.5% X39 3 训练井 X3ST 3 训练井 X42 4 训练井 X19 3 训练井 X34 2 训练井 X27 4 4 是 X52 1 1 是 X16 4 4 是 X2 4 4 是 X20 4 4 是 X25 4 4 是 X29 3 3 是 X56 3 1 否 -
[1] 赵晓明,葛家旺,谭程鹏,等. 深海水道储层构型及其对同沉积构造响应机理的研究现状与展望[J]. 中国海上油气,2019,31(5):76-87.
[2] ROYA,MATOS M,MARFURT K J. Automatic seismic facies classification with Kohonen Self organizing maps:a tutorial[J]. Geohorizons Journal of Society of Petroleum Geophysicists,2010,10(1):6-14.
[3] 梁平. 地震属性优化技术及BP小波神经网络储层参数预测的初步研究[D]. 成都: 成都理工大学, 2009.
[4] MUTTI E,NORMARK W R. Comparing examples of modern and ancient turbidite systems:problems and concepts[J]. Marine Clastic Sedimentology:Concepts and Case Studies,1987:1-38.
[5] 林煜,吴胜和,王星,等. 深水浊积水道体系构型样式研究:以西非尼日尔三角洲盆地某深水研究区为例[J]. 地质论评,2013,59(3):510-520. doi: 10.3969/j.issn.0371-5736.2013.03.011
[6] 赵晓明,吴胜和,刘丽. 尼日尔三角洲盆地Akpo油田新近系深水浊积水道储层构型表征[J]. 石油学报,2012,33(6):1049-1058. doi: 10.7623/syxb201206018
[7] 谢清惠,邓宏文,郭佳. 西非下刚果盆地深水曲流水道的地震响应特征与演化模式分析[J]. 石油物探,2013(6):655-661.
[8] 于水,程涛,陈莹. 尼日尔三角洲盆地深水沉积体系特征[J]. 地球科学:中国地质大学学报,2012,37(4):763-770.
[9] 吕明,王颖,陈莹. 尼日利亚深水区海底扇沉积模式成因探讨及勘探意义[J]. 中国海上油气,2008,20(4):275-282. doi: 10.3969/j.issn.1673-1506.2008.04.016
[10] NYANTAKYI E K,LI T,HU W S,et al. Structural and stratigraphic characteristics on distal parts of the outer fold and thrust belt of southern Niger Delta,Nigeria[J]. Arabian Journal of Geosciences,2015,8(9):6677-6695. doi: 10.1007/s12517-014-1727-x
[11] 冯鑫,韩文明,范洪耀. 西非深水目标高密度三维地震勘探效果[J]. 石油物探,2020,59(1):150-157. doi: 10.3969/j.issn.1000-1441.2020.01.017
[12] 关达,张秀容. 利用地震属性预测河道砂体[J]. 石油物探,2004(S1):56-58.
[13] 丁峰,胡光义,尹成,等. 基于地质信息约束的概率神经网络地震反射模式预测方法[J]. 中国海上油气,2018,30(1):127-131.
[14] 冯潇飞,赵晓明,谭程鹏,等. 深海弯曲水道内部一种特殊的沉积单元:凹岸坝[J]. 沉积学报,2020,38(2):440-450.
[15] 穆星. 利用地震几何属性和自组织神经网络进行地震相的自动识别[J]. 地质科技情报,2005(3):109-112. doi: 10.3969/j.issn.1000-7849.2005.03.022
[16] 吕慧, 鲍典, 王震, 等. 基于地震多属性的岩溶古河道识别及刻画[C]//SPG/SEG南京2020年国际地球物理会议论文集. 北京: 石油地球物理勘探编辑部, 2020: 816-819.
[17] 丁峰,尹成,朱振宇,等. 利用改进的自组织网络进行地震属性分析[J]. 西南石油大学学报(自然科学版),2009,31(4):47-51,200.
[18] 杜红斌,郭巧占. 概率神经网络及其在储层产能预测中的应用[J]. 石油仪器,2005(4):54-56,91.
[19] 罗浩然,尹成,丁峰,等. 概率神经网络的平滑参数分析及在地震属性分析中的应用[J]. 石油物探,2017,56(4):551-558. doi: 10.3969/j.issn.1000-1441.2017.04.011
-