摘要
我国南方页岩气地层厚度小于地震波长,陡倾角地层发育,具有较强的TTI各向异性特征,利用各向同性方法处理地震资料的结果信噪比和分辨率较低,且存在严重的井震深度误差。同时,TTI建模方法技术尚不成熟,成像方法实用化程度不高。针对上述问题,从准确的各向异性程函方程出发,推导了非双曲时距曲线方程,结合局部层析方法,充分利用测井数据,建立空间分布合理、数值较为准确的各向异性参数初始模型,采用加入综合正则化的等效参数联合层析反演方法进一步提高参数模型的精度,增加模型的中高波数信息。采用梯度法各向异性波场外推求解算法和双平台异构并行策略,实现CPU/GPU混合的TTI逆时偏移成像(TTI-RTM),有效降低了TTI-RTM成像处理的数据存储量和运算量,提高了TTI-RTM成像结果的有效性和实用性。将该方法应用于南方某页岩气探区实际地震资料处理,建立了符合地质规律、细节丰富的精细TTI各向异性参数模型,改善了TTI-RTM成像剖面的质量和精度,降低了井震深度误差,为后续储层精细描述和水平井开发提供了可靠的成果数据。
In the shale gas exploration area in south China,the stratum thickness is lower than the seismic wavelength,steep dip formations are present,and reservoirs exhibit strong TTI characteristics.Isotropic processing was inaccurate,producing low signal-to-noise ratio,poor resolution,and large depth errors compared to well data.TTI parameter modeling methods are not well developed,and TTI anisotropic imaging methods are not practical enough.To address these issues,we derived a non-hyperbolic time-distance curve equation based on the accurate anisotropic Eikonal equation.An initial anisotropic-parameter model with reasonable spatial distribution and accuracy was established by combining the local tomography method and well data.The accuracy of the parametric model was improved with equivalent parameter joint tomography inversion based on comprehensive regularization.The medium and high wavenumber model data was also improved.Using a gradient anisotropic wave field extrapolation algorithm and dual platform heterogeneous integration,we accomplished CPU/GPU hybrid TTI reverse time migration(TTI-RTM)imaging,effectively reducing storage and improving computing efficiency.Field data from a shale gas exploration area in south China was used to establish a fine TTI anisotropic parameter model based on the relevant geology.Then TTI-RTM improved the quality and accuracy of the imaging profile.Well-seismic depth error was eliminated,allowing reliable results to be obtained for subsequent reservoir description and horizontal well exploration and development.
作者
杨勤勇
郭恺
李博
刘小民
YANG Qinyong;GUO Kai;LI Bo;LIU Xiaomin(Sinopec Geophysical Research Institute,Nanjing 211103,China)
出处
《石油物探》
EI
CSCD
北大核心
2019年第6期882-889,897,共9页
Geophysical Prospecting For Petroleum
基金
国家高技术研究发展计划(863计划)课题(2011AA060303)资助~~
关键词
页岩气
TTI各向异性
参数建模
逆时偏移
井震深度误差
局部层析
等效参数
shale gas
TTI anisotropy
parameter modeling
reverse time migration
well-seismic depth error
local tomography
equivalent parameter