摘要
零偏VSP以水平地层为假设条件,当地层倾斜时,零偏VSP时深关系精度和走廊叠加记录精度出现误差,更严重的是,走廊叠加记录上观测井段之下地层的上行反射波成像位置错误,需进行倾角时差校正处理。针对复杂高陡构造零偏VSP资料处理,提出了空变倾角时差校正的基本概念和方法,利用零偏VSP上、下行波场计算反射界面倾角,建立二维层速度初始模型,通过射线追踪修正速度模型,使倾角空间变化的高陡构造地层的VSP上行反射波场得到充分拉平,不仅提高了零偏VSP走廊叠加记录的精度,而且观测井段之下地层的上行反射也得到有效拉平并能精确成像,从而实现观测井段之下地层深度的精确预测和波组特征预测,时深关系也校正到地震波自地面震源到井下检波器的铅直传播时间,有效解决了复杂高陡构造的零偏VSP时深关系、层位标定和钻前地层预测等问题,实际资料处理效果验证了方法的有效性。
Zero-offset VSP is based on the assumption of horizontal strata.For inclined layer,the time-depth relationship and seis mic corridor stack of zero-offset VSP is not accurate, and more seriously, the imaging position on corridor stack record under observation well section is incorrect, so we need to conduct DMO correction.For complex high-steep structure, we propose a modified method of space varying DMO correction for zero-offset VSP data.Firstly, the down-going and up-going waves are used to calculate dip angle of reflection interface, an initial 2D interval velocity model is established. Then, by tracing down-going and up-going wave's ray path with ray tracing algorithm, the above velocity model is amended, which can make the up-going waves of zero-offset VSP from complex high-steep structure come to flat.It not only enhances the accuracy of corridor stack record, but makes up going wave correctly imaging under observation well section. Therefore, the accurate prediction of formation depth and seismic reflections characteristics are achieved.Time-depth relationship is also corrected to direct path wave travel time in vertical depth, which effectively resolves problems such as time-depth relationship, seismic horizon calibration and pre-drilling prediction in complex highsteep structure.The actual data processing shows the effectiveness of the method.
出处
《石油物探》
EI
CSCD
北大核心
2017年第3期408-415,共8页
Geophysical Prospecting For Petroleum
基金
国家科技重大专项"塔里木前陆盆地油气富集规律
勘探技术与区带和目标优选"(2011ZX05003-04)资助~~
关键词
时深关系
层位标定
走廊叠加
倾角时差校正
VSP资料处理
钻前预测
time-depth relationship, seismic horizon calibration, corridor stack, DMO correction, VSP data processing, pre-drilling prediction