超深层目标随钻测井引导的动态地震成像

Dynamic seismic imaging driven by logging while drilling in ultra-deep targets

超深层油气储层目标上覆巨厚的复杂介质,基于地面观测数据估计的地震波传播速度存在误差,将导致地震成像结果不准确,增加深井钻探风险或勘探成本.随钻测井提供的地层速度信息可用于提高偏移速度模型的精度,从而通过叠前深度偏移重处理减小井震闭合误差,及时地重新定位深层的钻探目标,并为钻井轨迹调整提供科学依据.为此,本文提出一种随钻测井引导的动态地震成像方法.它在相对地质年代信息约束下,将测井资料处理获得的层速度用于修正井孔周围一定范围内的偏移速度模型,结合波动方程基准面延拓实现局部三维叠前深度偏移处理,利用图形处理单元(GPU)加速以便更新钻头前方的地震图像.基于实际地震与测井资料的仿真试验表明该方法技术具有可行性,预期可为面向深储层的智能导向钻井提供有力支撑.

The ultra-deep oil and gas targets are usually overlaid by extremely thick and complex media.The velocities of seismic waves estimated from the observed data at the surface easily deviate from the true values in the subsurface,which will lead to inaccurate seismic imaging and thus significantly increase the risk and cost of deep well drilling.Improving migration velocity model with the interval velocity obtained from well logging while drilling can drive prestack depth migration of the next round and reduce well-seismic mis-tie.Accordingly,we can relocate deep target and timely adjust the rest drilling trajectory.To meet the above requirements,we introduce a dynamic seismic imaging method driven by the logging while drilling technology.First,contrained by the relative geological time(RGT)volume,the migration velocity model around the well is updated according to the difference between Backus average of logging velocity and migration velocity at the well location.Then,the seismic imaging volume ahead the drill-bit is updated before the next drilling stage by combining wave-equation-based redatuming and local 3D prestack depth migration,which are leveraged by the massively parallel architecture of graphic processing units(GPUs)to accelerate the computation.The simulation test with the field seismic data and well-log velocity demonstrates the validity of the proposed approach,and reveals the potential to support intelligent drilling of the ultra-deep targets.