油藏描述中的井震时深转换技术研究

Research on time-depth conversion by well-to-seismic in reservoir characterization

油藏描述是油藏地球物理研究的重要内容之一,油藏描述的精度取决于时间域地震相解释和深度域测井相解释的精度以及两者综合解释的精度。显然,要实现时间域的地震空间信息与深度域的测井垂向信息的综合油藏描述(静态储层建模),就必然会面临地震与测井信息间的时深转换问题。一般情况下,时深转换的精度基本能够满足构造成图的要求,但是油藏描述中要求的时深转换精度远远高于构造成图的精度。因此,油藏描述中的时深转换将面临更多的问题,例如井震分辨率差异、井震时深转换速度的精度、井震时深转换方法、井震时深转换层位的选择、井震间基准面的选择等。以上问题直接影响井震时深转换的精度和油藏描述的能力。为此,提出了基于标准参考层位的井震时深转换方法、沿标准参考层的地震速度求取方法、井震基准面间校正方法、基于标准参考层的井震精细解释方法以及井震时深转换剩余因子校正方法。通过实例研究和严格的质量控制给出了误差小于1 m的井震时深转换结果。

Reservoir characterization is one of the important steps in reservoir geophysics research.However,the accuracy of reservoir characterization not only depends on the accuracy of time-domain seismic and depth-domain logging interpretation,but also depends on the precision of time-depth conversion.Absolutely,if we want to realize the comprehensive reservoir characterization of seismic spatial information in time domain and logging vertical information in depth domain(static reservoir modeling),we are confronted with the time-depth conversion between seismic and logging information. The accuracy of time-depth conversion basically satisfies the requirement of structural mapping.However,in reservoir characterization, the required accuracy of time-depth conversion is much higher than that of structural mapping.Therefore,it should be confronted with more problems when time-to-depth conversion is used in reservoir characterization,such as the difference of resolution from seismic and logging,accuracy of velocity for time-depth conversion, method of time-depth conversion,choice of unified key-horizon and datum between seismic and logging,etc.all the problems will directly impact precision of time-depth conversion and the ability of reservoir characterization.Therefore,we proposed a series of methods based on standard reference horizon,including time-depth conversion method,seismic velocity acquiring method,correction method,well-to-seismic fine interpretation method and residual factor correction method.By case study and strict QC we obtained time-depth conversion result with the error smaller than 1m.