地震岩石物理模板应用中的两个关键问题

Two key points of rock-physics template and its application

本文首先总结了地震岩石物理模板技术的应用现状,并指出目前业界应用中存在两个误区,一是基于岩石物理模板指示的多边形岩相分类,二是根据弹性参数相对变化量优选岩性/流体敏弹性因子,这两个关键环节会的问题在于多边形岩相分类不能考虑测量误差、随机噪声等的影响,另一个限制是不允许岩相重叠;仅根据弹性参数相对变化量优选岩性/流体敏弹性因子没有考虑各岩相的分布及标准偏差的差异性,得到的优选结果可能不客观.这两个方面都会影响定量地震解释精度,甚至可能导致得到错误的结果.针对这两个问题,文中给出了对应的解决方案,基于关键井及岩石物理模板的概率密度函数(PDF)执行贝叶斯分类能够较好避免岩相分类中的问题;根据弹性参数概率分布优选岩性/流体敏弹性因子综合考虑不同岩相均值及方差的差异,是相对更为客观的评价方法,并给出了具体公式,相比于传统做法,本文所作的改进能够提高定量地震解释结果的精度,避免落入解释陷阱.

Rock-physics template（RPT）was used broadly in oil and gas exploration and development recent years,and was considered as the key technology during quantitative seismic interpretation,but there are two important point we need to take care：one question is how to do facies classification,the other question is how to choose the most sensitive petro-elastic parameters group when making RPT.Currently,most of the geoscientist choose the most sensitive petro-elastic parameters based on the relative changes of each facies mean value,and then,using the polygon technology to do facies classification.The main problem in it is that,for facies classification,this does not take the random noise and measure error into consideration,and the result maybe not objective,even got the absolutely wrong outputs.In order to improve the accuracy of quantitative seismic interpretation,this paper advised to take both means value and standard deviation of each facies into consideration when choosing most sensitive petro-elastic parameters group,the criteria is the percentage of overlap area for each facies probabilitydistribution,and using Bayesian theory to do facies classification based on each facies probability density function and the instruction of RPT.Compared to conventional method, those two improvements for RPT application can tremendously improve the accuracy of quantitative seismic interpretation,has the benefit of avoid falling into the interpretation pitfalls.