基于三维空间响应分布函数的脉冲中子孔隙度测井快速模拟

Fast simulation of pulsed neutron porosity logging based on three-dimensional spatial response distribution function

应用蒙特卡洛方法进行中子测井正演模拟需要大量时间,无法满足实时反演等实际应用的生产要求。为此,通过蒙特卡洛方法得到不同位置单位地层对探测器计数的贡献即地层的三维空间响应分布函数,建立不同地层条件下的三维空间响应分布函数和热中子计数的数据库;再基于三维空间响应分布函数对不同区域地层进行积分得到不同区域地层对探测器计数的贡献;最后利用不同区域地层对探测器计数的贡献结合不同地层热中子计数实现了脉冲中子孔隙度测井曲线的快速模拟,并对比了不同地层条件下快速模拟方法与蒙特卡洛方法的计算结果。研究结果表明:(1)与蒙特卡洛方法相比,在直井和斜井中,快速模拟方法计算结果的最大误差分别为2.5 PU和3.0 PU;(2)对于不同岩性和含不同孔隙流体的地层,快速模拟方法与蒙特卡洛方法的计算结果基本一致;(3)快速模拟方法与蒙特卡洛方法相比将计算花费的时间从原本的几十小时降低到十几秒。结论认为,快速模拟方法不仅大幅度提高了正演速度,并且计算精度与传统的蒙特卡洛方法相近,为后续反演等应用打下了可靠的基础。

It takes much time to carry out forward simulation of neutron logging by using the Monte Carlo method,which cannot meet the requirements of real-time inversion and other actual applications.For this reason,this paper applies the Monte Carlo method to obtain the contribution of unit formation at different locations to detector counting,i.e.,the three-dimensional spatial response distribution function(3D-SRDF),and establishes the 3D-SRDF under different formation conditions and the database of thermal neutron counting.Then,the formations in different regions are integrated based on the 3D-SRDF,so as to obtain the contribution of formations in different regions to detector counting.Finally,fast simulation of pulsed neutron porosity logging curve is realized by using the contribution of formations in different regions to detector counting,combined with the thermal neutron counting of each formation.What’s more,the calculation results under different formation conditions by the fast simulation method are compared with those by the Monte Carlo method.And the following research results are obtained.First,compared with those by the Monte Carlo method,the maximum errors of calculation results by the fast simulation method are 2.5 PU and 3.0 PU for vertical wells and inclined wells,respectively.Second,for the formations with different lithologies and different pore fluids,the calculation results by the fast simulation method are basically consistent with those by the Monte Carlo method.Third,compared with the Monte Carlo method,the fast simulation method reduces the calculation time from tens of hours to only more than ten seconds.In conclusion,the fast simulation method improves the forward modeling speed greatly and its calculation accuracy is close to that of the traditional Monte Carlo method,which lays a reliable foundation for subsequent inversion and other applications.