Research On Matching Methods Of Logging And Seismic Based On Cross Frequence Band Rock Physics Model

In the process of inversion and interpretation,seismic data usually needs to be combined with logging data to compensate for the shortcomings of insufficient longitudinal resolution.Because there are differences in the scale,frequency and propagation path between logging data and seismic data,and the actual rock pore structure often leads to dispersion,which will make them differ in speed response.Therefore,the log data needs to be corrected for dispersion to better match the seismic data.In this paper,we need preprocess the well logging data at the beginning.First,the abnormal data is eliminated by median filtering.Then,the Walsh transform is used to perform preliminary layering on the logging data.Finally,the logging data is more finely layered according to the principle of minimum speed difference and minimum thickness,so that it retains more formation information.Secondly,three scale upscaling methods are adopted,which are simple weighted average method,Bacchus equivalent averaging method and correlation function(PCF)method,to reduce the sonic logging speed.So that the well logging data can be upscaled from logging frequency band to seismic frequency band,and compare with the same frequency band seismic VSP speed.Then,using three cross-band rock physics models,which are the standard viscoelastic model,the resonant Q model and extension of Gurevich's squirt-flow model(E-S)model based on the pore structure distribution,to correct the dispersion of logging data.Finally,the synthetic seismic records are made by using the frequency-reduced well data and the dispersion-corrected well data,and compared with the synthetic seismic records produced by the original logging data,and inserted into the well-side seismic traces to analyze the matching effect.By analyzing the final synthetic seismic record,it can be seen that the direct use of the logging data after the scale upscaling to make the synthetic record does not significant improve the time-depth relationship,and produces a large difference in the amplitude characteristics.In comparison,the PCF method considers the interaction between heterogeneous bodies in an inhomogeneous medium and the elastic scattering of waves propagating in the formation.The upscaling effect is slightly better than other two methods.Therefore,it can not to use the results of the scale upscaling to perform the well-seismic matching.And it is also necessary to use the scale upscaling methods in combination with the corresponding dispersion model.The synthetic seismic records made by the well data corrected by the three cross frequency band rock physics models have improved the effect.The layers can be well matched,the phase deviation is low,the amplitude characteristics are similar,and the correlation with the well-side seismic traces is also improved.In particular,the newly established E-S model realizes point-by-point dispersion correction of logging data,which makes it more advantageous in thin layer information display and fine layer calibration.