A Method For Processing Carbon Oxygen Ratio Spectral Data In Pulsed Neutron Full Spectrum Logging

At present,many onshore oil fields in China have entered a high water cut stage after a long period of water injection development,and the difficulty of extraction is becoming increasingly difficult.Therefore,it is particularly important to dynamically monitor residual oil saturation and improve oil recovery.Carbon oxygen ratio spectroscopy logging is an effective method for analyzing residual oil saturation in cased wells.This article studies the raw energy spectrum data obtained from the carbon oxygen ratio logging mode of the domestically produced PNST logging tool and proposes a set of carbon oxygen ratio spectroscopy data processing methods.This article first decompresses the raw spectral data recorded by the PNST logging tool and extracts various energy spectra required for this study;Secondly,preprocess the extracted energy spectrum data,including smoothing filtering,spectrum drift correction,and net spectrum generation;Then establish a suitable stratigraphic model,use MCNP5 software to perform Monte Carlo numerical simulation to obtain two sets of standard spectra of elements,namely capture and non elastic,and calculate the broadening coefficient of the energy spectrum by fitting the measured spectra of the calibration well.Gaussian broadening is performed on the simulated element standard spectra;Then,research is conducted on using the non negative least squares method to solve the spectrum and calculate the element yield.The mixed spectrum is obtained through forward modeling,and Gaussian noise is added to the mixed spectrum to invert the element yield.The error of this spectral solution method is calculated;Finally,three curves of element window counting rate,element yield,and element net counting rate,as well as element ratio curves,were obtained.The application effect of the processing method studied in this article was verified through actual well data processing.The research results show that the influence of statistical fluctuations is effectively reduced through five point three smoothing filtering,and the fluctuation changes of the energy spectrum are well preserved;The combination of channel by channel comparison method and function fitting method is used to determine the hydrogen peak.The iron peak is determined by calculating the correlation coefficient with the standard spectrum of iron element capture.After calculating the peak drift parameters,the effect of spectrum drift correction is good;By determining the net spectral coefficient through the hydrogen peak area ratio,the non elastic net spectrum obtained is more accurate and reliable;Simulate the detector response energy spectrum and detector surface flow spectrum separately,and obtain the element standard spectrum after convolution calculation.In the simulation,multiple methods such as gate element importance,weight window,and photon generation bias card are used to reduce simulation errors,effectively improving the quality of the simulation spectrum,and obtaining a good standard spectrum effect.The element yield was obtained by solving the spectrum using the non negative least squares method.After numerical simulation verification,it was found that the yield calculated by this method has high accuracy and small average error.Finally,through the data processing method studied in this article,three curves of element window counting rate,element yield,and element net counting rate,as well as element ratio curves,were obtained for actual well data processing.The practical application effect is good.This article studies a set of data processing methods based on PNST logging carbon oxygen ratio spectrum logging data,including data preprocessing,element standard spectrum simulation,and spectral analysis to determine element yield.This has certain significance for the improvement and development of carbon oxygen ratio logging data processing methods.