Methods and systems for accurate well-seismic tie based on stepwise matching and optimization

Patent metadata

Field	Value
Publication number	US-12535609-B2
Application number	US-202318338281-A
Country	US
Kind code	B2
Filing date	Jun 20, 2023
Priority date	Jun 12, 2023
Publication date	Jan 27, 2026
Grant date	Jan 27, 2026

Abstract

The present disclosure provides a method for an accurate well-seismic tie based on stepwise matching and optimization, including: determining target seismic data as a seismic trace near well and performing first data processing on the target seismic data; performing second data processing on the logging information to determine a reflection coefficient sequence; extracting seismic wavelets, convoluting the seismic wavelets with the reflection coefficient sequence, determining a synthetic seismic record of the target calibration object in a time domain, performing third data processing on the synthetic seismic record; determining a correlation between the seismic trace near well and the synthetic seismic record, determining a global time shift between the synthetic seismic record and the seismic trace near well, performing a preliminary alignment on the synthetic seismic record; and calculating a local time shift between the seismic trace near well and the synthetic seismic record and accurately calibrating the synthetic seismic record.

First claim

What is claimed is:

1 . A method for an accurate well-seismic tie based on stepwise matching and optimization, wherein the method is implemented by a processor and comprises:
based on seismic data and logging information, determining target seismic data as a seismic trace near well, and performing first data processing on the target seismic data; performing, based on a target calibration object, second data processing on the logging information to determine a reflection coefficient sequence; extracting seismic wavelets of the seismic trace near well, convoluting the seismic wavelets with the reflection coefficient sequence, determining a synthetic seismic record of the target calibration object in a time domain, and performing third data processing on the synthetic seismic record; determining, through a first preset algorithm, a correlation between the seismic trace near well and the synthetic seismic record, determining a global time shift between the synthetic seismic record and the seismic trace near well, and performing a preliminary alignment on the synthetic seismic record through the global time shift; and calculating a local time shift between the seismic trace near well and the synthetic seismic record through a second preset algorithm and accurately calibrating the synthetic seismic record through the local time shift.

2 . The method according to claim 1 , wherein the seismic data includes seismic data in the time domain, a seismic interpretation horizon, and the logging information includes at least one of sonic time difference logging information, density logging information, resistivity logging information, well coordinates, and a geological stratification.

3 . The method according to claim 2 , wherein the determining target seismic data as a seismic trace near well based on seismic data and logging information, and performing first data processing on the target seismic data includes:
calculating, based on the well coordinates and nearby seismic trace coordinates, a Euclidean distance between the well coordinates and the nearby seismic trace coordinates; and determining the seismic trace near well based on the Euclidean distance and performing the first data processing on the seismic trace near well.

4 . The method according to claim 1 , wherein the performing, based on a target calibration object, second data processing on the logging information to determine a reflection coefficient sequence includes:
preprocessing sonic time difference logging information, replacing an anomalous segment of the sonic time difference logging information, and obtaining preprocessed sonic time difference logging information; reconstructing density logging information, replacing an anomalous value of the density logging information, and obtaining reconstructed density logging information; smoothing the preprocessed sonic time difference logging information to obtain smoothed sonic time difference logging information; determining a longitudinal wave velocity sequence based on the smoothed sonic time difference logging information; and determining a reflection coefficient sequence in a depth domain based on the longitudinal wave velocity sequence and the reconstructed density logging information.

5 . The method according to claim 1 , wherein the extracting seismic wavelets of the seismic trace near well, convoluting the seismic wavelets with the reflection coefficient sequence, determining a synthetic seismic record of the target calibration object in a time domain, and performing third data processing on the synthetic seismic record includes:
converting a reflection coefficient sequence in a depth domain to a reflection coefficient sequence in the time domain based on sonic time difference logging information; determining, through a convolution model and error energy, a seismic wavelet with minimum error energy based on the reflection coefficient sequence in the time domain and a seismic trace near well record; and performing a convolution calculation based on the seismic wavelet and the reflection coefficient sequence in the time domain to determine the synthetic seismic record, and performing the third data processing on the synthetic seismic record.

6 . The method according to claim 1 , wherein the determining, through a first preset algorithm, a correlation between the seismic trace near well and the synthetic seismic record, determining a global time shift between the synthetic seismic record and the seismic trace near well, and performing a preliminary alignment on the synthetic seismic record through the global time shift includes:
determining a sampling plan for the seismic trace near well and the synthetic seismic record and determining a time window for each sampling point in the sampling plan; truncating the synthetic seismic record and the seismic trace near well based on the time window to generate a truncated synthetic seismic record and a truncated seismic trace near well, respectively; converting a geological stratification in a depth domain into a geological stratification in the time domain and calculating an amount of time deviation between the geological stratification in the time domain and a seismic horizon; determining at least one correlation between the truncated synthetic seismic record and the truncated seismic trace near well under at least one time shift, wherein a range of values for the at least one time shift is determined based on the amount of time deviation; obtaining a time shift corresponding to a maximum correlation in the at least one correlation as the global time shift based on the correlation; and performing the preliminary alignment on the synthetic seismic record through the global time shift and determining a preliminarily aligned synthetic seismic record.

7 . The method according to claim 6 , wherein the determining a sampling plan for the seismic trace near well and the synthetic seismic record includes:
generating at least one candidate sampling plan, the at least one candidate sampling plan including a sampling feature for sampling each of at least one time domain subinterval; determining, based on each of the at least one candidate sampling plan, a seismic trace near well, a seismic wavelet, and a reflection coefficient sequence corresponding to the candidate sampling plan; determining the global time shift based on the seismic trace near well, the seismic wavelet, and the reflection coefficient sequence corresponding to the candidate sampling plan; filtering the at least one candidate sampling plan, obtaining a first sampling plan set, and generating a first histogram distribution of global time shifts based on the global time shifts corresponding to candidate sampling plans in the first sampling plan set; and determining, based on the first sampling plan set and the first histogram distribution, a candidate sampling plan whose total downsampling rate exceeds a preset threshold and global time shift is the same as a global time shift with a highest frequency in the first histogram distribution in the first sampling plan set as the sampling plan.

8 . The method according to claim 6 , wherein determining the time window includes:
determining, through a time window determination model, a size of the time window based on the seismic trace near well and the synthetic seismic record, the time window determination model being a machine learning model.

9 . The method according to claim 8 , wherein the time window determination model includes a first embedding layer, a second embedding layer, and a determination layer, and the determining, through a time window determination model, a size of the time window includes:
determining, based on the seismic trace

Assignees

• Univ Southwest Petroleum

Inventors

• Cao Weiping
• Li Mengcheng
• Huang Xuri
• Tang Shuhang
• Yang Ran
• Li Luo
• Li Moyan
• Yang Ranran
• Li Siyao
• Li Xinwang

Classifications

• G01V2210/614

  Synthetically generated data · CPC title

• G01V11/00 Primary

  Prospecting or detecting by methods combining techniques covered by two or more of main groups G01V1/00 - G01V9/00 · CPC title

• G01V13/00

  Manufacturing, calibrating, cleaning, or repairing instruments or devices covered by groups G01V1/00 – G01V11/00 · CPC title

• G01V1/282 Primary

  Application of seismic models, synthetic seismograms · CPC title

• G01V1/366 Primary

  by correlation of seismic signals · CPC title