Improved Synthetic Seismic Trace Generation Using Check Shot Survey Data

Abstract

Abstract A review of sonic well log editing using check shot data shows how computer generated synthetics can be made which match the seismic data when the sonic log is otherwise unusable. Comparisons of three sets of data including sonic logs, seismic data, and check shot survey data for two wells from the Louisiana Gulf Coast show up major sources of common errors in both check shot and sonic surveys. A careful examination of the time-depth relationships implied by the integrated sonic log and the check shot data through the creation of a "mismatch" or discrepancy curve defines zones within which corrections can be made to the sonic log in order to compensate for mis-triggering of the tool or even normal errors in fractured or damaged rocks. Also, a good tie between the synthetic traces and the seismic data provides a cross check on the check shot survey. Corrections made to the sonic logs result in geologically consistent ties with the seismic data attaching a new level of confidence to the geophysical interpretation. Introduction The sonic log, now almost routinely used in exploration wells is growing in its usefulness to both exploration and production personnel. As seismic data becomes more reliable and higher resolution data is utilized more universally for stratigraphic interpretation, sonic logs perforce become the key in relating geology to the seismic record section. The creation of reliable synthetic seismograms for reflection analysis and correlation is a problem of careful editing and calibration of the sonic tool. Editing procedures are well known, and today's modern borehole compensated tools tend to minimize the number of obvious recording errors. Even so, before the synthetic seismogram is created, the log should be calibrated to match some sort of time-depth control or else small systematic errors in the log will lead to erroneous ties between the litho logy from the well and the reflection data. It is the purpose of this review paper to examine the usefulness of sonic log calibration methods using check shot survey data. Two wells drilled in Louisiana's Gulf Coast Area show discrepancies between the sonic-Log-derived time-depth relationships and their respective check shot surveys. It is especially important in this and other plays to know the depths not only to the objective, but also, to several geologic zones above the objective which dictate the planning of the drilling program. The obligation rests with the geophysicist to try to infer these depths from the seismic data. The first well, which I shall call "Well A" has a sonic log which exhibits a very large error due to undetected tool problems encountered during recording of the log. The second well, "Well B" exhibits some subtle discrepancies attributable to errors in the check shot survey interpretation. In both cases, a detailed examination of the discrepancy between the two types of data for each well has allowed the determination of the source of the errors and thereby, a reason for correcting them.