2D elastic FD modelling of multi-component vibroseis data acquired at a salt pillar

  • Venue:

    Geb. 06.42 - Room 001 (Seminar Room) / Online

  • Date:

    24.05.2022

  • Speaker:

    Abdullah Jarah

    GPI - KIT

  • Time:

    9:30 am

  • My name is Abdullah Jarah. I got the bachelor's degree in Geology (specialization: Geophysics)  in 2014 from Damascus University in Syria. I had excessive training in the Syrian Petroleum Company in the field of Geophysical and Geological interpretation in Damascus. I moved to Germany in 2016 and started my master's at KIT in WS 2018/2019. Languages:  English, German and Arabic.

Abstract

The progress in salt and potash mining has been increased rapidly in the past decades. However, the occurrence of gas and saline solutions on-site poses an economic and safety risk for subsurface mining activities. In this context the German Research Centre for Geosciences in Potsdam (GFZ) with other contributors carried out high-resolution explorations of cavernous structures within the salt formations. I performed 2D elastic finite-difference modelling of a 20 m × 20 m big salt pillar. The multi-component receivers are installed around the salt pillar. A high-frequency magnetostrictive vibrator source is used to excite a sweep signal from 100 Hz to 12 000 Hz; the shot positions surround the pillar. The preprocessing of the observed data prior to the FD modelling consists of receiver rotation, spreading transformation and time-windowing of the first arrival P-wave. In order to fully satisfy the free surface boundary condition in FD modelling, the vacuum formalism (VF) approach is utilized where the boundary conditions are treated implicitly. Zero-phase Klauder wavelet extracted from the observed data is used as a source wavelet to forward simulate the respective wavefields based on realistic P-wave velocities obtained through a traveltime tomography by GFZ. From this traveltime tomography model, the S-wave velocity model is extracted by employing the typical vp/vs ratio in the salt, whereas the density model is obtained by interpolating density values from geochemical boreholes in the salt pillar. A synthetic modelling shows a good agreement of the first arrivals of modelled and observed data, especially for body waves. Surface waves present modeling difficulties and their exact behavior could not be explained conclusively. Therefore, time windowing is implemented to only display the direct P- and S-waves and cut off all other phases. An improved data fit can be obtained with source-time-function inversion. Future studies should investigate the results to perform the entire full-waveform inversion (FWI) and take advantage of the other observed data which are obtained after inserting various fluids into the salt pillar to see the influence of these fluids on the salt deposits.