3D Seismic Sedimentology of Nearshore Subaqueous Fans – A Case Study from Dongying Depression, Eastern China - Advances in Data, Methods, Models and Their Applications in Geoscience - page 129

3. The pseudo-acoustic 3D seismic inversion based on Logs reconstruction
3.1 Methodology
The pseudo-acoustic 3D seismic inversion method different from the conventional 3D seismic
impedance inversion method not only in working through logs reconstruction, inversion,
interpolation and extrapolation, but also adding or replacing characteristic curves to the
density logs or, more commonly, velocity logs in order to achieve the ability to identify the
reservoir from the surrounding rock in the case of small impedance difference (Shen & Yang,
2006; Zhang et al., 2005). The potential reservoir may show no direct relationship with the
seismic reflection but can be distinguished from different lithological changes.
The velocity and the time-depth relationship after logs reconstruction may change so
deviations between seismic reflection horizon and synthetic seismogram calibration's
horizon should be established to reflect these changes (Luo et al., 2006). The pseudo-acoustic
seismic inversion results based on logs reconstruction may be not accurately reveals the
corresponding lithological changes of the target layers. To solve this problem, the zero
Mean-Based logs reconstruction techniques, which keeps the original time-depth
relationship unchanged, can be applied. The principle is to set the characteristic curves or
logs involved in seismic inversion to a mean of 0, that is ΣAi = 0 (Ai is characteristic curve
sample values of target layers). Adding or subtracting the normalized curve and acoustic
characteristics curve, then properly magnifying the normalized characteristics curve in
order to highlight lithology information. This process can be expressed as:
pseudo-acoustic curve = acoustic logs ± characteristic curves × K,
while K stands for the curve amplification factor.
As the characteristic curves keep the information of the target layer, and the velocity curves
of the upper and lower target layers are kept unchanged, so the original time-depth
relationship will remaine unchange (Luo et al., 2006).
The implementation process of this method includes the following: ① selection of the
characteristic curves; ② standardization of the characteristic curves; ③ normalization and
reconstruction of the pseudo-acoustic curve; ④ seismic wavelet extraction and the
establishment of the initial model; ⑤ pseudo-acoustic seismic inversion.
3.2 The pseudo-acoustic 3D seismic inversion based on logs reconstruction
1. Selection of the characteristic curves
To select the characteristic curves of the target layers, quantitative and semi-quantitative
correlations through statistical analysis are established between different lithologyies (such
as the conglomerate in the fan-root, sand-conglomerate in the mid-fan, mudstone, gypsum-
salt rock in the marginal-fan), effective reservoir (such as gas sand-conglomerate in the mid-
fan), logs (such as acoustic time (ac), natural gamma (gr), neutron porosity (cnl),
spontaneous potential (sp), and logging parameters that correspond to different lithology
types in different fans.
The results show that single logs parameter cannot identify the different lithologies in
different fans, but combinations of any two of logging parameters (ac, gr or sp) can
effectively indentify them to some extent. Further analysis also show that any two logs
parameter's combinations between ac, gr, and sp could distinguish the effective and
ineffective sand -conglomerate reservoir with a thicknesses greater than 6 m (Fig.4).
Therefore, we can use any two logs combination between ac, gr, and sp as the characteristic
curves.