Due to the existence of a rugged seafloor, strong lateral velocity changes across the seafloor interface cause the underlying strata to undulate with the seafloor, severely distorting the structural form, which cannot reflect the true appearance of the structure at all, seriously affecting the underlying strata. Seismic imaging of layers. In recent years, a variety of methods have been used to address the rugged seafloor, including laboratory forward simulations, testing of acquisition parameters, and timely deep conversion of processing methods. This has further revealed the essential characteristics of seismic wave propagation in deep-water rugged seafloor areas. , and the influence mechanism of rugged seafloor on seismic waves and factors of imaging distortion; through research on seismic processing in rugged seafloor areas, especially the suppression of diffracted multiple waves in rugged seafloor areas, the quality of seismic data has been improved; through layer replacement technology , wave field extension technology, pre-stack depth migration processing and other methods have been tested to determine the processing process of pre-stack depth migration on the rugged seabed, and solve the problem of structural distortion caused by the rugged seabed.
At the same time, under the conditions of long cables and large offsets, some conventional processing techniques can no longer be applied, such as dynamic correction based on hyperbolic reflection travel time, velocity analysis and horizontal superposition, and suppression of multiple waves. method. Recent international research on velocity analysis can be summarized into three aspects: first, the development of pre-stack velocity analysis methods based on non-hyperbolic reflection travel time equations; second, improved calculation methods of layer velocity; third, the rapid development of migration velocity analysis methods , which is related to the rise of prestack depth migration, mainly tomography methods.
(1) Analysis of existing seismic data
Some seismic data have been collected in deepwater areas over the years. Some data cannot be used due to their age. Therefore, the analysis of existing seismic data is mainly Seismic data collected in 1979 and 1997 were purposefully analyzed. The analysis mainly focuses on noise analysis, main interference wave types, multiple wave development and distribution, etc.
1. Noise analysis
Noise analysis mainly evaluates the distribution frequency band of surge noise and the main frequency band of inherent noise. The analysis methods we use are mainly FK analysis and spectrum analysis. Surge noise is mainly low-frequency noise, and its frequency band is mainly concentrated below 10Hz. The frequency band of inherent noise is mainly concentrated between 30 and 65Hz, and its main noise source is the propeller rotation of the seismic acquisition vessel.
2. Main interference wave types and multiple wave development and distribution analysis
Main interference wave types and multiple wave development and distribution analysis are mainly used to evaluate the interference wave types and multiple wave development. main frequency band. The analysis methods used are mainly FK analysis and spectrum analysis. The main type of interference waves is linear interference. The main frequency band distribution of linear interference waves is concentrated below 20Hz.
Multiple waves mainly manifest as multiple waves of equal length on the seafloor. Their frequency band distribution is very similar to that of primary waves. The main energy is concentrated between 30 and 60Hz. The energy is stronger than the primary effective reflection, covering up An effective primary wave reflection is obtained and reoccurs isochronously. Secondly, multiples also appear as diffracted multiples in rugged seafloor areas. Due to the rugged seafloor, the reflection of seafloor multiples on seismic profiles is also inconsistent. The seafloor is relatively flat. Due to the difference in velocity between multiples and normal formations, traditional methods such as demultiplexing in the Tau-P domain can be used. However, the ruggedness of the seafloor results in an oblique layer with a large angle on the seafloor. The multiple waves generated by this strong oblique layer on the seafloor are difficult to eliminate through conventional methods because their speed is not much different from that of the underlying strata. Elimination makes the lateral energy in the middle and deep layers of the seismic profile very uneven, causing the arcing phenomenon of the offset profile (Figure 5-1).
These multiple waves not only seriously interfere with the effective reflection in the depression, resulting in an extremely low signal-to-noise ratio of the seismic data in the depression, but also cause strong interference to the basement reflection, seriously affecting the seismic data in the area. Geological interpretation and research. Therefore, suppressing and eliminating multiple waves has become the focus of deepwater seismic data acquisition and processing.
Through analysis, the complex seafloor and underground structures are the main factors affecting the quality of data in the area. Deep water seismic data has the following characteristics: the seafloor structure is complex, the water depth changes drastically, and the side reflection and energy reflection in the slope zone are very weak; the noise is dominated by low-frequency interference, mid-to-deep high-frequency interference, and outliers; the frequency band in the shallow layer is wider than that in the mid-to-deep layer , the signal-to-noise ratio and resolution of the mid-deep layers are low; the multiple wave interference is mainly from the deep seafloor and long-period multiple waves, which have strong energy and scattered multiple waves; the velocity pickup of the mid-deep layers caused by the rugged basement is complicated.
Figure 5-1 Strongly diffracted multiple waves in a rugged seafloor area
(2) Processing technology methods
Based on the analysis of original data, the geology of the work area According to the investigation of the situation, combined with the geological tasks and processing requirements, the processing countermeasures adopted are: SRME, high-precision Radon and LIFT technology combined with multiple wave attenuation technology; multi-channel defolding through the series combination of deterministic wavelet processing and coastal bottom structure processing product technology to suppress the continuation phase; for slope zones with very low signal-to-noise ratios, spectrum shaping technology is used to improve the signal-to-noise ratio of data in this area; for places with serious random noise in depressions, multi-domain denoising technology is used to improve the signal-to-noise ratio; High-precision velocity analysis, dense control points in complex structural parts, careful comparison of the front and rear profiles of the target area, and repeated iterations to improve the accuracy and rationality of velocity analysis; use pre-stack depth migration to solve the imaging of rugged seafloor and high-steep structures in the area question.
1. Multiple wave attenuation technology
Attenuation of multiple waves is one of the key points and difficulties in this seismic data processing. Although there are many ways to suppress multiples, none can remove all multiple reflections under all conditions.
In view of the characteristics of multiple waves in the work area, after many tests, SRME (seabed multiple attenuation), high-precision Radon and LIFT multi-domain combined multiple attenuation technologies were adopted. Through three steps The method gradually suppresses multiple waves and achieves very ideal results.
In offshore seismic exploration, short-channel multiples are one of the most difficult coherent noises to deal with. Especially under the influence of shallow gas, short-channel multiples are even more difficult to suppress. The conventional technique for attenuating near-channel multiples is to predict the removal within the deconvolution combination. This technique is simple and effective, but while attenuating the multiples, the effective signal is also removed, destroying the integrity of the gather and causing trouble for subsequent processing. There will be some trouble.
This time we have developed a LIFT technology that effectively attenuates near-channel multiples. This technology simulates effective signals based on the AVO principle and performs signal-noise separation through local time windows. Practice has proven that this technology can not only effectively attenuate short-channel multiples, but also retain effective signals well, laying a solid foundation for subsequent processing.
2. Series combination deconvolution technology
Because the air gun is lowered to a certain depth from the sea surface when collecting marine seismic data, a relatively large amount of seismic energy will be generated immediately after the air gun explodes due to pressure. Large bubbles rise to the sea surface, and coupled with the interference of swells, the signal during the period swings back and forth, so a continuous phase is produced in the signal received by the geophone. The continuous phase generated by this acquisition is strongly reflected in the depths of both shallow and deep water areas, and some even cover effective signals. Therefore, in order to suppress the severe continuation phase, a series combination of deterministic wavelet deconvolution and multi-pass deconvolution was used, and a relatively ideal effect was achieved (Figure 5-2). Compared with statistical wavelet deconvolution, deterministic wavelet deconvolution is more targeted and effectively protects shallow signal and frequency amplitude characteristics.
Figure 5-2 Series deconvolution effect diagram
3. Spectrum shaping technology
For slope zones and low signal-to-noise ratio areas in the base, before iteration Spectrum shaping technology is used to improve the signal-to-noise ratio (Figure 5-3).
Figure 5-3 Comparison of spectrum shaping effects
4. Multi-domain denoising technology
The energy reflection in slopes, depressions, etc. is very weak, causing signal noise The signal-to-noise ratio is very low, and multi-domain denoising technology is adopted to improve the signal-to-noise ratio. The multi-domain denoising method uses the differences between signals and noise in different domains to maximize the difference between interference waves and effective waves. It uses quasi-three-dimensional FXY filtering, linear interference elimination, etc. in the gun domain and maximum offset domain. technology to improve the signal-to-noise ratio of seismic data (Figure 5-4).
Figure 5-4 Multi-domain seismic data signal-to-noise ratio comparison chart
5. High-precision velocity analysis technology
In conventional data processing methods, velocity analysis Coherence measures are commonly used. This method does not consider the effects of noise related to similar or interfering events, residual static corrections, non-hyperbolic time differences, and other non-random noise, thus affecting the time and velocity resolution. This processing uses the newly developed phase correlation statistical method. The advantage of this method is that it has higher and more reliable time and velocity resolution than conventional methods, and is more conducive to the analysis and explanation of small-amplitude structures.
Detection of time resolution: In the synthetic CDP gather, the middle interval of the two sets of events is 30ms. As can be seen from Figure 5-5, the phase-related statistical velocity spectrum is compared with the conventional velocity spectrum. The time resolution is significantly improved.
Detection of velocity resolution: Use two inphase axes at the same time but with different velocities. The velocity difference changes continuously from large to small. Observe the energy clusters in the velocity spectrum until they are inseparable. It can be seen from Figure 5-6 that when the energy groups in the conventional velocity spectrum are unclear, using this method, two energy groups at the same depth can be clearly separated, especially in the deep part, the effect is more obvious.
Figure 5-5 Comparison of resolutions of two velocity spectra
Figure 5-6 Comparison of energy clusters of two velocity spectra
6. Prestack depth migration (PSDM) imaging technology
The core problem of the rugged seafloor is: due to the rugged seafloor, the lateral velocity across the seafloor interface changes strongly, making the seismic ray path complex, and the time interval curve is non-double. Curve, the CMP gather in the conventional processing method is no longer the ultimate reflection point gather, and the superimposed profile is no longer a zero-offset profile, resulting in poor imaging of the underlying strata and serious distortion of the structural form. Chen Li, Ge Yong and others used theoretical models to discuss the effectiveness of using conventional time migration, post-stack depth migration and pre-stack depth migration technologies to solve seismic imaging problems in deep water and rugged seafloor. Through comparative analysis of various migration results of deepwater model data, it is concluded that neither conventional time migration nor post-stack depth migration can solve the problem of seismic imaging in rugged seafloor areas, while pre-stack depth migration is an effective method to solve this problem.
Prestack depth migration technology is usually used to achieve accurate migration imaging of complex structures and solve complex geological problems. For subsurface depth imaging, the most difficult problem is not the migration method but the establishment of a subsurface velocity model. Depth migration is an iterative process, which is a repeated process of continuously building a model, testing the model, running the migration, and correcting the model based on imaging.
Prestack depth migration basically makes no assumptions about the subsurface morphology. The velocity depth model is directly established using prestack data. The subsurface velocity can change both vertically and horizontally. The CMP gathers consider non-hyperbolic effects. The data volume thus obtained can not only improve the signal-to-noise ratio and ensure correct spatial homing, but also directly obtain geologically reasonable depth imaging data volume for geological interpretation. It is obviously a solution to the structural distortion caused by the rugged seafloor. Better way. Figure 5-7 shows the pre-stack time migration and pre-stack depth migration sections of the LW3-1 structure. Comparison shows that the time migration section has a large structural dip angle around the LW3-1 structural area and the imaging accuracy of the underlying stratigraphic structure. It is relatively low, the structure is unclear, and the structural morphology is seriously distorted. However, prestack depth migration has achieved significant improvements in vertical and horizontal resolution, relative maintenance of amplitude, energy focusing on complex structures, and structural morphology, and can meet the geological interpretation requirements. requirements.
Figure 5-7 Comparison of three-dimensional pre-stack time migration profile and pre-stack depth migration profile
After many tests and demonstrations, we chose rugged seafloor development and possible volcanic rocks. The developed Baiyun 6-1 structural area data were used for prestack depth migration tests.
Figure 5-8 (top) is the depth migration result profile of 04EC2458, and Figure 5-8 (bottom) is the final migration time profile of the survey line. From the comparison of depth and time profiles, the depth The profile maintains the original resolution and signal-to-noise ratio, and the profile appearance is relatively natural. The influence of the ruggedness of the seafloor in most areas has been basically eliminated. The events that follow the ruggedness of the seafloor have been basically flattened, reflecting the true structural form of the underground. However, in some areas (Left part of Figure 5-8) There is still the phenomenon of ups and downs of seismic reflections, indicating that the influence of seabed ruggedness has not been eliminated. Analyzing the situation where the influence of local seabed ruggedness has not been eliminated, it can be found that these unideal The seafloor above the situation is some shallower trenches. If you look closely, you can find that these shallower trenches are filled with thicker sediments (Figure 5-9). Through velocity analysis, it is found that the velocity of these sediment layers is very low, about 1670m/s, slightly higher than 1500m/s, but much lower than the formation velocity of 1820m/s on the uplift. Such low-velocity sediments may be some late-deposited silt.
Figure 5-8 Comparison of 04EC2458 pre-stack depth migration profile and final time migration profile
Figure 5-9 Seafloor rugged velocity analysis
Through calculation, If there is a 400ms trench on the seafloor, it may cause the underlying seismic phase to drop by up to 75ms. If there is filling material in the trench for 250ms, it may cause the underlying seismic phase to drop by up to 25ms. It can be seen that the structural shape of the underlying strata is Distortion is not only affected by the ruggedness of the seafloor itself, but also by the thickness of the filling in the trench. However, this factor was not considered in this pre-stack depth migration, so further improvements are needed in some local areas.
7. Post-stack LIFT processing technology to improve signal-to-noise ratio
Due to the complex structure of this work area, the signal-to-noise ratio and frequency components of shallow, medium and deep layers are very different, we use LIFT to Noise processing technology can effectively improve the quality of processing results (Figure 5-10).
Figure 5-10 LIFT technology signal-to-noise ratio comparison chart