WO2020087767A1

WO2020087767A1 - Velocity inversion method based on jointly collected station and three-dimensional seismic data

Info

Publication number: WO2020087767A1
Application number: PCT/CN2019/000145
Authority: WO
Inventors: 尚新民; 芮拥军; 崔庆辉; 赵胜天; 刁瑞; 冮明川; 隋旭强; 王蓬
Original assignee: 中国石油化工股份有限公司; 中国石油化工股份有限公司胜利油田分公司物探研究院
Priority date: 2018-10-31
Filing date: 2019-07-18
Publication date: 2020-05-07
Also published as: CN109100792A; CN109100792B

Abstract

A velocity inversion method based on jointly collected station and three-dimensional seismic data. The velocity inversion method based on jointly collected station and three-dimensional seismic data comprises station deployment position design (101), joint collection of station and three-dimensional seismic data, station and three-dimensional seismic observation system coordinate matching (102), station data interception (103), station and three-dimensional seismic data matching (104), station preliminary wave travel time pickup (105), three-dimensional seismic reflected wave travel time pickup (106), initial velocity model building (107), joint inversion of preliminary wave and reflected wave travel time (108) and velocity model output (109). The velocity inversion method based on jointly collected station and three-dimensional seismic data can reduce the non-uniqueness of deep velocity inversion, improve the accuracy of the deep velocity inversion and improve the imaging quality of deep seismic data.

Description

Velocity inversion method based on joint acquisition of station and 3D seismic data

Technical field

The invention relates to the field of oil and gas seismic exploration, in particular to a speed inversion method based on the joint acquisition of data by a station and a three-dimensional earthquake.

Background technique

Since the topic of seismic velocity imaging research content was set at the 54th Annual Meeting of Geophysical Prospectors Association held in Atlanta, after a lot of research represented by Daily, Somerstein, Bishop, Dyer and Worthington, seismic wave velocity inversion Theories, methods and techniques have been rapidly developed by numerical simulation. As for the important branch of joint velocity inversion when traveling with multiple waves, the current status of its development is as follows: In 1992, Zelt proposed an earthquake based on any type of seismic body wave while determining the velocity and interface morphology of the two-dimensional crust Inversion method when traveling. In 1995, Hua Biaolong deduced the formula for the partial derivative of the interface suitable for the reflection and transmission of seismic waves in two-dimensional layered media, and applied it to the joint inversion of interface and velocity in two-dimensional layered media. play. In 1997, McCaughey used transmission waves, wide-angle reflection and refraction travel to simultaneously invert the velocity and interface depth. This method stabilized the inversion by weighing the values of different dimensions in the coefficient matrix and adding regularization to the inversion Solution. In 1998, Zhang Jie proposed a non-linear refraction and reflection wave travel time inversion method, this method uses the refraction and reflection waves travel at the same time inversion. In 2003, James used transmission, wide-angle reflection, refraction, and multiple-wave travel to perform simultaneous three-dimensional velocities and interface velocities to minimize data errors and model roughness during inversion to find a suitable layer velocity-interface model. In 2003, Zhou proposed a deformable tomography method based on inversion interface morphology. Since most sedimentary basins have layered velocity structures, the current tomography method based on grid and node parameterization is difficult to reconstruct layered velocity structures with high accuracy. In 2006, when the model was parameterized, Zhou introduced multi-scale parameterization to develop the first to multi-scale deformable tomography method and the reflection scale deformable tomography method. Compared with single-scale tomography, multi-scale tomography The results are more stable and reliable. In 2006, Zhou Longquan deduced the formula for the partial derivative of the interface during seismic wave travel in a three-dimensional layered medium based on the principle of wave retrograde. Trial calculations prove that the method is effective for three-dimensional velocity structure and interface reconstruction. In 2011, Bai Chaoying and others introduced the weight coefficients of different seismic phase data and the normalization factors of different types of parameters into the inversion algorithm, and realized the joint inversion imaging of multi-seismic phase travel in 3D complex layered media.

Research on the joint inversion problem of multi-seismic phases has a long history, but it is facing the problem of basin-scale macroscopic velocity inversion, and the existing technology is based on the same observation system and the same detector, the solution is The problem of speed inversion accuracy. How to make good use of the advantages of the first arrival wave and the reflected wave to improve the accuracy of the inversion of the deep velocity model of oil and gas fields, no publicly published technical results, this is because of the limited economic cost and construction feasibility, the existing oil and gas seismic exploration and observation system The maximum offset is limited, and seismic wave information from deep layers cannot be received.

In recent years, Shengli Oilfield has witnessed a significant increase in deep oil and gas reserves and rich geological reserves of oil and natural gas. However, the overall exploration degree is relatively low, and it is difficult to implement structures and difficult to describe effective reservoirs. Although the high-precision three-dimensional data acquired twice greatly improves the imaging quality of deep seismic data, there are still some ambiguities in the requirements of geological and reservoir description. The main contradiction is the accuracy of the deep velocity model. To this end, we have invented a new velocity inversion method based on the joint acquisition of data from stations and 3D earthquakes to solve the above technical problems.

Summary of the invention

The object of the present invention is to provide a speed inversion method based on the joint acquisition of data from stations and three-dimensional earthquakes that can effectively improve the accuracy of deep layer velocity modeling.

The object of the present invention can be achieved by the following technical measures: a velocity inversion method based on the joint acquisition of data by the station and the 3D earthquake, the velocity inversion method based on the joint acquisition of data by the station and the 3D earthquake includes: Step 1: According to the 3D earthquake Observation system, combined with the actual surface conditions, lay out stations before the 3D seismic construction, and record continuously; Step 2: Match the station and the 3D seismic coordinate system to unify the coordinates of the two observation systems; Step 3: From continuous In the recorded station data, intercept the data within the shooting period; Step 4: Match the intercepted station record data with the three-dimensional seismic record data to synchronize the two data in time; Step 5: Match Pick up the first arrival time on the later station data; Step 6: Pick up the reflected arrival time on the 3D seismic data; Step 7: Establish the underground initial velocity model; Step 8: Use the picked station first arrival time Perform multi-scale joint tomographic inversion with 3D seismic reflection travel time; Step 9: When the inversion results converge, output the results as the final Speech velocity model.

The object of the present invention can also be achieved by the following technical measures:

In step 1, the layout of the station uses a parallel staggered pattern.

In step 1, the station used can record continuously for at least 4 days, the collected data is automatically stored on the memory card carried by itself, and it can continue to work after replacing the backup battery.

In step 1, the station adopts the method of digging and burying to ensure that the station is at least 60cm away from the surface.

In step 2, the coordinate matching method of the two observation systems is to use the coordinates of the detector at the point where the station and the detector coincide to obtain the coordinates of the unknown station, and to obtain the coordinates of the station that are consistent with the coordinates of the detector by solving the following equations :

Where (x, y) is the coordinates of the station to be sought, (x ₁ , y ₁ ), (x ₂ , y ₂ ), (x ₃ , y ₃ ) are the coordinates of the three stations near the station to be sought , D ₁ , d ₂ , and d ₃ are the distances from the station to the nearby three stations, and the distance is calculated by the coordinates measured by the handheld GPS device.

In step 3, intercept the data in the shooting time period from the continuously recorded station data, first sample and calculate the approximate time difference between the station timing system and the 3D seismic timing system, and then shoot each field shot in the 3D seismic field The time is the center, and the data within the time difference is intercepted before and after the output.

In step 4, before matching the intercepted station record data with the three-dimensional seismic record data, it is necessary to match the two recorded waveforms to keep the two recorded waveforms as consistent as possible.

In step 4, before matching the intercepted station record data with the three-dimensional seismic record data, it is necessary to filter and normalize the two records after waveform matching.

In step 4, the intercepted station record data is matched with the three-dimensional seismic record data by calculating the corresponding cross-correlation overlay value of the two records, and determining the accurate firing time on the station data by the maximum cross-correlation overlay value. And intercept the output station data according to the set data length.

In step 5, the first arrival pick-up on the station data is completed in the co-detection point data set of the station data, which is to extract the data from the common-shot point station into the co-detection point station data, and follow the deviation The shift distance is rearranged, and then the first arrival is automatically picked up using the energy ratio method of long and short time windows.

In step 6, the pick-up of the reflected wave arrival time on the 3D seismic data is done on the common shot point data set. When picking, the reflection in-phase axis with high signal-to-noise is selected as the standard reflection layer, which is performed automatically and manually. Pick up when the reflected wave arrives.

In step 7, the initial velocity model is a layered velocity model obtained by multiscale grid tomographic inversion of the first arrival travel time of the station data. The process is to discretize the actual geological model with grids of different sizes. Invert each scale model.

In step 8, the joint tomographic inversion of the first arrival travel time of the station and the 3D seismic reflection travel time is to discretize the underground velocity structure into a layered model, and to invert the fluctuations of the velocity interface and the velocity values within the layer.

In step 8, the joint inversion of the first arrival wave travel time of the station and the three-dimensional seismic reflection wave travel time is to calculate the perturbation amount of the change of the node position on each scale after multi-scale decomposition of the interface to the arrival time of the seismic wave, and use Inversion speed model of reflected wave travel time.

In step 8, the joint inversion of the first arrival travel time of the station and the three-dimensional seismic reflection travel time is to discretize the actual geological model with grids of different sizes, and invert the models at various scales simultaneously during the inversion process.

In step 8, in the joint tomographic inversion equation of the station's first arrival travel time and 3D seismic reflected wave travel time, both the first arrival wave and the reflected wave arrival time are used to invert the velocity model.

The speed inversion method based on the joint acquisition of data by stations and 3D earthquakes in the present invention changes the method of using a single observation system for conventional 3D earthquakes, and obtains a large offset by deploying sparse stations without substantially increasing the acquisition cost The first arrival wave information, combined with the three-dimensional seismic reflection information, to perform the first tomographic and reflected wave tomographic inversion can effectively improve the accuracy of deep velocity modeling. Using this method can reduce the multiple solutions of deep velocity inversion, improve the accuracy of deep velocity inversion, and improve the imaging quality of deep seismic data.

BRIEF DESCRIPTION

Figure 1 is a schematic diagram of the theoretical layered velocity model and ray tracing results of the Shengli oilfield A;

Figure 2 is a schematic diagram of the results of the first arrival wave inversion in the Shengli Oilfield Engineering Area;

Figure 3 is a schematic diagram of the results of inversion of reflected wave A in the Shengli Oilfield Engineering Area;

Figure 4 is a schematic diagram of the results of the joint inversion of the first arrival wave and reflected wave in the Shengli Oilfield Engineering Area;

5 is a three-dimensional seismic observation system diagram of the location (triangle) of the sparse station B in the Shengli Oilfield Engineering Area;

Figure 6 is a schematic diagram of a section of station data (common shot data set) intercepted by B of Shengli Oilfield;

7 is a schematic diagram showing the offset of the common station data set after the matching of Shengli Oilfield Engineering Area B is completed;

8 is a schematic diagram of the multi-scale joint tomographic inversion result of the first arrival wave and the reflected wave of a certain survey line of Shengli Oilfield Engineering Area B;

FIG. 9 is a flowchart of a specific embodiment of a speed inversion method based on combined data acquisition of a station and a three-dimensional earthquake according to the present invention.

detailed description

In order to make the above and other objects, features and advantages of the present invention more comprehensible, preferred embodiments are specifically listed below, together with the drawings, and described in detail as follows.

As shown in FIG. 9, FIG. 9 is a flowchart of the velocity inversion method based on the station and three-dimensional seismic joint acquisition data of the present invention.

Step 101: According to the 3D seismic observation system, combined with the actual surface conditions, the stations are laid out before the 3D seismic construction, and continuous recording is made; the layout of the stations adopts a parallel staggered pattern. The station used can record continuously for at least 4 days, the collected data can be automatically stored on the memory card carried by itself, and can continue to work after replacing the backup battery. The station adopts the method of digging and burying to ensure that the station is at least 60cm away from the surface.

Step 102: Match the station and the three-dimensional seismic coordinate system to unify the coordinates of the two observation systems; the coordinate matching method of the two observation systems is to use the detector coordinates of the point where the station and the detector coincide to obtain the unknown station The coordinates can be obtained by solving the following equations to obtain the coordinates of the station that are consistent with the coordinates of the geophone:

Step 103: Intercept the data in the blasting period from the continuously recorded station data; intercept the data in the blasting period from the continuously recorded station data, first sample the statistics of the station timing system and the three-dimensional seismic timing system Time difference range, then take the field shooting time of each shot recorded by the 3D seismic as the center, and intercept the data within the time difference range before and after output.

Step 104: Match the intercepted station record data with the 3D seismic record data to synchronize the two data in time; before matching the intercepted station record data with the 3D seismic record data, two The matching of the recorded waveforms makes the two recorded waveforms as consistent as possible. Before matching the intercepted station record data with the three-dimensional seismic record data, it is necessary to filter and normalize the two records after waveform matching. The intercepted station record data is matched with the three-dimensional seismic record data by calculating the corresponding cross-correlation superposition value of the two types of records, and determining the accurate shooting time on the station data by the maximum cross-correlation superposition value, and according to the set The data length intercepts the output station data.

Step 105: Picking the first arrival arrival time on the matched station data; picking the first arrival arrival time on the station data is completed in the common detection point data set of the station data, which is about to fire the station The station data is extracted into common detection station data, and rearranged according to the distance of the offset, and then the first arrival is automatically picked up using the energy ratio method of long and short time windows.

Step 106: Picking up the reflected wave arrival time on the 3D seismic data; picking up the reflected wave arrival time on the 3D seismic data is done in the common shot data set, and selecting the reflection in-phase axis with higher signal-to-noise as the standard when picking up The reflective layer picks up the arrival of the reflected wave in an automatic and manual manner.

Step 107: Use the first arrival wave multi-scale tomographic inversion results to establish an underground layered initial velocity model; use the first arrival wave travel time of the station data, perform a multi-scale grid tomographic inversion, and use the inversion results to establish the layered initial velocity model.

Step 108: Use the tomographic inversion of the first arrival travel time of the station and the 3D seismic reflected wave travel time; the tomographic inversion of the first arrival wave travel time of the station and the 3D seismic reflection wave travel time is to discretize the underground velocity structure as Layered model, inversion of velocity interface fluctuations and velocity values in layers. In the joint tomographic inversion of the first arrival wave travel time of the station and the three-dimensional seismic reflection wave travel time, the amount of perturbation of the seismic wave arrival time at each scale after the multi-scale decomposition of the interface is calculated, and the first arrival wave and the reflected wave travel time are used at the same time. Inversion speed model. In the joint tomographic inversion equation of the first arrival wave travel time of the station and the three-dimensional seismic reflection wave travel time, the actual geological model is discretized with grids of different sizes (scales), and the inversion of each scale model is simultaneously performed during the inversion process.

Step 109: When the inversion result converges, the result is output as the final inversion speed model.

The following is a specific example of applying the present invention. Example 1 is derived from the inversion result of a theoretical velocity model of Sinopec Shengli Oilfield. The effectiveness of the method can be verified through this example. The specific implementation is:

(1) According to the results of seismic data interpretation in the work area, establish a two-dimensional layered velocity model, as shown in Figure 1;

(2) When the first arrival wave and the reflected wave travel are obtained by ray tracing calculation, the gray broken line in Figure 1 is the first arrival ray tracing result, and the black broken line in Figure 1 is the reflected wave ray tracing result;

(3) Only the first arrival travel time inversion speed model is used, the result is shown in Figure 2, and the result is quite different from the real speed model;

(4) Only use the reflected wave travel time inversion speed model, the result is shown in Figure 3, the result is better than the inversion using the first arrival wave alone;

(5) The joint tomographic inversion using the travel time of the first arrival wave and the reflected wave is shown in Fig. 4. The combined inversion result is significantly better than the inversion result of the first arrival wave or the reflected wave alone.

Example 2 is derived from the joint acquisition project of a station and three-dimensional earthquake in Shengli Oilfield of Sinopec. The specific implementation mode is:

(1) According to the 3D seismic observation system, combined with the actual surface conditions, the station is laid out before the 3D seismic construction and continuous recording is performed. The location of the station and the 3D seismic observation system are shown in Figure 5.

(2) Match the station and the three-dimensional seismic coordinate system to unify the coordinates of the two observation systems.

(3) From the continuously recorded station data, intercept the data within the firing period. Figure 6 shows the intercepted station data of a certain shot.

(4) Match the intercepted station record data with the three-dimensional seismic record data to make the two data synchronized in time. Figure 7 shows the common station data set of a station after the matching is offset. Display the results.

(5) Pick up the first arrival on the matched station data.

(6) Pick up the reflected wave arrival time on the three-dimensional seismic data.

(7) Use the first-order multi-scale grid tomographic inversion results to establish the underground layer initial velocity model.

(8) Joint tomographic inversion is performed using the picked first arrival travel time of the station and the 3D seismic reflection wave travel time. Figure 8 shows the results of the joint tomographic inversion of the first arrival travel time of the station and the 3D seismic reflection travel time.

(9) When the inversion result converges, the result output is used as the final inversion speed model.

Claims

The velocity inversion method based on the joint acquisition of data by the station and the 3D earthquake is characterized in that the velocity inversion method based on the joint acquisition of data by the station and the 3D earthquake includes:

Step 1: According to the 3D seismic observation system, combined with the actual surface conditions, lay out stations before the 3D seismic construction and record continuously;

Step 2: Match the station and the 3D seismic coordinate system to unify the coordinates of the two observation systems;

Step 3: From the continuously recorded station data, intercept the data within the shooting period;

Step 4: Match the intercepted station record data with the three-dimensional seismic record data to synchronize the two data in time;

Step 5: Pick up the first arrival on the matched station data;

Step 6: Pick up the reflected wave arrival time on the 3D seismic data;

Step 7: Establish the underground initial velocity model;

Step 8: Perform multiscale joint tomographic inversion using the picked station first arrival travel time and 3D seismic reflected wave travel time;

Step 9: When the inversion results converge, output the results as the final inversion speed model.
The speed inversion method based on the combined data of the station and the three-dimensional seismic data according to claim 1, characterized in that, in step 1, the layout of the station adopts a parallel staggered pattern.
The method of speed inversion based on joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that, in step 1, the station used can record continuously for at least 4 days, and the collected data is automatically stored in its own carry On the memory card, you can continue to work after replacing the backup battery.
The speed inversion method based on the data collected by the station and the three-dimensional seismic data according to claim 1, characterized in that, in step 1, the station adopts the method of digging and burial to ensure that the station is at least 60cm away from the surface.
The speed inversion method based on the joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that, in step 2, the two coordinate matching methods of the observation system are the detection using the coincidence point of the station and the geophone The coordinates of the unknown station are obtained by the coordinates of the detector, and the coordinates of the station consistent with the coordinates of the detector are obtained by solving the following equations:

Where (x, y) is the coordinates of the station to be sought, (x 1 , y 1 ), (x 2 , y 2 ), (x 3 , y 3 ) are the coordinates of the three stations near the station to be sought , D 1 , d 2 , and d 3 are the distances from the station to the nearby three stations, and the distance is calculated by the coordinates measured by the handheld GPS device.
The method of speed inversion based on the joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that, in step 3, the data within the shooting period is intercepted from the continuously recorded station data, and the statistics are first sampled The approximate time difference between the station timing system and the 3D seismic timing system, and then taking the field shooting time of each shot recorded by the 3D seismic as the center, intercepting the data output within the time difference range before and after.
The method of speed inversion based on the joint acquisition of data from stations and 3D earthquakes according to claim 1, characterized in that, in step 4, before matching the intercepted station record data with the 3D seismic record data, it is necessary to perform The matching of the two recorded waveforms makes the two recorded waveforms as consistent as possible.
The method of speed inversion based on the joint acquisition of data from a station and a 3D earthquake according to claim 1, characterized in that in step 4, before matching the intercepted station record data with the 3D seismic record data, the The two records after waveform matching are filtered and normalized.
The method of speed inversion based on joint acquisition of data from a station and a 3D earthquake according to claim 1, characterized in that in step 4, the intercepted station record data is matched with the 3D seismic record data by calculation Corresponding two kinds of recorded cross-correlation superimposed values, the maximum cross-correlation superimposed value is used to determine the accurate firing time on the station data, and the output station data is intercepted according to the set data length.
The method of speed inversion based on the joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that, in step 5, the first arrival of the station data is picked up at the time of co-detection of the station data It is completed in the point data set, that is, the data of the common shot station is extracted into the data of the common detection station, and rearranged according to the distance of the offset, and then the first arrival is automatically picked up by the energy ratio method of the long and short time windows.
The method of velocity inversion based on the joint acquisition of data from stations and 3D seismic data according to claim 1, characterized in that, in step 6, the pick-up of the reflected wave arrival on the 3D seismic data is performed on the common shot point data set When picking, select the reflection in-phase axis with high signal-to-noise as the standard reflection layer, and pick up the reflected wave arrival time through automatic and manual methods.
The velocity inversion method based on the station and 3D seismic joint acquisition data according to claim 1, characterized in that in step 7, the initial velocity model is a multi-scale grid layer using the first arrival travel time of the station data Analyzing the layered velocity model obtained by inversion, the process is to discretize the actual geological model with grids of different sizes, while inverting the model at each scale.
The velocity inversion method based on the joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that in step 8, the joint tomographic inversion of the first arrival wave travel time of the station and the three-dimensional seismic reflection wave travel time is Discrete the underground velocity structure into a layered model, and invert the fluctuations of the velocity interface and the velocity values within the layer.
The velocity inversion method based on the joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that, in step 8, the joint inversion of the first arrival wave travel time of the station and the three-dimensional seismic reflection wave travel time is calculated by After the multi-scale decomposition of the interface, the changes in the position of nodes at various scales disturb the arrival time of seismic waves, and the travel time inversion velocity models of first arrival waves and reflected waves are used at the same time.
The velocity inversion method based on the joint acquisition of data from a station and a three-dimensional earthquake according to claim 1, characterized in that in step 8, the joint inversion of the first arrival wave travel time of the station and the three-dimensional seismic reflection wave travel time are different Large and small grids discretize the actual geological model, and simultaneously invert all scale models during the inversion process.
The speed inversion method based on the joint acquisition data of the station and the three-dimensional earthquake according to claim 1, characterized in that in step 8, the first tomographic travel time of the station and the three-dimensional seismic reflected wave travel time joint tomographic inversion equation At the same time, the first arrival wave and the reflected wave arrival are used to invert the velocity model.