WO2016011587A1 - Bridge calibration method for vertical seismic profile data of large inclined-angle area - Google Patents

Abstract

A bridge calibration method for vertical seismic profile (VSP) data of a large inclined-angle area, comprising the following steps: 1) collecting VSP data, and performing TAR compensation and channel equalization during VSP processing to obtain accurate unconverted NMO data; dividing the NMO data into two parts, one part comprising a description such as a volume header and a channel header in a SEG-Y data format and the other part comprising pure numerical values; 2) performing an NMO data conversion on the pure numerical values of the NMO data according to a specific method; and 3) outputting a through-well seismic profile, well log data, a well log profile, and the converted matrix data obtained in step 2) as images, combining the images, and performing VSP calibration based on actual drilling stratification data.

Description

 A vertical seismic data bridge calibration method for large dip areas

TECHNICAL FIELD The present invention belongs to the field of unconventional oil and gas seismic exploration, and is a vertical seismic bridge calibration method using a normal dynamic correction (NMO) data for large dip areas. BACKGROUND OF THE INVENTION Bridge calibration is one of the important uses of vertical seismic (VSP) data. It is the use of VSP data to establish the relationship between ground seismic reflection and the formation in the well. The usual practice is to use the VSP time-depth relationship to the NMO profile and corridor of the VSP. The superimposed profile, the cross-well seismic profile, the logging curve, and the lithology histogram are combined to obtain the VSP bridge calibration results. With the continuous expansion of VSP applications in recent years, VSP collections have been carried out in many large dip areas. The existing VSP corridor calibration process is: First process the obtained MO data, and then at a near wellhead location along a certain time (or Depth) Perform a narrow corridor resection, and then use the data superposition method to obtain the corridor superimposed section. Due to the existence of the dip angle of the formation, the upstream wave reflection in the data NMO wave field has an angle with the time axis (or the depth axis), so the existing method cannot be superimposed in phase to obtain a reliable VSP corridor profile, and the existing bridge type The calibration method is mainly based on the comparison of the wave resistance characteristics of the VSP corridor and the cross-section seismic section. The unreliable corridor greatly affects the accuracy of the VSP bridge calibration. SUMMARY OF THE INVENTION It is an object of the present invention to provide a calibration method for vertical seismic data bridge calibration in a large dip area using converted NMO data with accurate calibration and high reliability.

 The invention is achieved by the following steps:

1) Collecting VSP data, performing TAR compensation (true amplitude recovery) and track equalization processing in the VSP processing, and obtaining the pre-conversion NMO data; The NMO data is separated into two parts, one of which is a description part of a data head format SEGY, a head, and the like, and the other is a pure value part;

 The TAR compensation and the track equalization process ensure that the energy at different positions in the section is relatively stable, and no energy imbalance occurs due to the conversion process.

 The described process is identical in both the time domain and the depth domain.

 The data format is SEGY format.

 2) Perform NMO data conversion on the pure value part of the NMO data as follows:

 1 Arrange the NMO pure value parts in a matrix, the number of rows of the matrix is the maximum number of samples, and the number of columns is the total number of tracks included in the NMO data;

 2 Using the method of picking up from the first time in the seismic processing, picking up the original first arrival time series of the NMO pure value part data, each channel corresponding to a first arrival time;

 3 Perform an integer calculation on the first arrival time series, so that the time values of each point are integer multiples of the sampling rate, and the new first arrival time series is obtained.

 The calculation is performed as follows: r = int(-)*At ( 1 )

 At

 Where r is the calculated new first arrival time series, τ is the original first arrival time sequence picked up in step 2, At is the sampling rate of the data, and int is the rounding function;

 The sample rate is the data recording time divided by the maximum number of samples.

 4 segmenting each of the steps 1 matrix according to the new first arrival time sequence to obtain a new matrix containing the series of sub-matrices;

 5 will get a new matrix for data transformation operations;

 The data conversion operation is performed as follows:

 First, the right lower triangular matrix of the matrix is taken out, then the triangular array is flipped left and right, and the submatrix in the inverted triangular matrix is rearranged and returned to the submatrix position of the original triangular matrix, and then the new triangular array replacement step is performed. 5 The lower right triangle part of the new matrix, and finally the respective sub-matrices in the new matrix are expanded to obtain the converted matrix data of exactly the same size as step 1.

 6 Combining the description information of the roll head and the track head separated in step 1) with the converted matrix data to obtain new NMO seismic data in SEGY format;

 3) Perform VSP bridge calibration as follows:

The converted seismic data, the logging data, the logging profile, and the converted matrix data obtained in step 2) are image-outputted, and then combined, and the VSP calibration is performed with the solid-layered hierarchical data. The present invention is different from the current VSP bridge calibration method in that the VSP corridor superposition profile is no longer needed in the calibration result, because the seismic reflection axis near the wellhead has been transferred to the above series of conversions to

On the side of the NMO section, it can be directly compared with the ground seismic section. When avoiding the problem of superposition of the corridor in the large dip area, a more intuitive and reliable bridge calibration result is obtained. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of original NMO data (left) and converted NMO data (right); FIG. 2 is an original NMO wave field (left) and converted NMO (^) of actual VSP data. The near-wellhead cut-off section of the NMO wave field (left), the VSP corridor superimposed section (middle), and the VSP corridor are inlaid into the cross-well seismic section (right); Figure 4 shows the spliced contrast of the converted NMO wavefield and the cross-well seismic section; An example of VSP calibration for converting a NMO wave field to a well in the Tarim Kuqa area. DETAILED DESCRIPTION OF THE INVENTION The specific embodiments of the present invention are implemented by the following steps:

 1) Using the VSP data obtained from the set, the conventional seismic data processing is performed, wherein the two-step process of TAR compensation and track equalization affects the energy distribution in the profile, and the optimal parameters are determined by contrast observation, and then processed to obtain fidelity. Pre-conversion NMO data;

 The NMO data is separated into two parts, one of which is the description part of the data format SEGY, the head, and the like, and the other is the pure value part.

 2) Perform data conversion on the pure value part of the NMO data as follows:

1 Record the NMO pure value part obtained in the previous step, which is composed of X « elements, where ^ΐ: is the data sample value, and the double subscript indicates the position of the element in the matrix, as the position in the matrix is the first Line 1 歹 ij, whose physical meaning is the first sample value of the first track in the NMO data, and so on, is the total number of tracks included in the NMO data, ^ is the maximum number of data points, as shown in the following equation :

The sampling interval of the sampling interval ^Δέ and λ^, which is known from the SEGY data head and the head information, is divided by the maximum number of samples to obtain the sampling interval.

2 Using the first-time pick-up method in the seismic processing, picking up the VSP first arrival time of the NMO data, and obtaining the first arrival time sequence', which is composed of ^3⁄4 elements, each of which is the first arrival time in each track, and its value follows the VSP The depth of the collection increases and increases.

τ = j¾ ¾ ... ^ - <<...<-<ί ¾) (3)

3 Perform an integer sample calculation on time series ⁷ so that the time values of each point are integer multiples of the sampling rate, and obtain a new first arrival time sequence, as shown in equation (1), where ^r is the calculated new first arrival time. The sequence, ^τ is the time series picked up in step (2), ^Δί is the sampling rate of the data, and ^ί is the rounding function (the rounding step makes the first ^arrival precision lower, but the first ^arrival is only used for data conversion. and the data conversion process, the maximum time accuracy is the ^Δί, so that the beginning to the variation in the precision for the accuracy of the process flow is not affected. the (3) into (1) can be calculated new time series ^r Each element in the sequence corresponds to the first arrival time of the integer multiple of Δί, as shown in the following equation: = ... ] ( ^ <<...< d < 4) (4)

4 Since each element t' in Τ' represents a different first arrival time, according to this time, each data can be divided into + ¹ ) segments, and then (2) is transformed into:

! =

Where ^ is a segmented matrix, which consists of ( + 3⁄4 ^x sub-matrix columns. For the sake of simplicity, each sub-matrix is represented by the above formula. At this time, the double subscript indicates the position of the sub-matrix in the large matrix. :

 5 will get the new matrix for data transformation, according to the following method:

X

a) Take the upper right triangle of the upper step matrix -

 b) flipping the triangle array left and right to obtain, as shown in equation (8);

.bottom right

 C) rearranging the order of the sub-matrices in the inverted triangular matrix back to the original triangular matrix matrix position, and obtaining "', as shown in equation (9);

d) replacing the new triangular matrix with the lower right triangle of the matrix ^ in step (4) to obtain a new matrix ^y , as shown in equation (10); e) finally expanding each submatrix in the new matrix ⁵ ^, obtaining the step with 1) Converted matrix data of exactly the same size.

(8) - 0 0 0 ..,. _r 0 0 !

 0 0 0 .. 0

 0 0 0 . !

 0 Q 0 .. " 1}

0

 -i) ( 9 )

Figure 1 is a schematic diagram of NMO data before and after conversion, which visually indicates the change in sample position after the above calculation.

6 Combine the description information of the roll head and the track head separated in step 1) with the converted matrix data F to obtain new NMO seismic data in SEGY format. Figure 2 shows an example of the NMO data before and after the vertical seismic survey in the large dip area. It can be seen that the fidelity conversion data is obtained according to the above method.

 3) Perform VSP bridge calibration as follows:

 The converted seismic profile, the logging data, the logging profile, and the converted NMO data obtained in step 2) are image-outputted, and then combined, and then the VSP calibration is performed with the solid-layered hierarchical data. Different from the current VSP bridge calibration method, the VSP corridor superposition profile is no longer needed in the calibration results. Because of the above series of transformations, the seismic reflection axis near the wellhead has been transferred to the NMO profile side, and the ground seismic profile. It can be directly compared, and the bridge calibration results can be obtained more intuitively and reliably while avoiding the problem of overlapping the corridors in the large dip area.

 Fig. 3 is an effect diagram of the existing corridor resection and superposition method for seismic section calibration, and Fig. 4 is an effect diagram of the new method. The comparison shows that the new method has obvious advantages. Figure 5 shows a calibration example of a new method for a VSP project in the Tarim region. As can be seen from the calibration chart, the new method preserves all the functions of the existing bridge calibration, while avoiding the fact that the formation reflections in the large dip area cannot be superimposed in phase. The problem is that the NMO profile and the seismic profile are seamlessly connected at the well location after conversion, and the wave resistance characteristics are intuitive and reliable. The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and modifications may be made without departing from the spirit and scope of the invention. Simplifications, which are equivalent replacement means, are included in the scope of the present invention.

Claims

Request

1. A vertical seismic data bridge calibration method using a large dip area, which is characterized by the following steps:

1) Collecting VSP data, performing TAR compensation and channel equalization processing in VSP processing, and obtaining pre-conversion NMO data;

 Separating the NMO data into two parts, one of which is the description part of the data format SEGY, the head of the track, and the like, and the second part is a pure value part;

 2) Perform NMO data conversion on the pure value part of the NMO data as follows:

 1 Arrange the NMO pure value parts in a matrix, the number of rows of the matrix is the maximum number of samples, and the number of columns is the total number of tracks included in the NMO data;

 2 Using the method of picking up from the first time in the seismic processing, picking up the original first arrival time series of the NMO pure value part data, each channel corresponding to a first arrival time;

3 Perform an integer-like calculation on the first-time time series so that the time values of each point are integer multiples of the sampling rate, and obtain a new first-time time series ^τ ':

 4 segmenting each of the steps 1 matrix according to the new first arrival time sequence to obtain a new matrix containing the series of sub-matrices;

 5 will get a new matrix for data transformation operations;

 6 Combine the description information of the roll head and the track head separated in step 1) with the converted matrix data to obtain new seismic data in SEGY format;

 3) Perform VSP bridge calibration as follows:

 The converted seismic data, the logging data, the logging profile, and the converted matrix data obtained in step 2) are image-outputted, and then combined, and the VSP calibration is performed with the solid-layered hierarchical data.

 2. The method for calibrating a vertical seismic data bridge using a large dip area according to claim 1, wherein: the step 1) the TAR compensation and the channel equalization processing are to ensure that the energy at different positions in the section is relatively stable. , there is no energy imbalance due to conversion processing.

 3. The vertical seismic data bridge calibration method using a large dip area according to claim 1, wherein: step 1) the track equalization processing is exactly the same in the time domain and the depth domain.

 4. The method of claim 1, wherein the step data format is SEGY format.

5. The vertical seismic data bridge calibration method using a large dip area according to claim 1, wherein: step 2) said integer integer calculation is performed according to the following formula: r' = int(1) * At

 At

 Where r is the calculated new first arrival time series, τ is the original first arrival time sequence picked up in step 2, Δ is the sampling rate of the data, and int is the rounding function;

 The sample rate is the data recording time divided by the maximum number of samples.

 6. The vertical seismic data bridge calibration method using a large dip area according to claim 1, wherein: step 2) said data conversion operation is performed according to the following method:

 First, the right lower triangular matrix of the matrix is taken out, then the triangular array is flipped left and right, and the submatrix in the inverted triangular matrix is rearranged and returned to the submatrix position of the original triangular matrix, and then the new triangular array replacement step is performed. 5 The lower right triangle part of the new matrix, and finally the respective sub-matrices in the new matrix are expanded to obtain the converted matrix data of exactly the same size as step 1.