A seismic exploration method based on borehole information structural constraints

Abstract

This invention discloses a seismic exploration method based on borehole information structural constraints. It employs a drilling-while-drilling source to enable seismic wave excitation at any time during drilling. By continuously changing the source position during drilling, seismic wave data from multiple time periods can be acquired. The first numbered acquisition period is then inverted and imaged separately using conventional seismic inversion. Subsequent time periods are then inverted and imaged using time-shifted source data constrained by borehole information structural constraints. This means that the structural constraints of the previous time period's inversion results and the borehole information up to the current acquisition period are incorporated into the traditional seismic inversion objective function, effectively controlling the differences in subsequent time periods and resulting in higher resolution structural constraint inversion. This allows for accurate capture of changes in anomaly exposure during drilling, ultimately enabling effective differentiation of anomalies within the detection range. This achieves mutual verification between drilling and geophysical data, enabling precise detection of the drilling area.

Description

Technical Field

[0001] This invention belongs to the field of seismic exploration technology, specifically a seismic exploration method based on borehole information structural constraints. Background Technology

[0002] Seismic exploration is a key technology in geology and geophysics used to detect subsurface structures. It infers the properties and structure of subsurface rock strata by generating seismic waves at the surface or underground and recording their propagation. Traditional seismic exploration methods typically rely on actively generated seismic data from artificial sources such as impacts and explosives. However, this method has limitations, such as the influence of subsurface medium properties, signal aliasing from anomalies, and relatively low exploration accuracy. To overcome these limitations, borehole seismic exploration technology has emerged. Borehole seismic exploration can generate seismic waves at depths or far from the exploration surface, thereby obtaining more accurate reflection signals from anomalous media and providing more precise local subsurface medium information.

[0003] Traditional time-lapse seismic methods involve repeated three-dimensional (two-dimensional or other) seismic exploration of the subsurface medium at different times. By subtracting the time-delayed seismic images, the static properties of the subsurface medium are eliminated, thus directly obtaining the imaging results of the dynamic changes in the medium. However, this method is only suitable for situations where the subsurface medium changes rapidly within a certain period. Currently, most subsurface media undergo changes over a long period without human intervention, and the information about the subsurface medium is known, falling within the scope of monitoring. Therefore, the acquisition cycle is long, often with an interval of more than six months between two acquisitions, and the observation system is fixed. Thus, the aforementioned traditional time-lapse seismic interpretation method is not suitable for situations in borehole seismic exploration where the information of the exploration area is unknown, and the spatiotemporal location of the seismic source changes with the drilling progress. There is an urgent need to research rapid imaging algorithms for such borehole time-lapse seismic sources.

[0004] Full Waveform Inversion (FWI) is an advanced seismic data processing technique that optimizes the physical properties of the subsurface medium by minimizing the discrepancy between observed data and model predictions, providing high-resolution subsurface velocity models. However, traditional FWI methods require a highly accurate initial model, and the subsequently acquired actual seismic data often contains significant noise and incomplete data acquisition. Furthermore, the lack of effective constraints on the results during the inversion process ultimately limits the accuracy and resolution of the inversion.

[0005] Therefore, the research direction of this invention is to provide an improved seismic exploration method that not only makes up for the shortcomings of the full waveform inversion algorithm that requires a good initial model, but also takes into account the continuity of excitation and acquisition information in time and space of borehole sources, and constrains the inversion process based on borehole information structure, so as to effectively improve the accuracy and resolution of seismic exploration. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a seismic exploration method based on borehole information structure constraints. This method not only overcomes the shortcomings of the full waveform inversion algorithm, which requires a good initial model, but also considers the continuity of excitation and acquisition information in time and space of the borehole source. Based on the inversion process constrained by borehole information structure, it ultimately effectively improves the accuracy and resolution of seismic exploration.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: a seismic exploration method based on borehole information structural constraints, comprising the following steps:

[0008] Step 1: Determine the drilling and exploration distance and data collection time period: Based on prior geological data, obtain the required exploration range, determine and number the number of data collection time periods during the drilling process, and ensure that the number of data collection times is the same within each data collection time period.

[0009] Step 2, Borehole Information Acquisition: Drill the first borehole into the detection area and install multiple geophones inside the borehole. Then, connect the drilling source to the drill rod and drill bit and drill the second borehole into the detection area. During the drilling of the second borehole, according to the acquisition time set in Step 1, the drill bit stops drilling and the drilling source is used to excite seismic waves at each time point. Each geophone receives the waveform data to acquire the borehole information for that time until the acquisition process for all time points is completed.

[0010] Step 3, First Inversion Imaging: For the data of the first numbered acquisition period, use the data within that period to set up an initial model, and use the conjugate gradient optimization algorithm to perform a separate full waveform inversion on the data to obtain the inversion imaging results for that acquisition period;

[0011] Step 4, Second Inversion Imaging: For the data of the second acquisition period, the inversion imaging result of the first acquisition period is used as the initial model for the second inversion. The objective function of the second inversion based on the borehole information structure constraint is established by combining the borehole information of the first and second acquisition periods. Then, the conjugate gradient optimization algorithm is used to perform full waveform inversion on the data of the second acquisition period to obtain the inversion imaging result of this acquisition period.

[0012] Step 5, Third Inversion Imaging: For the data of the third acquisition period, the inversion imaging results of the second acquisition period are used as the initial model for the third inversion. The objective function of the third inversion based on the borehole information structure constraint is established by combining the borehole information from the first to the third acquisition periods. Then, the conjugate gradient optimization algorithm is used to perform full waveform inversion on the data of the third acquisition period to obtain the inversion imaging results of this acquisition period.

[0013] Step Six: Seismic Exploration Imaging: Step Five is repeated for each subsequent time period until all numbered acquisition time period data are inverted, thereby forming the seismic exploration imaging of the required exploration area.

[0014] Furthermore, the number of data acquisition periods in step one is at least three. Setting at least three data acquisition periods is necessary to ensure that the method of this invention achieves the required accuracy.

[0015] Furthermore, in step one, the time interval between each sampling moment within each sampling period is the same. By setting the same sampling interval, it is possible to further observe the geological changes in the area to be detected at different times within the same time period, thereby more accurately revealing geological information and the development of anomalies during drilling.

[0016] Furthermore, the objective function used in the first inversion in step three is specifically:

[0017]

[0018] Where, m T1 This is the inversion imaging result from the first acquisition time period, W d Let be the covariance matrix of the data fitting term, where d is the forward operator and dbos,1 is the seismic data of the first acquisition period.

[0019] Furthermore, the objective function used in the second inversion in step four is specifically:

[0020]

[0021] Where, m T2 This is the inversion imaging result from the second acquisition time period, W d Let be the covariance matrix of the data fitting term, generally defined as the identity matrix, d be the forward operator, dbos,2 be the seismic data from the second acquisition period, β be the regularization parameter of the gradient structure constraint, and W be the covariance matrix of the data fitting term. m Let m be the covariance matrix of the borehole information structure constraint. T1+T2 This represents the model constraint information for borehole information collected during the first and second data collection periods. The model constraint information serves two purposes: first, it constructs the initial model for the next inversion based on the model results of the previous inversion (this step is equivalent to constraining the spatial location and morphology of the anomaly); second, it optimizes the properties of the constrained model based on the parameters and borehole information from the first inversion. The purpose of the model constraint information is twofold: first, to constrain the anomaly's morphology and size, i.e., geometric information; and second, to constrain the anomaly's wave velocity, density, etc., i.e., elastic parameters.

[0022] Furthermore, the objective function used in the third inversion in step five is specifically:

[0023]

[0024] Where, m T3 This is the inversion imaging result from the third acquisition period, W d Let be the covariance matrix of the data fitting term, d be the forward operator, dbos,3 be the seismic data from the third acquisition period, β be the regularization parameter of the gradient structure constraint, and W be the covariance matrix of the data fitting term. m Let m be the covariance matrix of the borehole information structure constraint. T1+T2+T3 This represents the model constraint information for borehole data collected during the three data acquisition periods, from the first to the third.

[0025] Furthermore, the borehole information refers to the lithological properties and elastic parameters obtained during the drilling process.

[0026] Due to the ambiguity of full-waveform seismic inversion, standalone seismic inversion cannot accurately obtain the spatial location and elastic parameters of anomalies, and may also exhibit artifacts. To address this, this invention employs a drilling-while-source (WSBS) system that allows for seismic wave excitation during drilling, with drilling stopped at any time. By continuously changing the source position during one drilling operation while a geophone in another borehole continuously receives data, seismic wave data from multiple time periods can be acquired. First, conventional seismic inversion is used for standalone imaging of the first numbered acquisition period. In subsequent periods, time-shifted source data inversion imaging with borehole information structural constraints is performed. This involves incorporating the inversion results from the previous period and structural constraints from the borehole information up to the current acquisition period into the traditional seismic inversion objective function, effectively controlling the differences between subsequent periods and resulting in higher resolution structural constraint inversion. This allows for accurate capture of changes in anomaly exposure during drilling, ultimately enabling effective differentiation of anomalies within the detection range. This achieves mutual verification between drilling and geophysical data, enabling precise detection of the drilling area. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of borehole exploration according to the present invention;

[0028] Figure 2 This is a flowchart of the time-shifted full-waveform inversion imaging method of the present invention;

[0029] Figure 3 It is a geological model diagram to be revealed in experimental verification;

[0030] Figure 4 This is a graph showing the data acquisition and inversion results during the T1 acquisition period as demonstrated in the experiment.

[0031] (a) is a schematic diagram of data acquisition; (b) is the result of a separate inversion.

[0032] Figure 5 This is a graph showing the data acquisition and inversion results during the T2 acquisition period as demonstrated in the experiment;

[0033] (a) is a schematic diagram of data acquisition; (b) is the result of inversion alone; and (c) is the result of inversion with T1 and T2 borehole information constraints added in this invention.

[0034] Figure 6 This is a graph showing the data acquisition and inversion results during the T3 acquisition period as demonstrated in the experiment;

[0035] (a) is a schematic diagram of data acquisition; (b) is the result of inversion alone; and (c) is the result of inversion with the constraint of borehole information from T1 to T3 in this invention.

[0036] Figure 7 This is a graph showing the data acquisition and inversion results during the T4 acquisition period as demonstrated in the experiment;

[0037] (a) is a schematic diagram of data acquisition; (b) is the result of inversion alone; and (c) is the result of inversion with the constraint of borehole information from T1 to T4 in this invention. Detailed Implementation

[0038] The present invention will be further described below.

[0039] like Figure 2 As shown, the present invention includes the following steps:

[0040] Step 1: Determine the drilling and sampling distance and time periods: Based on prior geological data, determine the required exploration range and identify n sampling time periods during drilling, numbered T1, T2, T3…Tn. The number of sampling moments within each time period is the same, and the interval between each sampling moment is also the same. By setting the same sampling interval, it is possible to further observe the geological changes in the area to be explored at different times within the same time period, thereby more accurately revealing geological information and anomaly development during drilling.

[0041] Step 2, Borehole Information Acquisition: A first borehole is drilled into the detection area using drill rod 1, and multiple geophones are installed within the borehole. Then, a second borehole is drilled into the detection area after connecting the drilling seismic source to drill rod 2 and the drill bit. During the drilling of the second borehole, according to the acquisition times set in Step 1, the drill bit stops drilling and the drilling seismic source is used to excite seismic waves at each time point. Each geophone receives the waveform data to acquire the borehole information for that time, until the acquisition process for all times is completed. The drilling seismic source is existing equipment that can be purchased directly from the market; this invention only utilizes its drilling excitation function.

[0042] Step 3, First Inversion Imaging: For the data acquired during time period T1, an initial model is set up using the data within that time period, and the conjugate gradient optimization algorithm is used to perform a separate full-waveform inversion of the data to obtain the inversion imaging results for that acquisition period. The objective function used for the inversion is as follows:

[0043]

[0044] Where, m T1 This is the inversion imaging result from the T1 acquisition period, W d Let be the covariance matrix of the data fitting term, where d is the forward operator and dbos,1 is the seismic data acquired during the T1 acquisition period.

[0045] Step 4: Second Inversion Imaging: For the data acquired during period T2, the inversion imaging results from period T1 are used as the initial model for the second inversion. A target function for the second inversion, based on borehole information structural constraints, is established by combining the borehole information from both periods T1 and T2. Then, the conjugate gradient optimization algorithm is used to perform full waveform inversion on the data acquired during period T2 to obtain the inversion imaging results for that period. The specific target function used for the inversion is as follows:

[0046]

[0047] Where, m T2 This is the inversion imaging result from the T2 acquisition period, W d Let be the covariance matrix of the data fitting term, generally defined as the identity matrix, d be the forward operator, dbos,2 be the seismic data acquired during the T2 acquisition period, β be the regularization parameter of the gradient structure constraint, and W be the covariance matrix of the data fitting term. m Let m be the covariance matrix of the borehole information structure constraint. T1+T2 This represents the model constraint information for borehole data collected during two time periods, T1 and T2. This model constraint information serves two purposes: first, it constructs the initial model for the next inversion based on the model results of the previous inversion (this step is equivalent to constraining the spatial location and morphology of the anomaly); second, it optimizes the properties of this constrained model based on the parameters from the first inversion and the borehole information. The purpose of the model constraint information is twofold: first, to constrain the anomaly's shape and size, i.e., geometric information; and second, to constrain the anomaly's wave velocity, density, etc., i.e., elastic parameters. For example, if the actual anomaly is 5x5 in size with a wave velocity of 1500 m / s, the first inversion result yields an anomaly of 7x7 in size and a wave velocity of 1300 m / s. The second inversion introduces structural constraints, and the constraint model is also set based on the results of the first inversion and the borehole information, for example, set to 6x6 with a wave velocity of 1400 m / s. The resulting inversion shows an anomaly of 5x5 in size and a wave velocity of 1450 m / s. This allows for more accurate inversion results.

[0048] Step 5, Third Inversion Imaging: For the data acquired during period T3, the inversion imaging results from period T2 are used as the initial model for the third inversion. Combined with borehole information from periods T1 to T3, an objective function for the third inversion based on borehole information structural constraints is established. Then, the conjugate gradient optimization algorithm is used to perform full waveform inversion on the data acquired during period T3 to obtain the inversion imaging results for that period. The specific objective function used for the inversion is as follows:

[0049]

[0050] Where, m T3 This is the inversion imaging result from the T3 acquisition period, W d Let be the covariance matrix of the data fitting term, d be the forward operator, dbos,3 be the seismic data acquired during the T3 acquisition period, β be the regularization parameter of the gradient structure constraint, and W be the covariance matrix of the data fitting term. m Let m be the covariance matrix of the borehole information structure constraint. T1+T2+T3 This represents the model constraint information for the borehole data collected during the first three data acquisition periods, from T1 to T3.

[0051] Step Six: Seismic Exploration Imaging: Step Five is repeated for each subsequent time period until all numbered acquisition time period data are inverted, thereby forming the seismic exploration imaging of the required exploration area.

[0052] Experiments have shown that:

[0053] To verify the effectiveness of the inversion imaging method of this invention, a numerical simulation experiment was conducted, using the area to be drilled ahead of the tunnel face as the experimental setup. To simplify calculations, the model size was set to 100m × 50m, the background wave velocity was set to 1000m / s, and the wave velocities of all anomalies were set to 1500m / s. dx = dz = 1m, dt = 0.0001s, and T = 0.17s. Three anomalies were set, designated as anomaly 1 to 3. The specific geological model is as follows... Figure 3 As shown; next, an observation system was set up, with geophones covering the entire model space along the borehole axis, spaced 5m apart, for a total of 20 detectors; during the drilling process, 20 acquisition times were set, acquiring the seismic waves generated by the drill bit source every 5m of drill bit advance. The 20 acquisition times were divided into 4 acquisition periods, meaning each acquisition period included 5 acquisition times of excitation. The inversion parameters were consistent with the forward modeling process. In particular, to quickly verify the effectiveness of the method of this invention, each inversion was only set to 10 iterations, using the conjugate gradient optimization algorithm for inversion iterative calculation, and an adaptive step size calculation method to calculate the step size of each iteration.

[0054] First, the data collected in time period T1 was inverted separately. The inversion results are as follows: Figure 4 As shown in (b), it is compared with Figure 4(a) Comparison with the actual geological model during the T1 acquisition period shows that, since the seismic waves generated by the drill bit source during this period are mostly reflected waves from the left side of the anomaly and the signal energy of the distant anomaly 3 is weak, only the approximate locations of anomaly 1 and anomaly 2 are reflected, while the center velocity differs greatly from the real model, and there are many false anomalies.

[0055] The drill bit continued drilling, and the data from the T2 acquisition period were inverted separately, as well as inverted with borehole information structure constraints. The inversion results are as follows: Figure 5 As shown, firstly Figure 5 (b) and Figure 5 (c) and Figure 4 (b) A comparison shows that, due to the change in the spatial position of the drill bit, the acquired signals include reflected wave signals from the right side of anomaly 1, supplementary information from anomaly 2, and reflected signals from the left side of anomaly 3. Therefore, both methods can invert the positions of the three anomalies, and the overall inversion effect of each anomaly is better than the individual inversion during the T1 acquisition period; Figure 5 (b) and Figure 5 (c) It can be seen that the inversion effect is better after adding the borehole information structure constraint. The location of the anomaly is more accurately delineated and the center velocity is closer to the real model. However, due to the model structure constraint during the T1 acquisition period, the anomaly 3 can only be seen in approximate location.

[0056] Subsequently, the borehole information acquired by the drill bit during the T3 acquisition period was inverted separately, and the borehole information structure constraint was also inverted. The inversion results are as follows: Figure 6 As shown, firstly Figure 6 (b) and Figure 5 (b) A comparison shows that, since the drill bit is located between anomaly 2 and anomaly 3 at this time, the acquired signal mainly covers the right side of the model. Therefore, the inversion position of anomaly 3 is clearer, and the right-side inversion effect of anomaly 1 and anomaly 2 is better than that of the acquisition periods T1 and T2; Figure 6 (b) and Figure 6 (c) It can be seen that the inversion results of the positions and central velocities of the three anomalies after adding borehole information structural constraints are better than the individual inversion results; comparison Figure 6 (c) and Figure 5 (c) It can be seen that, without affecting the inversion effect of anomaly 1 and anomaly 2, anomaly 3 is clearly revealed.

[0057] Finally, the seismic data generated by the drill bit during the T4 acquisition period were inverted separately and inverted with borehole information structural constraints. The inversion results are as follows: Figure 7 As shown, comparison Figure 7 (b) and Figure 6(b) It can be seen that, since the drill bit further penetrated into the right region of the detection model at this time, the inversion results of the right side of anomaly 3 and anomaly 2 were better, while the inversion result of anomaly 1 was not as good as that of the T3 acquisition period; in comparison Figure 7 (b) and Figure 7 (c) It can be seen that after adding the structural constraints of the borehole information during the T3 acquisition period, most of the false anomalies in the inversion were eliminated, and the center velocity value was closer to the real model; compared with Figure 7 (c) and Figure 6 (c) It can be seen that the location of the inverted anomaly is more convergent and more consistent with the size of the anomaly in the real model, and the velocity value of the inversion center can be further optimized.

[0058] The above experiments demonstrate that the present invention can effectively combine borehole seismic information for full waveform inversion, accurately reveal the location of unknown anomalies during drilling as time progresses and the drill bit position changes, and ultimately achieve integrated geological interpretation of drilling and geophysical exploration.

[0059] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A seismic exploration method based on borehole information structural constraints, characterized in that, Includes the following steps: Step 1: Determine the drilling and exploration distance and data collection time period: Based on prior geological data, obtain the required exploration range, determine and number the number of data collection time periods during the drilling process, and ensure that the number of data collection times is the same within each data collection time period. Step 2, Borehole Information Acquisition: Drill the first borehole into the detection area and install multiple geophones inside the borehole. Then, connect the drilling source to the drill rod and drill bit and drill the second borehole into the detection area. During the drilling of the second borehole, according to the acquisition time set in Step 1, the drill bit stops drilling and the drilling source is used to excite seismic waves at each time point. Each geophone receives the waveform data to acquire the borehole information for that time until the acquisition process for all time points is completed. Step 3, First Inversion Imaging: For the data of the first numbered acquisition period, use the data within that period to set up an initial model, and use the conjugate gradient optimization algorithm to perform a separate full waveform inversion on the data to obtain the inversion imaging results for that acquisition period; Step 4: Second Inversion Imaging: For the data from the second acquisition period, the inversion imaging results from the first acquisition period are used as the initial model for the second inversion. A target function for the second inversion, based on borehole information structural constraints, is established by combining the borehole information from both the first and second acquisition periods. Then, the conjugate gradient optimization algorithm is used to perform full waveform inversion on the data from the second acquisition period to obtain the inversion imaging results for that period. The specific target function for the second inversion is: in, These are the inversion imaging results from the second acquisition period. Let be the covariance matrix of the data fitting term, and d be the forward operator. This refers to the earthquake data collected during the second time period. It is the regularization parameter of the gradient structure constraint. Let be the covariance matrix of the borehole information structural constraints. The model constraint information represents the borehole information collected during the first and second data collection periods. Step 5, Third Inversion Imaging: For the data from the third acquisition period, the inversion imaging results from the second acquisition period are used as the initial model for the third inversion. Combined with the borehole information from the first to third acquisition periods, an objective function for the third inversion based on borehole information structural constraints is established. Then, the conjugate gradient optimization algorithm is used to perform full waveform inversion on the data from the third acquisition period to obtain the inversion imaging results for that acquisition period. The specific objective function for the third inversion is: in, These are the inversion imaging results from the third acquisition period. Let be the covariance matrix of the data fitting term, and d be the forward operator. This refers to the seismic data collected during the third acquisition period. It is the regularization parameter of the gradient structure constraint. Let be the covariance matrix of the borehole information structural constraints. Model constraint information representing borehole information from the first to the third data collection period; Step Six: Seismic Exploration Imaging: Step Five is repeated for each subsequent time period. The inversion results of the previous time period and the borehole information structure constraints up to the current acquisition time period are introduced into the traditional seismic inversion objective function until all numbered acquisition time period data are inverted, thereby forming the seismic exploration imaging of the required exploration area.

2. The seismic exploration method based on borehole information structural constraints according to claim 1, characterized in that, The number of data collection periods in step one is at least three.

3. The seismic exploration method based on borehole information structural constraints according to claim 1, characterized in that, In step one, the interval between each collection moment within each collection period is the same.

4. The seismic exploration method based on borehole information structural constraints according to claim 1, characterized in that, The objective function used in the first inversion in step three is as follows: in, These are the inversion imaging results from the first acquisition period. Let be the covariance matrix of the data fitting term, where d is the forward operator. This refers to the earthquake data collected during the first time period.

5. The seismic exploration method based on borehole information structural constraints according to claim 1, characterized in that, The borehole information refers to the lithological properties and elastic parameters obtained during the drilling process.

Images