Coal seam exploration logging gas reservoir identification method

Abstract

The invention belongs to the technical field of coal bed gas detection, and discloses a coal bed exploration logging gas reservoir identification method, which comprises the steps of data acquisition, data preprocessing, gas reservoir preliminary identification, multi-dimensional cross validation, and gas content and development value evaluation. According to the coal seam exploration logging gas reservoir identification method, firstly, association of logging, a rock core and three-dimensional seismic data is established by means of well-seismic calibration on the basis of multi-source data acquisition, then a three-dimensional cross validation closed loop is formed through rock core logging point verification and well-seismic space verification, and the precision of single well local identification is guaranteed through rock core experimental data; according to the method, seismic data are used for checking the spatial suitability of a gas reservoir, a structure height part and a fracture area, the boundary of the gas reservoir is corrected in combination with wave impedance abnormal characteristics, the limitation of traditional single rock core verification is eliminated, it is ensured that the recognition result is highly consistent with a geological structure rule, and accurate parameters are obtained through fusion of rock core actual measurement and logging inversion.

Description

Technical Field

[0001] This invention relates to the field of coalbed methane detection technology, specifically a method for identifying gas layers in coalbed exploration wells. Background Technology

[0002] The method of identifying gas layers in coal seam exploration logging is a key technology connecting underground reservoirs and surface development decisions in coalbed methane exploration and development. Essentially, it analyzes logging data obtained during drilling to solve the core problem of whether underground coal seams contain gas and whether such gas has development value. Coalbed methane exhibits strong heterogeneity; within the same coal seam, some areas contain gas while others do not, and the physical properties of gas layers differ from those of water and dry layers. One of the core functions of logging is to distinguish these strata types through differences in physical response, avoiding blind exploration. Coal seam exploration is costly, but logging evaluation of gas layers can effectively reduce the risk of failed exploration wells, screen favorable blocks, and divide favorable and unfavorable gas-bearing areas by comparing logging data from multiple wells, thus avoiding excessive investment in unfavorable areas.

[0003] Meanwhile, as a new type of clean energy, coalbed methane plays a vital role in the development of the national economy through its exploration and development. Currently, the exploration, development, and well location deployment of coalbed methane in China are mainly based on 3D seismic exploration. As an area exploration method, 3D seismic exploration can provide geological information such as the spatial distribution, fracture system, and lithology of coal seams, and has been widely used in coalfields.

[0004] However, existing technologies have the following problems: Currently, the verification mechanism for coalbed methane exploration is simplistic and relies excessively on core test data. This not only results in high sampling costs and limited coverage, making it difficult to achieve verification across the entire well section, but also fails to incorporate spatial verification using 3D seismic data. Consequently, it is prone to becoming disconnected from geological structural patterns, and the application of results lacks quantitative standards. It only outputs results for gas-bearing and non-gas-bearing layers, making it impossible to assess development value and leading to blind deployment of exploration wells. Summary of the Invention

[0005] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a method for identifying gas layers in coal seam exploration wells. This method has advantages such as multiple verification dimensions and quantitative evaluation results, and solves problems such as a single verification mechanism, detachment from geological background, and vague qualitative conclusions.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a method for identifying gas layers in coal seam exploration wells, comprising the following steps: S1. Data Acquisition: Acquire conventional logging data, core test data, and 3D seismic data of the target coal seam area from the undetermined exploration wells in the target coal seam area; S2. Data Preprocessing: Outlier removal, depth realignment, and missing value imputation are performed on well logging data, and standardization is carried out to form a unified standardized data volume; parameters related to gas layer identification are extracted from core experimental data, and the parameters are cleaned, outliers are removed, and missing data is imputed; noise reduction, filtering, and amplitude compensation are performed on the three-dimensional seismic data of the target coal seam area. Based on the coordinate information of the well location corresponding to the logging data, the well location coordinates are input using seismic data interpretation software to generate a three-dimensional seismic profile of the well location; a seismic record is synthesized by combining the logging data and the three-dimensional seismic data, and then the synthesized seismic record is superimposed on the seismic profile. On the aligned seismic profile, the positions of key strata such as coal seams are marked to complete the well-seismic calibration. S3. Preliminary gas layer identification: Based on the standardized logging data volume, using known gas-bearing core experimental data as the training set, a logging response model for coalbed methane layers is established, and the gas layer identification threshold values ​​for each parameter are determined. Through the logging response model and threshold values, logging data of the entire well section are calculated to identify depth segments that meet the characteristics of gas layers. Combining multi-well information, the depth segments and ranges that meet the conditions are found, potential gas layers are delineated, and the preliminary gas layer identification is completed. S4. Core Logging Verification: Based on the core sampling depth, within the depth range corresponding to the preliminary gas layer identification results, verification points are set up at a density of one core verification point every 2m. For each verification point, the gas saturation measured in the core laboratory is extracted, and the gas probability of the corresponding depth is read. The single-point deviation value is calculated, the number of verification points with a deviation ≤10% is counted, and the core logging matching degree is calculated. Well-seismic cross-verification: Using the well-seismic calibration relationship, the delineated potential gas layer range is superimposed on the three-dimensional seismic profile. The spatial overlap between the structural high position, fault zone and gas layer is compared. At the same time, GIS tools are used to calculate the overlap area and overlap degree between the gas layer range and the favorable structural zone. If it is higher than 70%, it is qualified. If it is lower than 70%, the gas layer boundary is further corrected by combining the seismic wave impedance anomaly characteristics. Verification result determination: When the core logging matching degree is ≥85% and the well-seismic spatial overlap degree is ≥70%, the gas layer identification result is confirmed to be valid. If the above conditions are not met, return to S3 for re-identification and screening until the verification result meets the standard. S5. Gas Content and Development Assessment: Using gas layer thickness, gas saturation, permeability, and gas content as input parameters, the weight of each parameter is determined by the analytic hierarchy process (AHP) to calculate the gas layer development value index.

[0007] Preferably, the well logging data in S1 includes resistivity, sonic transit time, density, neutron porosity, and natural gamma curve; the core experimental data includes gas saturation, permeability, and porosity; and the three-dimensional seismic data includes wave impedance, amplitude properties, and structural interpretation results data.

[0008] Preferably, in S2, outlier removal adopts a dual-verification mode of mathematical statistics and geological verification; in the missing interpolation, short missing segments adopt linear interpolation, medium and long missing segments adopt neighboring well analogy, and missing key coal seam segments adopt core calibration; the standardization process adopts regional Z-Score standardization; the gas layer identification parameters extracted from the core experimental data include porosity, permeability, gas saturation and coal content; and the data cleaning adopts repeated experimental verification and box plot outlier detection.

[0009] Preferably, in S2, the denoising of the three-dimensional seismic data adopts a step-by-step process of surface wave suppression and random noise attenuation. The filtering adopts bandpass filtering with a passband range of 15-85Hz to remove low-frequency interference and high-frequency noise. The amplitude compensation adopts a combination of spherical diffusion compensation and absorption attenuation compensation.

[0010] Preferably, in S3, the logging response model uses logging parameters at the corresponding depth of gas-bearing cores as training samples, establishes the model using a random forest algorithm, and outputs the gas-bearing probability prediction result. The gas layer identification threshold value adopts the method of statistical analysis of core gas saturation: the coal seam is divided into three intervals: shallow, medium, and deep. The critical values ​​of logging parameters corresponding to gas-bearing cores are statistically analyzed in each interval to form a partitioned threshold value system.

[0011] Preferably, the single-point deviation value in S4 is calculated using the following formula:

[0012] Where Log_prob is the gas-bearing probability at the depth corresponding to the verification point, and Core_Sg is the measured gas saturation in the core laboratory at the verification point. The absolute value sign is used to eliminate the positive and negative influences of the deviation, only reflecting the magnitude of the difference. The core logging matching degree is the percentage of verification points with a deviation ≤10% out of the total number of verification points, as shown in the following formula:

[0013] Where P is the core logging matching degree, M is the number of valid verification points with a deviation ≤10%, and N is the total number of valid verification points, which is the number after removing invalid points such as core breakage and logging data distortion.

[0014] Preferably, the core verification point layout in S4 should avoid the fracture zone: when there is a seismically interpreted fracture within the initially identified gas layer segment, the verification points should be densified within a 3-meter range on both sides of the fracture, with a density of 1 verification point per 1 meter after densification, to ensure the accuracy of the matching degree calculation of the fracture-affected area.

[0015] Preferably, the formula for the value index in S5 is:

[0016] Where V is the development value index, the higher the V value, the stronger the development value; W_H is the weight of gas layer thickness, where H is the gas layer thickness; W_Sg is the weight of gas saturation, where Sg is the gas saturation; W_K is the weight of permeability, where K is the permeability; and W_G is the weight of gas content, where G is the gas content. Before calculating the development value index V, the gas layer thickness, gas saturation, permeability, and gas content are normalized to make their value range 0.1-1.

[0017] Preferably, the hierarchical analysis method described in S5 involves constructing a target layer and a criterion layer to determine the parameter weights. The criterion layer includes four parameters: gas layer thickness, gas saturation, permeability, and gas content. The importance of the parameters is scored using the 1-9 scaling method, a judgment matrix is ​​constructed, and after passing a consistency check, the weights of each parameter are finally determined.

[0018] Preferably, in the well-seismic cross-validation in S4, the criteria for defining the favorable structural zone are as follows: High structural locations: the core areas of anticlines and domes in earthquake interpretation, with a structural amplitude ≥ 5m; Favorable fault zone: within 10-50m of an open normal fault or extensional-shear fault, where the reservoirs on both sides of the fault are well connected and there is no mudstone layer to block them. Anomaly zone of seismic impedance: The seismic wave impedance value is 10%-20% lower than the average wave impedance of the surrounding rock within ±5m of the target coal seam, and is consistent with the range of gas layer wave impedance calibrated by well logging. The favorable structural zone is the superposition range of the above-mentioned areas, and the continuous area formed after superposition is used as the reference area for calculating the spatial overlap of well-seismic space.

[0019] (III) Beneficial Effects Compared with the prior art, the present invention provides a method for identifying gas layers in coal seam exploration wells, which has the following beneficial effects: This method for identifying gas layers in coal seam exploration wells first establishes a correlation between well logging, core data, and 3D seismic data based on multi-source data acquisition and well-seismic calibration. Then, a three-dimensional cross-validation closed loop is formed through core logging point verification and well-seismic spatial verification. Core experimental data ensures the accuracy of local identification in single wells, while seismic data verifies the spatial compatibility of gas layers with structural highs and fault zones. Furthermore, wave impedance anomaly characteristics are combined to correct gas layer boundaries, overcoming the limitations of traditional single core verification and ensuring a high degree of consistency between identification results and geological structural patterns. Accurate parameters are obtained through the fusion of core measurements and well logging inversion. Then, the analytic hierarchy process (AHP) is used to assign weights that meet commercial mining needs. Through value index calculation, gas layers are divided into three levels: priority development, potential development, and temporarily non-development, outputting quantitative classification results. This provides a clear basis for investment decisions regarding well deployment and effectively avoids cost waste caused by blind exploration. Attached Figure Description

[0020] Figure 1 This invention presents a schematic diagram of the overall process for a method of identifying gas layers in coal seam exploration wells; Figure 2 This invention presents a schematic diagram of the data preprocessing process for a method of identifying gas layers in coal seam exploration wells. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 and Figure 2 This invention provides a technical solution: a method for identifying gas layers in coal seam exploration wells, comprising the following steps: S1. Data Acquisition: Acquire conventional logging data, core test data, and 3D seismic data of the target coal seam area from the undetermined exploration wells in the target coal seam area.

[0023] The logging data in S1 includes resistivity, sonic transit time, density, neutron porosity, and natural gamma curve; the core experimental data includes gas saturation, permeability, and porosity; and the 3D seismic data includes wave impedance, amplitude properties, and structural interpretation results.

[0024] In this step, well logging data can directly reflect the physical properties of coal seams, providing basic physical parameters for gas layer identification. Different curves reflect formation characteristics from different angles. For example, resistivity is sensitive to gas content, sonic transit time reflects the density of rocks, and core experimental data can provide direct physical property parameters of gas layers, serving as the standard true values ​​for verifying well logging interpretation and model construction. Among these, gas saturation is the core indicator for judging gas layers. 3D seismic data enables regional-scale structural and lithological prediction, providing a macroscopic background for well-seismic combined gas layer identification. Wave impedance can indirectly reflect changes in lithology and gas content.

[0025] S2. Data Preprocessing: Outlier removal, depth realignment, and missing value imputation are performed on well logging data, and standardization is carried out to form a unified standardized data volume; parameters related to gas layer identification are extracted from core experimental data, and the parameters are cleaned, outliers are removed, and missing data is imputed; noise reduction, filtering, and amplitude compensation are performed on the three-dimensional seismic data of the target coal seam area. Based on the coordinate information of the well location corresponding to the logging data, the well location coordinates are input using seismic data interpretation software to generate a three-dimensional seismic profile of the well location. Seismic records are synthesized by combining logging data and three-dimensional seismic data. Then, the synthesized seismic records are superimposed on the seismic profile. On the aligned seismic profile, the positions of key strata such as coal seams are marked to complete the well-seismic calibration.

[0026] In S2, outlier removal employs a dual-verification model combining mathematical statistics and geological validation. For missing data imputation, linear interpolation is used for short missing segments, neighboring well analogy is used for medium and long missing segments, and core calibration is used for missing key coal seam segments. Standardization is achieved using a regionalized Z-Score standardization method. Gas layer identification parameters extracted from core experimental data include porosity, permeability, gas saturation, and coal content. Data cleaning utilizes repeated experimental verification and box plot outlier detection. Noise reduction of 3D seismic data employs a step-by-step process of surface wave suppression and random noise attenuation. Bandpass filtering is used, with a passband range of 15-85Hz to remove low-frequency interference and high-frequency noise. Amplitude compensation uses a combination of spherical diffusion compensation and absorption attenuation compensation.

[0027] In this step, outlier removal employs a combination of mathematical statistics and geological verification. The mathematical statistics utilize the 3σ criterion and box plot method, while geological verification provides dual validation. This approach identifies data exceeding normal fluctuation ranges using mathematical methods and eliminates false anomalies by incorporating lithological characteristics, preventing outlier data from interfering with gas reservoir feature extraction. Furthermore, the imputation of missing values ​​in both well logging and core experimental data is identical, allowing for differentiated methods for missing segments of varying lengths. Short missing segments (≤5m) use linear interpolation to ensure data continuity, while medium-to-long missing segments (5-20m) employ adjacent well analogy combined with geological consistency interpolation. Key coal seam segments are imputed using core calibration to ensure interpolation accuracy, preventing missed or incorrect depth segment identification due to data gaps. A regionalized Z-Score standard is also used. The quantitative method can eliminate measurement system errors of different wellbores and instruments, such as the difference in resistivity benchmark values ​​between different wells, so that the logging data of each well are in the same quantitative standard. The four parameters extracted from the core experimental data, namely porosity, permeability, gas saturation and coal content, constitute the core of reservoir evaluation. Porosity and permeability characterize the reservoir capacity, gas saturation directly indicates gas content, and coal content accurately defines the target reservoir. These data will serve as labels for training logging models and verifying results. Data cleaning in the core experimental data adopts repeated experimental verification and box plot outlier detection. Repeated experiments ensure the accuracy of individual data points from the source, while box plots robustly identify discrete outlier experimental values ​​from the perspective of the whole dataset, which ensures that the standard itself is pure and reliable. Meanwhile, 3D seismic data can better protect effective reflection signals while removing noise, and the passband range is set to 15-85Hz. This range usually contains the frequency components with the highest signal-to-noise ratio in seismic data that best reflect reservoir characteristics, providing a high-resolution data foundation for subsequent fine identification of coal seams and gas layers. Spherical diffusion compensation in amplitude compensation can restore the energy attenuation caused by wavefront expansion, and absorption attenuation compensation can restore the high-frequency loss caused by the absorption of seismic waves by the strata. The combination of the two makes the amplitude energy of deep and shallow reflected waves comparable.

[0028] S3. Preliminary gas layer identification: Based on the standardized logging data volume, using known gas-bearing core experimental data as the training set, a logging response model for coal seam gas layers is established, and the gas layer identification threshold values ​​for each parameter are determined. Through the logging response model and threshold values, logging data of the entire well section are calculated to identify depth segments that meet the characteristics of gas layers. Combining multi-well information, the depth segments and ranges that meet the conditions are found, potential gas layers are delineated, and the preliminary gas layer identification is completed.

[0029] In S3, the logging response model uses logging parameters at the corresponding depth of gas-bearing cores as training samples, and adopts the random forest algorithm to build the model and output the gas-bearing probability prediction results. The gas layer identification threshold value adopts the method of statistical analysis of core gas saturation: the coal seam is divided into three intervals: shallow, medium and deep. The critical values ​​of logging parameters corresponding to gas-bearing cores are statistically analyzed in each interval to form a partition threshold value system.

[0030] In this step, the well logging response model employs the random forest algorithm to address complex nonlinear problems and provides probabilistic and interpretable prediction results. The threshold values ​​are determined using a zonal statistical approach based on core gas saturation: the coal seam is divided into three intervals—shallow, medium, and deep—and the critical values ​​of the logging parameters corresponding to gas-bearing cores are statistically analyzed for each interval, forming a zonal threshold value system. However, different coal seam depths result in significant differences in geological environment. Shallow coal seams may have lower compaction and weaker gas-bearing stability, leading to different critical values ​​for logging parameters compared to medium-deep coal seams. Medium-deep coal seams are influenced by formation pressure. The relationship between pore structure and gas saturation may change. By dividing the burial depth into three intervals (shallow, medium, and deep), and statistically analyzing the critical values ​​of logging parameters corresponding to gas-bearing cores in each interval, a zonal threshold value system is formed. This system avoids the problems of missed detection of shallow burial layers and misjudgment of medium and deep layers caused by uniform standards. It makes the identification standard more in line with the gas layer characteristics under different geological conditions, greatly improving the practicality and accuracy of the threshold value. By combining the logging response model with the zonal threshold value, logging data of the entire well section is calculated and judged segment by segment, achieving screening without omissions.

[0031] S4. Multi-dimensional cross-validation: Core logging verification: Based on the core sampling depth, within the depth range corresponding to the preliminary gas layer identification results, verification points are set up at a density of one core verification point every 2m. For each verification point, the gas saturation measured in the core laboratory is extracted, and the gas probability of the corresponding depth is read. The single-point deviation value is calculated, the number of verification points with a deviation ≤10% is counted, and the core logging matching degree is calculated. Well-seismic cross-verification: Using the well-seismic calibration relationship, the delineated potential gas layer range is superimposed on the three-dimensional seismic profile. The spatial overlap between the structural high position, fault zone and gas layer is compared. At the same time, GIS tools are used to calculate the overlap area and overlap degree between the gas layer range and the favorable structural zone. If it is higher than 70%, it is qualified. If it is lower than 70%, the gas layer boundary is further corrected by combining the seismic wave impedance anomaly characteristics. Verification result determination: When the core logging matching degree is ≥85% and the well-seismic spatial overlap is ≥70%, the gas layer identification result is confirmed to be valid. If the above conditions are not met, return to S3 for re-identification and screening until the verification result meets the standard.

[0032] In S4, the single-point deviation value is calculated using the following formula:

[0033] Where Log_prob is the gas-bearing probability at the depth corresponding to the verification point, Core_Sg is the measured gas saturation in the core laboratory at the verification point, and the absolute value sign is used to eliminate the positive and negative influence of the deviation, only reflecting the magnitude of the difference. The core logging matching degree is the percentage of verification points with a deviation ≤10% out of the total number of verification points, as shown in the following formula:

[0034] Where P is the core logging matching degree, M is the number of valid verification points with a deviation ≤10%, and N is the total number of valid verification points, which is the number after removing invalid points such as core breakage and logging data distortion. In the cross-validation of well and seismic data in S4, the criteria for defining favorable structural zones are as follows: High structural locations: the core areas of anticlines and domes in earthquake interpretation, with a structural amplitude ≥ 5m; Favorable fault zone: within 10-50m of an open normal fault or extensional-shear fault, where the reservoirs on both sides of the fault are well connected and there is no mudstone layer to block them. Anomaly zone of seismic impedance: The seismic wave impedance value is 10%-20% lower than the average wave impedance of the surrounding rock within ±5m of the target coal seam, and is consistent with the range of gas layer wave impedance calibrated by well logging. The favorable structural zone is the superposition range of the above-mentioned areas, and the continuous area formed after superposition is used as the reference area for calculating the spatial overlap of well-seismic space.

[0035] In this step, cross-verification between core logging and seismic structures corrects initial identification deviations, clarifies gas layer boundaries, and ensures that identification results conform to the patterns of single-well logging and align with regional geological structure characteristics. Simultaneously, verification points are deployed at a density of at least one point every 2 meters, with the density increased to one point per meter on both sides of the fault zone. This ensures verification points cover both conventional and structurally complex areas. The potential gas layer range is superimposed on the seismic profile, and GIS tools are used to calculate the overlap area with favorable structural zones. An overlap of ≥70% indicates that the gas layer distribution matches geological patterns. If it is below 70%, combined with seismic impedance anomalies, gas layers typically exhibit low impedance, correcting the gas layer boundaries and eliminating pseudo-gas layers exceeding favorable structural zones. However, gas layers that meet impedance anomalies are missed. A dual threshold of "core logging matching degree ≥85% and well-seismic overlap degree ≥70%" is set, forming a closed-loop control system where rework is required if the standards are not met. When the standards are met, the process returns to S3, re-optimizes characteristic parameters or judgment criteria, and re-identifies and verifies until both indicators meet the standards, avoiding invalid results due to insufficient point accuracy or unreasonable surface distribution.

[0036] S5. Gas Content and Development Assessment: Using gas layer thickness, gas saturation, permeability, and gas content as input parameters, the weight of each parameter is determined by the analytic hierarchy process (AHP) to calculate the gas layer development value index.

[0037] The formula for the development value index in S5 is:

[0038] Where V is the development value index, the higher the V value, the stronger the development value; W_H is the weight of gas layer thickness, where H is the gas layer thickness; W_Sg is the weight of gas saturation, where Sg is the gas saturation; W_K is the weight of permeability, where K is the permeability; and W_G is the weight of gas content, where G is the gas content. Before calculating the development value index V, the gas layer thickness, gas saturation, permeability, and gas content are normalized to make their value range 0.1-1. In S5, the analytic hierarchy process (AHP) constructs a target layer and a criterion layer to determine the parameter weights. The criterion layer includes four parameters: gas layer thickness, gas saturation, permeability, and gas content. The importance of the parameters is scored using the 1-9 scale, a judgment matrix is ​​constructed, and after passing the consistency test, the weights of each parameter are finally determined.

[0039] In this step, based on the confirmation of the gas layer's validity, the development value of the gas layer is quantified to provide data support for whether to develop it, the development sequence, and the selection of development technology. Simultaneously, H determines the gas layer's reserve scale, Sg and G directly reflect gas abundance, and K determines gas production capacity; all four together embody the development value. Using the analytic hierarchy process (AHP), the structure of the target layer and criterion layer is constructed. The importance of parameters is scored using a 1-9 scale, a judgment matrix is ​​constructed, and consistency checks are performed to ensure that the weight allocation conforms to engineering practice and avoids subjective assignment. The four parameters are first normalized to eliminate dimensional differences, and then the index is calculated using the formula V=W_H×H+W_Sg×Sg+W_K×K+W_G×G. A higher V value represents stronger development value and can be directly used for gas layer classification; for example, V≥0.8 is a primary development layer, and 0.6-0.8 is a secondary layer. Simultaneously, the development value index quantifies the gas layer's development potential, providing a scientific basis for exploration and development decisions and determining priority development areas.

[0040] Example: Taking a gently dipping coal seam area in a mining area as the research object, the coal seam dip angle in this area is 8°-15°, the structure is simple, there are no large faults, only a few small joints, and the stratigraphic continuity is good. The specific implementation process is as follows: S1. Data Acquisition: Conventional logging curves from wells 1, 2, and 3 in the area were acquired, including resistivity, sonic transit time, density, neutron porosity, and natural gamma. The curves showed good continuity, with no significant data interruptions in small joint sections. Core samples were taken from the No. 2 and No. 5 main coal seam sections of the three wells to obtain key parameters. The gas saturation ranged from 45% to 82%, the permeability from 0.3 to 4.8 mD, and the porosity from 6% to 14%. The coal content was recorded simultaneously. The core sampling rate was over 90%, and the data was reliable. Three-dimensional seismic data covering a 200 km² work area were acquired. After processing, wave impedance data, amplitude attributes, and structural interpretation results were obtained. The seismic data revealed that the area is a broad and gentle anticline structure with no large fault response.

[0041] S2, Data Preprocessing: First, quality control was performed on the well logging data. Outliers were removed using a combination of mathematical statistics and geological verification. For a short, missing 4-meter density curve in Well 2, linear interpolation was used for repair. For a medium-to-long missing 12-meter sonic curve in Well 3, analogous interpolation was performed by referencing the curve trend of the neighboring Well 1, which has a similar geological background. All well logging data were standardized using the regionalized Z-Score method to eliminate inter-well systematic errors. Core data underwent box plot testing and comparison with repeated experimental records to remove individual outliers, ensuring accuracy. The training set was kept pure, and the 3D seismic data underwent surface wave suppression, random noise attenuation, 15-85Hz bandpass filtering, and a combination of spherical diffusion and absorption attenuation compensation, which effectively improved the signal-to-noise ratio and fidelity. Subsequently, synthetic seismic records for each well were created using acoustic waves and density curves, and finely calibrated with the seismic traces near the wells. By finely adjusting the time-depth relationship, the synthetic records and seismic profiles were made to highly match the main wave group characteristics. The strong reflection phase axes of coal seams #2 and #5 were successfully and accurately marked on the seismic profiles, completing the well-seismic calibration.

[0042] S3. Preliminary gas layer identification: Using gas-bearing core samples and their corresponding standardized logging data as the training set, a logging response model was constructed using the random forest algorithm. The model output a continuous gas-bearing probability curve for the entire well section. Simultaneously, based on the coal seam burial depth (shallow <800m, middle 800-1200m, deep >1200m), a zoning threshold system was established. Applying this model and zoning thresholds, the entire well section of three wells was processed. Multiple depth segments matching the characteristics of gas-bearing layers were identified in the No. 2 and No. 5 coal seams. Combining multi-well information, a contiguous potential gas-bearing layer distribution area located in the core of the anticline was initially delineated on the plane.

[0043] S4. Multi-dimensional cross-validation: Core logging verification: Within the initially identified gas-bearing zone, 125 verification points were set up at a density of one point every 2 meters. The deviation between the measured gas saturation in the core and the probability of gas content in the logging was calculated for each point. It was found that there were 112 verification points with a deviation ≤10%. The core logging matching degree P = (112 / 125) × 100% = 89.6%, which is higher than the threshold of 85%.

[0044] Well-seismic cross-validation: The delineated potential gas-bearing layer area is overlaid onto 3D seismic data, and GIS tools are used to calculate its overlap area with favorable structural zones. In this example, the main areas are the high parts of anticlines and regions where the wave impedance value is lower than 15% of the average value of the surrounding rock. The calculated spatial overlap reaches 76%, which is higher than the 70% threshold.

[0045] Verification results: Since the core logging matching degree and well-seismic spatial overlap both meet the standards, the gas layer identification result of S3 is confirmed to be valid and no rework is required.

[0046] S5, Gas Content and Development Assessment: For the verified gas reservoir, its average thickness H=6.5m, average gas saturation Sg=65%, average permeability K=2.1mD, and calculated gas content G=12.5m³ / t were extracted. The weights of each parameter were determined using the analytic hierarchy process (AHP): gas content W_G=0.4, gas saturation W_Sg=0.3, permeability W_K=0.2, and reservoir thickness W_H=0.1. After normalizing each parameter, they were substituted into the development value index formula.

[0047] The average development value index V of the gas reservoir in this area was calculated to be 0.78. According to the classification standard, V≥0.8 is classified as Level 1 and 0.6-0.8 is classified as Level 2. This area is rated as a Level 2 potential area with high development value. It is recommended to prioritize the deployment of test well groups for pilot development tests.

[0048] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.

[0049] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for identifying gas layers in coal seam exploration wells, characterized in that, Includes the following steps: S1. Data Acquisition: Acquire conventional logging data, core test data, and 3D seismic data of the target coal seam area from the undetermined exploration wells in the target coal seam area; S2. Data Preprocessing: Outlier removal, depth realignment, and missing value imputation are performed on well logging data, and standardization is carried out to form a unified standardized data volume; parameters related to gas layer identification are extracted from core experimental data, and the parameters are cleaned, outliers are removed, and missing data is imputed; noise reduction, filtering, and amplitude compensation are performed on the three-dimensional seismic data of the target coal seam area. Based on the coordinate information of the well location corresponding to the logging data, the well location coordinates are input using seismic data interpretation software to generate a three-dimensional seismic profile of the well location; a seismic record is synthesized by combining the logging data and the three-dimensional seismic data, and then the synthesized seismic record is superimposed on the seismic profile. On the aligned seismic profile, the positions of key strata such as coal seams are marked to complete the well-seismic calibration. S3. Preliminary gas layer identification: Based on the standardized logging data volume, using known gas-bearing core experimental data as the training set, a logging response model for coalbed methane layers is established, and the gas layer identification threshold values ​​for each parameter are determined. Through the logging response model and threshold values, logging data of the entire well section are calculated to identify depth segments that meet the characteristics of gas layers. Combining multi-well information, the depth segments and ranges that meet the conditions are found, potential gas layers are delineated, and the preliminary gas layer identification is completed. S4. Multi-dimensional cross-validation: Core logging verification: Based on the core sampling depth, within the depth range corresponding to the preliminary gas layer identification results, verification points are set up at a density of one core verification point every 2m. For each verification point, the gas saturation measured in the core laboratory is extracted, and the gas probability of the corresponding depth is read. The single-point deviation value is calculated, the number of verification points with a deviation ≤10% is counted, and the core logging matching degree is calculated. Well-seismic cross-verification: Using the well-seismic calibration relationship, the delineated potential gas layer range is superimposed on the three-dimensional seismic profile. The spatial overlap between the structural high position, fault zone and gas layer is compared. At the same time, GIS tools are used to calculate the overlap area and overlap degree between the gas layer range and the favorable structural zone. If it is higher than 70%, it is qualified. If it is lower than 70%, the gas layer boundary is further corrected by combining the seismic wave impedance anomaly characteristics. Verification result determination: When the core logging matching degree is ≥85% and the well-seismic spatial overlap degree is ≥70%, the gas layer identification result is confirmed to be valid. If the above conditions are not met, return to S3 for re-identification and screening until the verification result meets the standard. S5. Gas Content and Development Assessment: Using gas layer thickness, gas saturation, permeability, and gas content as input parameters, the weight of each parameter is determined by the analytic hierarchy process (AHP) to calculate the gas layer development value index.

2. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: The well logging data mentioned in S1 includes resistivity, sonic transit time, density, neutron porosity, and natural gamma curve; the core experimental data includes gas saturation, permeability, and porosity; and the three-dimensional seismic data includes wave impedance, amplitude properties, and structural interpretation results.

3. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: In S2, outlier removal employs a dual-verification model combining mathematical statistics and geological validation. In the missing data imputation, short missing segments are imputed using linear interpolation, medium-to-long missing segments are imputed using neighboring well analogy, and missing key coal seam segments are imputed using core calibration. The standardization process employs a regionalized Z-Score standardization method. The gas layer identification parameters extracted from the core experimental data include porosity, permeability, gas saturation, and coal content. Data cleaning employs repeated experimental verification and box plot outlier detection.

4. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: In S2, the denoising of the three-dimensional seismic data adopts a step-by-step process of surface wave suppression and random noise attenuation. The filtering adopts bandpass filtering with a passband range of 15-85Hz to remove low-frequency interference and high-frequency noise. The amplitude compensation adopts a combination of spherical diffusion compensation and absorption attenuation compensation.

5. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: The logging response model described in S3 uses logging parameters at the corresponding depth of gas-bearing cores as training samples, and establishes the model using the random forest algorithm to output the gas-bearing probability prediction results. The gas layer identification threshold value adopts the method of statistical analysis of core gas saturation: the coal seam is divided into three intervals: shallow, medium and deep. The critical values ​​of logging parameters corresponding to gas-bearing cores are statistically analyzed in each interval to form a partitioned threshold value system.

6. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: The single-point deviation value mentioned in S4 is calculated using the following formula: Where Log_prob is the gas-bearing probability at the depth corresponding to the verification point, and Core_Sg is the measured gas saturation in the core laboratory at the verification point. The absolute value sign is used to eliminate the positive and negative influences of the deviation, only reflecting the magnitude of the difference. The core logging matching degree is the percentage of verification points with a deviation ≤10% out of the total number of verification points, as shown in the following formula: Where P is the core logging matching degree, M is the number of valid verification points with a deviation ≤10%, and N is the total number of valid verification points, which is the number after removing invalid points such as core breakage and logging data distortion.

7. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: In S4, the layout of core verification points should avoid fracture zones: when there are seismically interpreted fractures within the initially identified gas-bearing strata, verification points should be densified within a 3-meter range on both sides of the fracture, with a density of 1 verification point per 1 meter, to ensure the accuracy of the matching degree calculation for the fracture-affected zone.

8. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: The formula for the development value index in S5 is: Where V is the development value index, the higher the V value, the stronger the development value; W_H is the weight of gas layer thickness, where H is the gas layer thickness; W_Sg is the weight of gas saturation, where Sg is the gas saturation; W_K is the weight of permeability, where K is the permeability; and W_G is the weight of gas content, where G is the gas content. Before calculating the development value index V, the gas layer thickness, gas saturation, permeability, and gas content are normalized to make their value range 0.1-1.

9. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: The analytic hierarchy process described in S5 involves constructing a target layer and a criterion layer to determine the parameter weights. The criterion layer includes four parameters: gas layer thickness, gas saturation, permeability, and gas content. The importance of the parameters is scored using a 1-9 scale, a judgment matrix is ​​constructed, and after passing a consistency check, the weights of each parameter are finally determined.

10. The method for identifying gas layers in coal seam exploration wells according to claim 1, characterized in that: In the cross-validation of well and seismic data in S4, the criteria for defining the favorable structural zone are as follows: High structural locations: the core areas of anticlines and domes in seismic interpretation, with a structural amplitude ≥ 5m; Favorable fault zone: within 10-50m of an open normal fault or extensional-shear fault, where the reservoirs on both sides of the fault are well connected and there is no mudstone layer to block them. Anomaly zone of seismic impedance: The seismic wave impedance value is 10%-20% lower than the average wave impedance of the surrounding rock within ±5m of the target coal seam, and is consistent with the range of gas layer wave impedance calibrated by well logging. The favorable structural zone is the superposition range of the above-mentioned areas, and the continuous area formed after superposition is used as the reference area for calculating the spatial overlap of well-seismic space.

Images