Deep coal rock gas horizontal well drilling rapid steering method based on real-time logging data

Abstract

This invention relates to the field of oil and gas drilling engineering technology, specifically to a rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data, comprising: real-time acquisition of multi-source logging data; establishment of the functional relationship between idle friction torque and well depth; real-time calculation of effective rock-breaking torque of the drill bit; and MSE (Mean Separation of Forms). b Model building and real-time calculation; establishment and dynamic correction of lithology identification threshold: using logging and well logging data from adjacent wells, a statistical classification model for lithology identification is established; real-time guidance decision-making and trajectory adjustment: this application realizes real-time lithology identification, formation early warning and working condition judgment, thereby guiding precise adjustment of wellbore trajectory. This solution does not require additional equipment, is adaptable to complex drilling conditions, completely solves the data lag problem, significantly improves reservoir encounter rate and drilling efficiency, reduces operating costs and risks, and provides reliable technical support for the efficient development of deep coal and gas formations.

Description

Technical Field

[0001] This invention relates to the field of oil and gas drilling engineering technology, specifically to a rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data. Background Technology

[0002] Deep coalbed methane is an important unconventional natural gas resource with huge development potential. However, deep coal seams are usually characterized by high ground stress, high ground temperature and low permeability, which poses challenges to efficient development. Horizontal well technology has become a key means to achieve efficient development of this type of resource by significantly increasing the contact area between the wellbore and the coal seam. In this process, improving the drilling length and drilling rate of the reservoir target through geological steering technology is the core link to ensure high production of deep coalbed methane wells.

[0003] Currently, geological steering in horizontal wells mainly relies on logging while drilling (LWD) technology. This technology determines the position of the drill bit in the formation by real-time monitoring parameters such as gamma and resistivity. It has been successfully applied in drilling operations with horizontal sections of up to 3,000 meters in deep coal and gas formations. However, the existing LWD technology has a fundamental limitation: its sensors are usually installed in the drill string assembly at a distance of about 10 meters or even further from the drill bit. This physical distance leads to a significant lag in formation evaluation data. When the LWD sensor detects that the trajectory deviates from the reservoir, the drill bit may have already left the formation a certain distance, resulting in untimely determination of trajectory departure and seriously affecting drilling efficiency.

[0004] The consequences of this data lag are severe. Due to the delay in correction instructions, not only will the reservoir encounter rate decrease, but the difficulty of trajectory reversal will also increase, and it may even trigger complex sidetracking operations, ultimately negatively impacting drilling efficiency and wellbore quality. Therefore, there is an urgent need in this field for a method that can achieve zero-lag or near-real-time formation interface identification to overcome the inherent defects of LWD technology, provide more timely and accurate decision support for horizontal well drilling guidance, and thus significantly improve the drilling efficiency and development effect of deep coal and gas horizontal wells. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data, which solves the problem that the existing technology cannot meet the requirements for real-time and accurate directional drilling of deep coal and gas horizontal wells.

[0006] To solve the above problems, the technical solution of the present invention is as follows: a rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data, comprising the following steps: S1. During the drilling process, multi-source logging data, including drilling pressure, top drive speed, total surface torque, mechanical drilling speed and displacement, are collected in real time. S2. Using conventional short trip drilling operations, the drill bit is lifted off the bottom of the well. While maintaining normal circulation flow and rotation, the idling friction torque of the drill string is measured. Through multiple measurements, the functional relationship between the idling friction torque and the well depth is established. S3. During normal drilling, determine the idling friction torque at the current well depth based on the functional relationship obtained in S2, and remove the idling friction torque from the total torque on the ground to obtain the effective rock-breaking torque of the drill bit in real time. S4. Based on the effective rock-breaking torque, drilling pressure, mechanical drilling speed, top drive speed, and downhole power drill bit output speed, construct the drill bit rock-breaking specific energy (MSE). b Model, and calculate real-time MSE b value; S5. Establishment and dynamic correction of lithology identification threshold: Utilize logging and well logging data from adjacent wells to establish a statistical classification model for lithology identification; S6, when real-time MSE b When the value exceeds the threshold range corresponding to the target reservoir, it is determined that the drill bit has drilled out of the reservoir and an early warning is issued. Simultaneously, based on the MSE... b The changing trend guides the adjustment of the wellbore trajectory.

[0007] Furthermore, in S2, the functional relationship between the idling friction torque and the well depth is fitted using a quadratic exponential model, expressed as follows: ,in, Let be the idling friction torque of the drill string, a, b, and c be the fitting coefficients, and D be the well depth.

[0008] Furthermore, in S3, the effective rock-breaking torque of the drill bit ,in, This represents the total torque on the ground.

[0009] Furthermore, in S4, the MSE b The model's expression is , Among them, MSE b This represents the actual rock-breaking energy utilization rate of the drill bit. For drilling pressure, The cross-sectional area of ​​the wellbore. Rotary / top drive speed, The output speed of the downhole power drilling tool. For mechanical drilling speed, This represents the actual rotational speed of the drill bit.

[0010] Furthermore, S5 includes: S51. Based on the logging data of adjacent wells, the rock strength profile is obtained by inversion and used as a known lithology label; S52, Using MSE bThe model recalculates the logging data from adjacent wells to generate an MSE aligned with the lithology label depth. b curve; S53, Based on the aforementioned lithological tags and MSE b The correspondence of the curves, setting the initial MSE. b Lithology identification threshold; S54, Initial MSE b The lithology identification threshold is applied to real-time guidance of new wells for dynamic correction.

[0011] Furthermore, S53 includes: S531. Based on the lithology labels, determine the MSE corresponding to each type of lithology. b The probability density distributions of the datasets were fitted using Gaussian mixture models to quantify MSE under different lithologies. b The probabilistic characteristics of the value; S532. Based on the Bayesian discrimination criterion, calculate the optimal decision boundary between the probability distributions of adjacent lithology categories, and set the optimal decision boundary as the initial lithology identification threshold.

[0012] Furthermore, S531 includes: the expression for the Gaussian mixture model is: , Where x is MSE b The value, K, is the number of Gaussian components. The weight of the k-th component is... The mean is variance is The Gaussian probability density function.

[0013] Furthermore, S532 includes the following: the expression for the Bayesian discriminant criterion is: , in, It is the posterior probability, that is, the probability of observing MSE. b After value x, the lithology is The probability, It is a priori probability, that is, lithology obtained from data from adjacent wells. The prevalence of It represents the total probability and is a normalized constant.

[0014] Furthermore, in S54, the dynamic correction method is as follows: when the accuracy of the initially set threshold in the new well application reaches the preset standard, the threshold is fixed; otherwise, it is updated incrementally based on the measured data of the new well. and The threshold is iteratively optimized until the accuracy requirement is met.

[0015] Furthermore, in S6, the wellbore trajectory adjustment strategy is as follows: When the bottom is determined to be the source, the guidance tool is instructed to increase the well inclination angle to make the trajectory dip upwards and return to the reservoir; When it is determined that the well has been pushed out, the command guide tool reduces the well inclination angle to make the trajectory dip downwards and return to the reservoir; Adjustment range based on MSE b The value of the deviation exceeding the threshold and the distance between the current well inclination angle and the reservoir boundary are combined to determine the value. Compared with the prior art, the present invention has the following beneficial effects: by constructing an improved drill bit mechanical energy (MSE) b The model uses real-time ground logging data, such as drilling pressure, torque, and rotational speed, to directly calculate the rock-breaking energy at the drill bit, thus bypassing the data lag of more than 10 meters caused by the installation position of the logging-while-drilling (LWD) sensor. This allows the system to issue an early warning based on the jump in MSEb value the moment the drill bit emerges from the reservoir, achieving zero-hysteresis or near-real-time identification of the formation interface.

[0016] Fitting MSE using Gaussian mixture model b The probability distribution of values ​​is analyzed, and the optimal classification threshold is determined based on the Bayesian discrimination criterion. With the goal of minimizing the probability of classification errors, the threshold for distinguishing different lithologies, such as coal seams, mudstone, and limestone, is scientifically set, which greatly improves the accuracy of layer detection and lithology identification. Combined with the trajectory adjustment strategy clearly defined in S6, it can guide the drill bit to return to the target reservoir quickly and accurately, thereby effectively improving the reservoir encounter rate and drilling length throughout the well section.

[0017] Intelligent guidance upgrades can be achieved solely based on conventional logging data, without the need for new specialized instruments or equipment or changes to existing drilling processes. This avoids additional equipment procurement, installation, and maintenance costs. By measuring the idle friction torque during conventional short trips in the drilling process, no additional work procedures are required, simplifying the operation and improving drilling efficiency. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the endoscope main unit of the present invention; Figure 2 This is a schematic diagram of the fitting function image of the relationship between the drill string idling torque and well depth in reference well B of this invention; Figure 3 The MSE of the model of this invention and the Teale model in each well section b A comparison chart of values; Figure 4 This is a schematic diagram of the predicted profile of well A; Figure 5 This is a fitted curve representing the relationship between Tp and well depth D in the target well A of this invention. Detailed Implementation

[0019] This invention provides a rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data. Its core lies in the innovative data processing method that transforms conventional surface logging data into a drill bit mechanical energy (MSEb) index that can reflect the rock breaking state of the drill bit at the bottom of the well and the changes in formation lithology in real time and accurately. This overcomes the formation identification delay problem caused by the lag in sensor installation position in traditional logging while drilling (LWD) technology, and realizes real-time auxiliary guidance with zero lag for horizontal well trajectory.

[0020] The flowchart of the method of this invention is as follows: Figure 1 As shown.

[0021] A rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data includes the following steps: S1. During the drilling process, multi-source logging data, including drilling pressure, top drive speed, total surface torque, mechanical drilling speed and displacement, are collected in real time. S2. Using conventional short trip drilling operations, the drill bit is lifted off the bottom of the well. While maintaining normal circulation flow and rotation, the idling friction torque of the drill string is measured. Through multiple measurements, the functional relationship between the idling friction torque and the well depth is established. S3. During normal drilling, determine the idling friction torque at the current well depth based on the functional relationship obtained in S2, and remove the idling friction torque from the total torque on the ground to obtain the effective rock-breaking torque of the drill bit in real time. S4. Based on the effective rock-breaking torque, drilling pressure, mechanical drilling speed, top drive speed, and downhole power drill bit output speed, construct the drill bit rock-breaking specific energy (MSE). b Model, and calculate real-time MSE b value; S5. Establishment and dynamic correction of lithology identification threshold: Utilize logging and well logging data from adjacent wells to establish a statistical classification model for lithology identification; S6, when real-time MSE b When the value exceeds the threshold range corresponding to the target reservoir, it is determined that the drill bit has drilled out of the reservoir and an early warning is issued. Simultaneously, based on the MSE... b The changing trend guides the adjustment of the wellbore trajectory.

[0022] Specifically, taking Well A as the target well, which is a deep coal-gas horizontal well, the target reservoir is a coal seam (soft rock), and the top and bottom plates are limestone (hard rock). The designed length of the horizontal section is 1500m. During drilling, it is necessary to avoid the drill bit penetrating the reservoir and entering the hard rock of the top and bottom plates. Well B, with similar geological conditions and complete logging data, is selected as a reference well. A lithology identification threshold is established using data from Well B, and then applied to the real-time drilling guidance of Well A. The rapid drilling guidance method for deep coal-gas horizontal wells based on real-time logging data includes the following steps: S1: Real-time acquisition of multi-source logging data Based on the existing logging-while-drilling system, core drilling parameters were collected in real time during the drilling of Well A. The collection frequency was consistent with the conventional sampling frequency of the logging system, at 0.25 meters per point. The collected parameters included: Drilling pressure (W): Range 50-80kN; Top drive speed (N): Range 90-120 RPM; Total ground torque (Ta): Range 20-30kN m; Mechanical drilling rate (v): range 25-35 m / h; Displacement (Q): Range 2000-2100L / min.

[0023] The logging system automatically records and stores the above parameters, forming a continuous real-time data sequence, which provides basic data support for subsequent calculations. Table 1 shows some logging data of Well A during the normal drilling phase on July 5, 2025.

[0024] Table 1. Some drilling parameters automatically recorded by the logging system during normal drilling of Well A

[0025] S2: Establishment of the functional relationship between idling friction torque and well depth Utilizing conventional short trip-in / out operations during drilling, without adding extra procedures or affecting normal drilling efficiency, and adapting to existing drilling practices, the drill string idling friction torque Tp is measured at different well depths. Operation: When drilling to a depth of 4732.78m in Well A, a short trip operation was performed, raising the drill string by 5m to completely remove the drill bit from the bottom of the well. The drill bit position was 4724.93m. The top drive speed was kept constant at 111RPM and the displacement was kept constant at 2087L / min. The drill pipe was idled to clean the wellbore. Data acquisition: The logging system records torque data during idling, which is the idling friction torque Tp at the current well depth. Table 2 shows some logging data examples for this period, as follows; Function Fitting: The above operation was repeated at different well depths (4500m, 4600m, 4700m, 4732.78m) in target well A to obtain multiple sets of data pairs between well depth D and corresponding idling friction torque Tp. A quadratic exponential model was used to fit the data, and the model expression is as follows: Where a, b, and c are the fitting coefficients, obtained by the maximum likelihood estimation method: a = 7.22146, b = -0.0061, c = 1.82895 × 10⁻⁶. -6 The model has an adjusted coefficient of determination (Adj.R-Square) of 0.98654, indicating extremely high fitting accuracy. It can accurately characterize the relationship between Tp and well depth D. The fitting curve is shown in the figure below. Figure 5 As shown.

[0026] Taking reference well B as an example, a functional relationship between idling friction torque and well depth was established using its short tripping data. The fitted curve is shown below. Figure 2 As shown.

[0027] Table 2. Partial drilling parameters automatically recorded by the logging system during normal rotating circulation and rock clearing when the drill bit of Well A is lifted from the bottom of the well.

[0028] S3: Real-time calculation of effective rock-breaking torque of the drill bit Determine the current well depth Tp: When well A is drilled to a depth of 4738.3m, substitute the well depth D=4738.3m into the quadratic exponential model established in step S2 to calculate the idling friction torque Tp=17.65kN at the current well depth. m; Calculate the effective rock-breaking torque: according to the formula Where Ta is the average total surface torque at the current well depth collected in step S1, calculated from the data in Table 1 as Ta = 25.03 kN. Therefore, the effective rock-breaking torque Te = 25.03 - 17.65 = 7.38 kN. m; Continuous calculation throughout the well section: During the subsequent drilling of Well A, the current well depth D is acquired in real time, and the corresponding Tp is calculated by fitting the model. Combined with the real-time acquired Ta, the effective rock-breaking torque Te is continuously output, realizing the real-time and continuous calculation of Te throughout the well section.

[0029] By employing a simple calculation logic of total surface torque minus idling friction torque, the effective rock-breaking torque is directly separated. The principle is clear, the calculation is efficient, and continuous real-time output across the entire well section is possible. This solves the industry pain point of not being able to directly obtain drill bit torque in deep coal and gas horizontal wells. The accurate acquisition of the effective rock-breaking torque eliminates interference from drill string friction on the rock-breaking signal, ensuring subsequent MSE (Mean Separation Emissions). b The model calculations focus solely on drill bit rock-breaking energy consumption, significantly improving the accuracy of rock strength characterization; the real-time calculation mode ensures that rock-breaking torque data is synchronized with the drilling process, providing MSE (Mechanical, Electrical, and Sequence) data. b The real-time updates of the values ​​provide crucial input, ensuring the timeliness of subsequent lithology identification and early warning of stratigraphic emergence.

[0030] S4:MSE b Model building and real-time computing Model parameters determined: Wellbore cross-sectional area Ab: The wellbore diameter of well A is 215.9 mm, therefore Ab = π × (0.2159 / 2)² ≈ 0.0366 m²; Output speed of downhole power drill bit Nm: According to the displacement-speed relationship provided by the power drill bit manufacturer, when the displacement Q=2087L / min, Nm=120RPM; Actual drill bit rotation speed: N + Nm = 111 + 120 = 231 RPM; MSE b Value calculation: Using the above parameters, W=64.68kN, Ab=0.0366m², N+Nm=231RPM, Te=7.38kN Substituting m and v = 31.2 m / h into MSE b The model, the model expression is: Substituting the numerical values, we get: MSE b =64.68 / 0.0366+[120×π×231×7.38] / (0.0366×31.2)≈1767.21+556.89≈2324.1MPa. The result here is for illustrative purposes only. The actual result needs to be dynamically calculated based on real-time parameters. Real-time output: During the drilling process of Well A, the MSE is continuously calculated based on the parameters acquired in real time and the calculated Te. b Values, generating continuous MSEs b The curve directly reflects the change in rock strength at the bottom of the well.

[0031] like Figure 3 As shown, this is the MSE of the present invention. b Comparison of MSE values ​​between the model and the traditional Teale model in different well sections: In sections with low well inclination, the trends of both are similar due to less friction interference; however, in sliding drilling sections, the Teale model loses its formation detection capability due to top drive stall, resulting in lower MSE values. b The value is basically 0, but this model, by introducing the downhole power drill bit speed Nm, can still output a fluctuation curve that is highly correlated with the lithological profile. In the horizontal section, the Teale model's MSE value mainly reflects the drill string friction energy consumption, while this model, after removing friction interference, can accurately characterize the rock strength and rock breaking conditions, further verifying the reliability and adaptability of this model to complex working conditions.

[0032] The model incorporates the output rotational speed of the downhole power drill string, correcting for the actual drill bit rotational speed. This solves the problem of rotational speed failure in the sliding drilling section of the traditional Teale model, achieving MSE under all operating conditions. b Continuous value calculation adapts to complex drilling scenarios in deep coal and gas horizontal wells; it normalizes multiple parameters such as drilling pressure and rotation speed into a single energy index, simplifying the rock strength identification process and enabling technicians to use intuitive MSE (Metal Strength Index) data. b Value changes allow for the perception of stratigraphic lithological differences, reducing decision-making complexity; real-time calculation and output functions enhance MSE's capabilities. bThe curve can dynamically reflect the changes in rock strength at the bottom of the well. Compared with traditional lagging logging data, it realizes real-time feedback of formation characteristics and provides core decision-making basis for rapid guidance.

[0033] S5: Establishment and Dynamic Correction of Lithological Identification Thresholds S51: Inversion of rock strength profile from adjacent well to determine lithological labels. Using conventional logging data from adjacent well B (natural gamma, resistivity, sonic transit time, and density), a multivariate linear regression rock mechanics model is employed to invert the rock compressive strength profile of the entire well section. Combined with the core analysis results from well B and the comprehensive logging interpretation report, the lithological labels corresponding to each meter of well depth are determined: coal seam compressive strength 10-30 MPa, mudstone compressive strength 50-150 MPa, and limestone compressive strength 500-800 MPa, forming a three-dimensional dataset of well depth-lithological label-rock strength for well B.

[0034] S52: B Well MSE b Curve generation and depth alignment Collect logging data from the adjacent well (Well B), including drilling pressure, top drive speed, total surface torque, mechanical drilling speed, and displacement. Using steps S1-S4, calculate the MSE (Mean Segregation Estimate) for the entire section of Well B. b Value, generate B well MSE b The curve, through the depth correction parameters of well B, is precisely aligned with the lithology labels obtained from S51 according to depth, ensuring that each lithology label corresponds to a unique MSE at its depth. b value.

[0035] S53: Setting the initial lithology identification threshold S531: Gaussian mixture model fitting: MSE of Well B by lithology label b The datasets were divided into three categories: coal seam, mudstone, and limestone. For each category, a Gaussian mixture model was used to fit the probability density distribution. Gaussian mixture model fitting can accurately quantify the MSE of different lithologies. b The probability distribution characteristics of the values ​​are more scientific and statistically significant than traditional empirical threshold settings, reducing subjective errors. The model expression is as follows: , Where x is MSE b The value, Lj represents the j-th type of lithology, j=1,2,3 correspond to coal seam, mudstone, and limestone respectively, and K=2 represents the number of Gaussian components. The weight of the k-th component is... The mean, To find the variance, the parameters are obtained by solving the maximum likelihood estimation method: Coal seam (L1): =0.65, =180MPa =25MPa²; =0.35, =220MPa =30MPa²; Mudstone (L2): =0.70, =320MPa, =40MPa²; =0.30, =380MPa =45MPa²; Limestone (L3): =0.60, =650MPa, =80MPa²; =0.40, =720MPa, =90MPa²; S532: Bayesian discriminant criterion for threshold determination: Based on the fitted probability density distribution, according to the Bayesian discriminant criterion... Calculate the optimal decision boundary for adjacent lithologies, where the prior probability is... Based on the lithological thickness ratio of well B, P(L1) = 0.45, P(L2) = 0.30, P(L3) = 0.25, and the following calculations are obtained: The optimal decision boundary for coal seams and mudstone, with a threshold of 1: 280 MPa; The optimal decision boundary for mudstone and limestone, with a threshold of 2: 480 MPa; Therefore, the initial lithology identification threshold system is: MSE b <280MPa indicates a coal seam (target reservoir); 280MPa≤MSEb<480MPa indicates mudstone; and MSEb≥480MPa indicates limestone.

[0036] The optimal decision boundary determined by the Bayesian discrimination criterion aims to minimize the probability of classification errors, which significantly improves the accuracy of lithology differentiation and provides a reliable standard for lithology identification in Well A.

[0037] S54: Threshold Dynamic Correction The initial threshold is applied to the real-time guidance of well A, and the threshold is optimized through a prediction-verification-correction mechanism: Verification process: When well A was drilled to a depth of 4800m, the real-time MSE was... b The value remained at 220 MPa, indicating a coal seam. Subsequent logging data confirmed that the current location was indeed a coal seam, proving the threshold was valid. Correction process: When drilling reached a depth of 4850m, the real-time MSE was...b The value suddenly rose to 300 MPa, initially identified as mudstone based on the threshold. However, subsequent core sampling revealed a coal seam containing argillaceous interlayers. Therefore, the Gaussian mixture model parameters were incrementally updated based on this measured data. Adjusted to 185MPa The pressure was adjusted to 28 MPa², and the optimal decision boundary for the coal seam and mudstone was recalculated and corrected to 300 MPa to ensure that the threshold was adapted to the geological characteristics of well A.

[0038] S6: Real-time Guidance Decision Making and Trajectory Adjustment Lithology identification and stratigraphic early warning: When well A is drilled to a depth of 4900m, real-time MSE b The value suddenly increased from 230MPa to 550MPa and remained outside the target reservoir threshold range (<300MPa) for 5 minutes. Combined with the well inclination angle data, the current well inclination angle is 89°, which is determined to be bottom-out. The drill bit has drilled out of the reservoir bottom plate and entered the limestone. The system immediately issued a formation emergence warning. Operating condition identification: Simultaneously monitoring the idling friction torque Tp calculated in step S2, it was found that Tp remained stable at 18.2 MPa. Combined with MSE... b The rise was determined to be due to changes in formation lithology, rather than poor wellbore cleaning or pressure buildup; Trajectory adjustment: Based on the MSEb exceeding the threshold by 550-300=250MPa, the estimated distance (approximately 2m) between the current well inclination angle of 89° and the reservoir boundary, and engineering experience, the rotary steering tool is instructed to increase the well inclination angle by 1°, and the adjusted well inclination angle is 90°, so that the trajectory tilts upward and returns to the reservoir. Adjustment and verification: Continuously monitor MSE after adjustment. b Value, MSE in 10 minutes b The value dropped to 260 MPa, returning to the reservoir threshold range, verifying the effectiveness of the trajectory adjustment, and the drill bit successfully returned to the coal seam.

[0039] like Figure 4 The diagram shows a trajectory prediction profile for well B based on the method of this invention: when the drill bit reaches the bottom, the MSE... b The value suddenly increased from a typical low value (<200MPa) in the reservoir to 800MPa, and after the guide trajectory dipped upward by 1°, the MSE b The value drops back to 500 MPa, separating from the hard rock floor; when jacking occurs, MSE b The value also shows an upward trend. By adjusting the downward dip by 3.6°, it can be returned to the reservoir, which intuitively presents the complete guiding process of early warning-judgment-adjustment of the present invention, and further verifies the practicality of the technical solution.

[0040] By applying the method of this invention to well A, the following effects were achieved: The lag time for out-of-strata identification is ≤1 minute, which completely solves the problem of lag in stratigraphic evaluation data compared to the physical lag of about 10 meters in traditional LWD technology. The reservoir encounter rate reached 96.8%, which is 8.3 percentage points higher than that of Well B (which used traditional LWD directional drilling and had a reservoir encounter rate of 88.5%). The drilling cycle in the horizontal section was shortened by 12 days, the average mechanical drilling rate was increased by 32%, and no sidetracking operations were required due to formation failure, which significantly reduced drilling costs.

[0041] This invention addresses the pain points of lagging LWD (Low Driving) technology data and inaccurate traditional models in the drilling guidance of deep coal and gas horizontal wells. It proposes a rapid guidance scheme based on real-time logging data: by collecting multi-source drilling parameters, separating drill string friction torque from effective rock-breaking torque, and combining this with the downhole power drill bit rotation speed, a Mean Squared Equation (MSE) is constructed. b The model, based on logging and well logging data from Well B, establishes lithology identification thresholds through Gaussian mixture model fitting and Bayesian discrimination criteria, and dynamically optimizes them during new well drilling. Based on the comparison between real-time MSEb values ​​and the thresholds, it achieves real-time lithology identification, formation warning, and working condition assessment, thereby guiding precise adjustment of the wellbore trajectory. This solution requires no additional equipment, is adaptable to complex drilling conditions, completely solves the data lag problem, significantly improves reservoir encounter rate and drilling efficiency, reduces operating costs and risks, and provides reliable technical support for the efficient development of deep coal and gas formations.

[0042] The above specific embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A rapid directional drilling method for deep coal and gas horizontal wells based on real-time logging data, characterized in that, Includes the following steps: S1. During the drilling process, multi-source logging data, including drilling pressure, top drive speed, total surface torque, mechanical drilling speed and displacement, are collected in real time. S2. Using conventional short trip drilling operations, the drill bit is lifted off the bottom of the well. While maintaining normal circulation flow and rotation, the idling friction torque of the drill string is measured. Through multiple measurements, the functional relationship between the idling friction torque and the well depth is established. S3. During normal drilling, determine the idling friction torque at the current well depth based on the functional relationship obtained in S2, and remove the idling friction torque from the total torque on the ground to obtain the effective rock-breaking torque of the drill bit in real time. S4. Based on the effective rock-breaking torque, drilling pressure, mechanical drilling speed, top drive speed, and downhole power drill bit output speed, construct the drill bit rock-breaking specific energy (MSE). b Model, and calculate real-time MSE b value; S5. Establishment and dynamic correction of lithology identification threshold: Utilize logging and well logging data from adjacent wells to establish a statistical classification model for lithology identification; S6, when real-time MSE b When the value exceeds the threshold range corresponding to the target reservoir, it is determined that the drill bit has drilled out of the reservoir and an early warning is issued. Simultaneously, based on the MSE... b The changing trend guides the adjustment of the wellbore trajectory.

2. The drilling rapid steering method according to claim 1, characterized in that: In S2, the functional relationship between the idling friction torque and the well depth is fitted using a quadratic exponential model, expressed as follows: ,in, Let be the idling friction torque of the drill string, a, b, and c be the fitting coefficients, and D be the well depth.

3. The drilling rapid steering method according to claim 2, characterized in that: In S3, the effective rock-breaking torque of the drill bit ,in, This represents the total torque on the ground.

4. The drilling rapid steering method according to claim 3, characterized in that: In S4, the MSE b The model's expression is , Among them, MSE b This represents the actual rock-breaking energy utilization rate of the drill bit. For drilling pressure, The cross-sectional area of ​​the wellbore. This refers to the rotational speed of the turntable / top drive. The output speed of the downhole power drilling tool. For mechanical drilling speed, This represents the actual rotational speed of the drill bit.

5. The drilling rapid guidance method according to claim 4, characterized in that, S5 include: S51. Based on the logging data of adjacent wells, the rock strength profile is obtained by inversion and used as a known lithology label; S52, Using MSE b The model recalculates the logging data from adjacent wells to generate an MSE aligned with the lithology label depth. b curve; S53, Based on the aforementioned lithological tags and MSE b The correspondence of the curves, setting the initial MSE. b Lithology identification threshold; S54, Initial MSE b The lithology identification threshold is applied to real-time guidance of new wells for dynamic correction.

6. The drilling rapid guidance method according to claim 5, characterized in that, S53 includes: S531. Based on the lithology labels, determine the MSE corresponding to each type of lithology. b The probability density distributions of the datasets were fitted using Gaussian mixture models to quantify MSE under different lithologies. b The probabilistic characteristics of the value; S532. Based on the Bayesian discrimination criterion, calculate the optimal decision boundary between the probability distributions of adjacent lithology categories, and set the optimal decision boundary as the initial lithology identification threshold.

7. The drilling rapid guidance method according to claim 6, characterized in that, S531 includes: The expression for the Gaussian mixture model is: , Where x is MSE b The value, K, is the number of Gaussian components. The weight of the k-th component is... The mean is variance is The Gaussian probability density function.

8. The drilling rapid steering method according to claim 7, characterized in that: S532 includes the following: The expression for the Bayesian discriminant criterion is: , in, It is the posterior probability, that is, the probability of observing MSE. b After value x, the lithology is The probability, It is a priori probability, that is, lithology obtained from data from adjacent wells. The prevalence of It represents the total probability and is a normalized constant.

9. The drilling rapid steering method according to claim 8, characterized in that: In S54, the dynamic correction method is as follows: when the accuracy of the initially set threshold in the new well application reaches the preset standard, the threshold is fixed; otherwise, it is updated incrementally based on the measured data of the new well. and The threshold is iteratively optimized until the accuracy requirement is met.

10. The drilling rapid steering method according to claim 9, characterized in that: In S6, the wellbore trajectory adjustment strategy is as follows: When the bottom is determined to be the source, the guidance tool is instructed to increase the well inclination angle to make the trajectory dip upwards and return to the reservoir; When it is determined that the well has been pushed out, the command guide tool reduces the well inclination angle to make the trajectory dip downwards and return to the reservoir; Adjustment range based on MSE b The value is determined by a combination of factors, including the magnitude of the exceedance threshold, the current well inclination angle, and the distance to the reservoir boundary.

Images