A semi-automatic real-time tracking method for formation interfaces in logging resistivity while drilling

Abstract

This invention discloses a real-time semi-automatic method for tracking formation interfaces in drilling resistivity logging, comprising: extracting formation information from logging data of adjacent wells / pilot wells; establishing an initial interpretation model to determine whether the formation is in a build-up section; determining the location of the formation interface in the build-up section; enabling an automatic interface tracking mode for horizontal well sections to quickly invert the upper and lower interface positions of the target layer; determining whether the error between simulated data and measured data exceeds a threshold; enabling an expert mode to identify the formation interface and manually adjusting the formation interface; and obtaining the formation interface distribution throughout the well section. This invention performs sliding windowing processing on logging data, establishes an initial interpretation model for adjacent wells / pilot wells, and uses a pilot model for manual interactive adjustment to determine the location of the build-up section during drilling of deviated / horizontal wells. In horizontal well sections, it uses a combination of automatic interface tracking mode and expert mode to identify formation interfaces, making full use of the resistivity-constrained inversion model and human experience.

Description

Technical Field

[0001] This invention relates to the field of petroleum exploration technology, and in particular to a real-time semi-automatic method for tracking formation interfaces in logging-while-drilling resistivity. Background Technology

[0002] During the drilling of deviated / horizontal wells, real-time tracking of formation interfaces is one of the primary issues that needs to be addressed in logging-while-drilling (LWD) geosteering. It is crucial for accurate wellbore entry, controlling the wellbore trajectory within the target formation, and maximizing oil and gas resource development. While LWD electromagnetic resistivity logging instruments, combined with instrument response, allow for manual tracking of formation interfaces through interactive formation correlation, this method is cumbersome and time-consuming, making it unsuitable for real-time tracking during LWD. Automatic LWD electromagnetic resistivity logging inversion, on the other hand, suffers from multiple solutions. Therefore, accurate and rapid real-time tracking of formation interfaces is one of the key issues that real-time geosteering in LWD must address. Furthermore, combining interactive formation correlation and automatic tracking methods simultaneously is essential for achieving real-time formation interface tracking in LWD electromagnetic resistivity logging. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention discloses a real-time semi-automatic method for tracking formation interfaces in drilling resistivity logging, aiming to provide formation interface imaging information for real-time formation interface tracking.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A semi-automatic real-time tracking method for formation interfaces in logging-while-drilling resistivity includes the following steps:

[0006] s1. Input logging data from adjacent wells / pilot wells and highly deviated wells, and extract formation information from the logging data of adjacent wells / pilot wells, including layer thickness, gamma data and resistivity, etc.

[0007] s2. Based on the formation information of adjacent wells / pilot wells, establish an initial interpretation model, i.e., a pilot model, to determine whether the current formation is in the build-up section. If so, proceed to s3; otherwise, proceed to s4.

[0008] s3. Determine the location of the stratigraphic interface of the orthographic section: First, compare the difference between the model gamma value and the measured value at the current location point, and manually adjust the interface to keep the gamma response consistent; use the updated stratigraphic model to simulate the apparent resistivity response; compare the simulated data with the measured response. If they are basically consistent, proceed with the processing of the subsequent orthographic section; otherwise, manually adjust the location of the stratigraphic interface until the measured and simulated curve results are consistent; based on the gamma and resistivity data, if the instrument enters the target layer, execute s4.

[0009] s4. Enable automatic interface tracking mode for horizontal well sections, construct resistivity constraint interpretation model, combine known prior information, adopt constraint inversion model and nonlinear optimization method to quickly invert the upper and lower interface positions of the target layer, and calculate the corresponding simulation data;

[0010] s5. Determine whether the error between the simulated data and the measured data exceeds the threshold. If yes, proceed to step s6; otherwise, proceed to step s7.

[0011] s6. Enable expert mode to identify stratigraphic interfaces, and comprehensively judge whether there are situations such as entering faults, thin interlayers, or adjacent layers of the target layer by combining geological information and well logging data. Combine rapid forward modeling, interactive comparison and inversion methods to manually adjust stratigraphic interfaces.

[0012] s7. Execute s4-s6 sequentially on the data window of the horizontal section to obtain the formation interface distribution of the entire well section.

[0013] Optionally, step s4, the step of quickly inverting the positions of the upper and lower interfaces of the target layer, specifically includes:

[0014] s4.1. Based on three types of prior information: pilot model information established from logging data of adjacent wells / guide wells, formation resistivity and interface distance transformation model information established through neural network algorithm training, and formation interface information obtained by processing adjacent sections; all three types of prior information can provide initial value information such as the thickness of the layer where the instrument is located, resistivity, distance from the formation interface, and surrounding rock resistivity.

[0015] s4.2. Based on the number of existing data curves, establish a multi-layer inversion model, considering only the current layer interface location or resistivity is unknown, while the surrounding rock resistivity and layer thickness are known (information from the pilot model):

[0016] If there are more than two curves, then the parameters to be inverted include three, namely H. up H down and R t Among them, R t H represents the resistivity of the target layer. up H represents the distance from the instrument to the upper stratigraphic interface. down Indicates the distance between the instrument and the lower stratum interface;

[0017] If there are two curves, then the parameters to be inverted include two, namely H. up and H down ;

[0018] If there is only one resistivity curve, assuming the current layer thickness is fixed, the parameters to be inverted include one H. up ;

[0019] s4.3. Convert the apparent resistivity curve into phase difference and amplitude ratio. Combined with the inversion model established in s4.2, use the regularized Gauss-Newton algorithm to directly invert the location of the formation interface. If the fitting difference obtained from the inversion is less than the threshold, then the subsequent horizontal segments are inverted to update the interface location in real time and automatically.

[0020] Optionally, step s6, which involves enabling expert mode to identify the formation interface and manually adjusting the formation interface, specifically includes:

[0021] s6.1. Compare the measured gamma curve with the gamma curve of the current layer. If the two are basically the same, but the apparent resistivity curve is much higher than the resistivity value of the current layer, it means that the instrument is close to the formation interface. At this time, manually adjust the formation model interactively. If the gamma curve changes abruptly, it means that the instrument has penetrated the target layer, i.e. the current layer. Execute s6.2.

[0022] s6.2. Compare the gamma value of the newly entered layer with the gamma value of the adjacent layer of the target layer in the pilot model. If the two are basically the same, it means that the instrument has passed through the target layer and entered the adjacent layer. At this time, continue to start the automatic interface inversion described in s4. Otherwise, it means that the instrument has entered a complex stratigraphic structure.

[0023] s6.3. Compare the gamma and resistivity data of the current layer with those of other layers near the target layer in the pilot model. If they are consistent, it means that the instrument has penetrated the fault and drilled into other layers indicated in the pilot model. Therefore, use interactive forward modeling to adjust this part to a fault structure. If they are inconsistent, proceed to s6.4.

[0024] s6.4. The instrument has entered an isolated sand body or thin interlayer. Therefore, new layers or lenses are added to the pilot model, and their resistivity and interface position are continuously adjusted until the simulated response matches the measured response.

[0025] The beneficial effects of this invention are as follows: It performs sliding windowing processing on well logging data to establish an initial interpretation model for adjacent / pilot wells. During drilling of deviated / horizontal wells, the pilot model established based on adjacent or pilot wells is used for manual interactive adjustments to determine the location of the build-up section. In horizontal well sections, a combination of automatic interface tracking mode and expert mode is used to identify formation interfaces, making full use of resistivity-constrained inversion models and human experience. Compared to traditional methods, this method has a higher processing speed, provides accurate formation interface information, and contributes to more precise geological steering decisions and improved real-time processing and interpretation efficiency. Attached Figure Description

[0026] Figure 1 This is a flowchart of a real-time semi-automatic tracking method for formation interface in logging resistivity while drilling, as described in this invention.

[0027] Figure 2This is a 2D planar display diagram of measured data from a horizontal well in this invention;

[0028] Figure 3 shows the semi-automatic processing results of formation interfaces in horizontal wells: (a) the result after manual interactive interpretation of the build-up section; (b) the automatic inversion result of formation interfaces in the horizontal section; (c) the interlayer-fault section, tracked by the expert mode interactive interface.

[0029] Figure 4 The following are schematic diagrams of the automatic inversion model of formation interface under resistivity constraint in horizontal wells in this invention: (a) three-parameter inversion model; (b) two-parameter inversion model; (c) single-parameter inversion model;

[0030] Figure 5 The following is a simplified expert interpretation model of the stratigraphic structure in this invention: (a) a schematic diagram of the stratigraphic model with the instrument close to the strata; (b) a schematic diagram of the instrument passing through the target layer and entering the adjacent layer of the pilot model;

[0031] Figure 6 The following are expert interpretation models of complex stratigraphic structures in this invention: (a) Schematic diagram of the instrument passing through a fault structure model; (b) Schematic diagram of the stratigraphic structure passing through thin interlayers and isolated sand bodies in this invention;

[0032] Figure 7 This is a graph showing the results of logging resistivity data and semi-automatic interface tracking for a horizontal well in this invention. Detailed Implementation

[0033] 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.

[0034] A semi-automatic real-time tracking method for formation interfaces in logging-while-drilling resistivity, such as... Figure 1 As shown, it includes the following steps:

[0035] s1. For example Figure 2 As shown, input logging data from adjacent wells / pilot wells and highly deviated wells. The upper right of the figure shows the measured resistivity data, and the lower right shows the horizontal well drilling trajectory. P36H and P22H represent apparent resistivity curves with a frequency of 2MHz and source distances of 36 inches and 22 inches, respectively. Formation information, including layer thickness, gamma data, resistivity, and formation lithology, is extracted from the logging data of adjacent wells / pilot wells.

[0036] s2. Based on the formation information of adjacent wells / pilot wells, establish an initial interpretation model, i.e., a pilot model, to determine whether the current formation is in the build-up section. If so, proceed to s3; otherwise, proceed to s4.

[0037] s3. Determine the location of the stratigraphic interface for the directional drilling section: First, compare the difference between the model gamma value and the measured value at the current location point, and manually adjust the interface to ensure consistent gamma response; use the updated stratigraphic model to simulate the apparent resistivity response; compare the simulated data with the measured response. If they are basically consistent, proceed with the processing of subsequent directional drilling sections; otherwise, manually adjust the stratigraphic interface location until the measured value is consistent. Figure 3a The results of the solid line and the simulation (curve) are consistent. The results of the interaction tracking of the stratigraphic interface in the sloping section are shown in Figure 3(a) at a lateral depth of 60-85m. In the figure, P36H_S and P22H_s represent the inversion and reconstruction resistivity curves with source distances of 36 inches and 22 inches, respectively. According to the gamma and resistivity data, if the instrument enters the target layer, execute s4.

[0038] The above design facilitates the separate processing of the orogenic section and the horizontal section strata, thereby improving the accuracy of tracing the stratigraphic interface. Generally, in the orogenic section, resistivity, gamma value, density value, etc., change with the penetrated strata, and combining this with human experience makes the judgment more accurate.

[0039] s4. Enable automatic interface tracking mode for horizontal well sections, construct resistivity constraint interpretation model, combine known prior information, adopt constraint inversion model and nonlinear optimization method to quickly invert the upper and lower interface positions of the target layer, as shown in Figure 3(b), and calculate the corresponding simulation data;

[0040] Specifically, it includes:

[0041] s4.1. Based on three types of prior information: pilot model information established from logging data of adjacent wells / guide wells, formation resistivity and interface distance transformation model information established through neural network algorithm training, and formation interface information obtained by processing adjacent sections; all three types of prior information can provide initial value information such as the thickness of the layer where the instrument is located, resistivity, distance from the formation interface, and surrounding rock resistivity.

[0042] s4.2. Based on the number of existing data curves, establish a multi-layer inversion model, considering only the current layer interface location or resistivity is unknown, while the surrounding rock resistivity and layer thickness are known (information from the pilot model):

[0043] like Figure 4 As shown in Figure a, if the number of curves is greater than two, then the parameters to be inverted include three, namely H. up H down and R t Among them, R t H represents the resistivity of the target layer. up H represents the distance from the instrument to the upper stratigraphic interface. down Indicates the distance between the instrument and the lower stratum interface;

[0044] like Figure 4 As shown in b, if there are two curves, then the parameters to be inverted include two, namely H. up and H down ;

[0045] like Figure 4 As shown in c, if there is only one resistivity curve, assuming the current layer thickness is fixed, the parameters to be inverted include one H. up ;

[0046] s4.3. Convert the apparent resistivity curve into phase difference and amplitude ratio. Combined with the inversion model established in s4.2, use the regularized Gauss-Newton algorithm to directly invert the location of the formation interface. If the fitting difference obtained from the inversion is less than the threshold, then the subsequent horizontal segments are inverted to update the interface location in real time and automatically.

[0047] s5. Determine whether the error between the simulated data and the measured data exceeds the threshold. If yes, proceed to step s6; otherwise, proceed to step s7.

[0048] s6. Enable expert mode to identify stratigraphic interfaces, and comprehensively judge whether there are situations such as entering faults, thin interlayers, or adjacent layers of the target layer by combining geological information and well logging data. Combine rapid forward modeling, interactive comparison and inversion methods to manually adjust stratigraphic interfaces.

[0049] Specifically, it includes:

[0050] s6.1. Compare the measured gamma curve with the current layer gamma curve. If the two are basically the same, but the apparent resistivity curve is much higher than the current layer resistivity value, then... Figure 7 The lateral depth range of 105m to 170m indicates that the instrument is close to the formation interface; the interpretation model is shown below. Figure 5 a; At this point, manually adjust the formation model interactively; if the gamma curve changes abruptly, it indicates that the instrument has penetrated the target layer, i.e., the current layer, and execute s6.2;

[0051] s6.2. Compare the gamma value of the newly entered layer with the gamma value of the adjacent layer of the target layer in the lead model. If the two are basically the same, it indicates that the instrument has penetrated the target layer and entered the adjacent layer. See [link to relevant documentation]. Figure 5 Model shown in b. For Figure 2 The example shown, with a lateral depth of 90m to 155m, exemplifies this situation; the stratigraphic interface tracing results are available in [link to example]. Figure 3b The middle section; at this point, continue to start the automatic interface inversion described in s4, otherwise it indicates that the instrument has entered a complex stratigraphic structure;

[0052] s6.3. Compare the gamma and resistivity data of the current layer with those of other layers near the target layer in the pilot model. If they are largely consistent, it indicates that the instrument has traversed the fault and drilled into other layers indicated in the pilot model. Therefore, interactive forward modeling is used to adjust this part to a fault structure, see [link to relevant documentation]. Figure 6The model shown in figure a. For Figure 2 In the example shown, the resistivity drops significantly in the lateral depth range of 185m to 230m. This drop is consistent with the gamma and resistivity measurements of other layers in the pilot model, indicating the presence of a fault. The manual tracing results of this stratigraphic interface, processed using expert mode, are shown below. Figure 3c The inner right dashed box; if the two are inconsistent, execute s6.4;

[0053] s6.4. The instrument has entered an isolated sand body or thin interlayer, therefore, it is necessary to add new layers or lens structures to the pilot model, such as... Figure 6 Model b. Its resistivity and interface position were continuously adjusted until the simulated response matched the measured response. Figure 2 In the example shown, within the lateral depth range of 160m to 185m, the shallow-depth curve P22H showed no anomalies, while the deep-depth curve P36H showed a significant increase, indicating the presence of other anomalies within the detection range of the deep-depth curve. Combining the gamma curve and resistivity, two isolated sand bodies were ultimately added, as shown in [reference needed]. Figure 3c Dotted frame on the left.

[0054] s7. Execute s4-s6 sequentially on the data window of the horizontal section to obtain the formation interface distribution of the entire well section.

[0055] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A semi-automatic real-time tracking method for formation interfaces in logging-while-drilling resistivity, characterized in that, Includes the following steps: s1. Input logging data from adjacent wells / pilot wells and highly deviated wells, and extract formation information from the logging data of adjacent wells / pilot wells; s2. Establish an initial interpretation model based on the formation information of adjacent wells / pilot wells, and determine whether the formation is in the build-up section. If so, proceed to s3; otherwise, proceed to s4. s3. Determine the location of the stratigraphic interface of the sloping section: First, compare the difference between the model gamma value and the measured value at the current location point, and manually adjust the interface to make the gamma response consistent; use the updated stratigraphic model to simulate the apparent resistivity response; compare the simulated data with the measured response. If the two are basically consistent, then proceed with the processing of the subsequent sloping section; otherwise, manually adjust the location of the stratigraphic interface until the measured and simulated curve results are consistent. s4. Enable automatic interface tracking mode for horizontal well sections to quickly invert the positions of the upper and lower interfaces of the target layer and calculate the corresponding simulation data; s5. Determine whether the error between the simulated data and the measured data exceeds the threshold. If yes, proceed to step s6; otherwise, proceed to step s7. s6. Enable expert mode to identify formation interfaces and manually adjust formation interfaces; s7. Execute s4-s6 sequentially on the data window of the horizontal section to obtain the formation interface distribution of the entire well section; Step s4, the step of quickly inverting the positions of the upper and lower interfaces of the target layer, specifically includes: s4.

1. Based on three types of prior information: pilot model information established from logging data of adjacent wells / pilot wells, formation resistivity and interface distance transformation model information established through neural network algorithm training, and formation interface information obtained by processing adjacent sections; s4.

2. Based on the number of existing data curves, a multi-layer inversion model is established, considering only the current layer interface location or resistivity is unknown, while the surrounding rock resistivity and layer thickness are known: If there are more than two curves, then the parameters to be inverted include three, namely: H up、 H down and R t ;in, R t Indicates the resistivity of the target layer. H up Indicates the distance of the instrument from the upper stratum interface. H down Indicates the distance between the instrument and the lower stratum interface; If there are two curves, then the parameters to be inverted include two, namely... H up and H down ; If there is only one resistivity curve, assuming the current layer thickness is fixed, the parameters to be inverted include one. H up ; s4.

3. Convert the apparent resistivity curve into phase difference and amplitude ratio. Combined with the inversion model established in s4.2, use the regularized Gauss-Newton algorithm to directly invert the location of the formation interface. If the fitting difference obtained from the inversion is less than the threshold, then the subsequent horizontal segments are inverted to update the interface location in real time and automatically.

2. The method for real-time semi-automatic tracking of formation interfaces in logging-while-drilling resistivity according to claim 1, characterized in that, Step s6, which involves enabling expert mode to identify the formation interface and manually adjusting the formation interface, specifically includes: s6.

1. Compare the measured gamma curve with the gamma curve of the current layer. If the two are basically the same, but the apparent resistivity curve is much higher than the resistivity value of the current layer, it means that the instrument is close to the formation interface. At this time, manually adjust the formation model interactively. If the gamma curve changes abruptly, it means that the instrument has penetrated the target layer. Execute s6.

2. s6.

2. Compare the gamma value of the newly entered layer with the gamma value of the adjacent layer of the target layer in the pilot model. If the two are basically the same, it means that the instrument has passed through the target layer and entered the adjacent layer. At this time, continue to start the automatic interface inversion described in s4. Otherwise, it means that the instrument has entered a complex stratigraphic structure. s6.

3. Compare the gamma and resistivity data of the current layer with those of other layers near the target layer in the pilot model. If they are relatively consistent, it indicates that the instrument has penetrated the fault and drilled into other layers indicated in the pilot model. Therefore, use interactive forward modeling to adjust this part to a fault structure. If they are inconsistent, proceed to s6.

4. s6.

4. The instrument enters an isolated sand body or thin interlayer. Therefore, new layers or lenses are added to the pilot model, and their resistivity and interface position are continuously adjusted until the simulated response matches the measured response.

Images