A method for predicting water-locking at the shale oil reservoir scale
By combining core experiments and numerical simulations to predict the degree of water-locking at the shale oil reservoir scale, the dynamic evolution of water-locking damage in shale oil reservoirs at the reservoir scale was solved, enabling accurate prediction of water-locking damage and optimization of production parameters, thereby improving the development efficiency of shale oil reservoirs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA UNIV OF PETROLEUM (EAST CHINA)
- Filing Date
- 2026-07-02
- Publication Date
- 2026-07-31
AI Technical Summary
Existing technologies are insufficient to accurately characterize the dynamic evolution of water-locking damage at the shale oil reservoir scale, especially in the relationship between fracturing fluid migration and distribution and production loss throughout the entire process of well shut-in, flowback, and production. This makes it impossible to directly generalize core experimental results to the reservoir scale.
Conventional core experiments combined with nuclear magnetic resonance (NMR) technology were used to monitor the migration and distribution of fracturing fluid in the fracture-matrix system. A numerical model of the CMG-IMEX simulator was established, calibrated with core experiment results, and extended to reservoir space and time scales to predict the degree of water-locking damage.
It enables quantitative characterization and dynamic evaluation of water-locking at the shale reservoir scale, optimizes well shut-in time and production parameters, reduces water-locking damage, and improves shale oil reservoir development efficiency.
Smart Images
Figure CN122490865A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of unconventional oil and gas extraction technology, and in particular to a method for predicting the degree of water lock at the shale oil reservoir scale. Background Technology
[0002] Shale oil, as a typical unconventional oil and gas resource, is an important component for ensuring my country's energy security and optimizing its energy structure. Due to the generally ultra-low porosity and ultra-low permeability characteristics of shale oil reservoirs, their economic development heavily relies on horizontal well volumetric fracturing technology. During fracturing, a large amount of fracturing fluid enters the matrix or unsupported fractures and is difficult to flow back in time, easily causing water-locking damage in the pore throat and fracture-matrix system, thereby reducing oil phase permeability and affecting subsequent production.
[0003] Existing research on water-locking phenomena in shale oil is mostly concentrated at the laboratory core scale. The experimental spatial and temporal scales are significantly smaller than those of actual reservoir development processes, making it difficult to directly characterize the dynamic evolution of water-locking damage at both temporal and spatial scales. Furthermore, existing studies typically focus on the well-clogging stage or a single production phase, failing to adequately characterize the continuous relationship between fracturing fluid migration and distribution and production losses throughout the entire process of well clogging, flowback, and production.
[0004] Core test results cannot be directly extrapolated to the reservoir scale, and it is difficult to evaluate the impact of parameters such as different well-clogging times, production pressure differentials, and production times on water-locking levels. Therefore, it is necessary to establish a method that couples physical simulation, NMR dynamic monitoring, and numerical simulation to predict water-locking levels and optimize production parameters at the shale reservoir scale. Summary of the Invention
[0005] To address the aforementioned technical problems, this invention discloses a method for predicting the degree of water-locking in shale oil reservoirs. This method monitors the migration and distribution of fracturing fluid (water phase) in the fracture-matrix system using conventional core experiments and nuclear magnetic resonance (NMR) technology. It obtains the changes in water saturation and water-locking damage coefficients throughout the entire production process of the core sample. Then, a numerical model calibrated with the core experiment results is established using the CMG-IMEX simulator. Subsequently, the model's grid size and simulation time are modified to extend it to the reservoir space and time scale, thereby predicting the dynamic evolution of water-locking damage in actual reservoir production. This provides a basis for optimizing well shut-in time, production pressure differential, and production regime.
[0006] To achieve the above objectives, the present invention adopts the following technical solution:
[0007] A method for predicting water-locking at the shale oil reservoir scale, specifically including the following steps:
[0008] Step (1) Shale core simulation experiment;
[0009] Using a core displacement system, a water-locking phenomenon simulation experiment was conducted on the core of the target shale reservoir to simulate the entire process of well stagnation, flowback and production after fracturing of the shale reservoir. The laboratory timescale was used to obtain the changes in the one-dimensional frequency-coded water phase signal and the cumulative oil production during the entire process of well stagnation, flowback and production.
[0010] Step (2) Experimental data processing;
[0011] The experimental data obtained in step (1) are processed to obtain the water saturation change curve of the core throughout the entire production stage, and the water-lock damage coefficient is used to quantify the water-lock damage.
[0012] Step (3) Establish a core data model and fit the experimental results;
[0013] A numerical model consistent with the core experiment size was established using the CMG-IMEX simulator. The numerical model characterizing the above core experiment was obtained by fitting the water saturation change and water-locking damage coefficient during the entire core production stage.
[0014] Step (4) extends the fitted numerical model to the reservoir scale;
[0015] The model is extended from spatial and temporal scales to the reservoir scale to obtain a reservoir-scale water-lock dynamic evolution prediction model. Based on the calibrated core-scale model, the reservoir-scale water-lock prediction model divided by different fracture networks can be simulated by extending the model length in the I direction. The number of grids and grid width in the J and K directions of the reservoir-scale model are consistent with the core-scale model. Only the number of grids and grid width in the I direction are adjusted to achieve the extension from the core scale to the reservoir scale.
[0016] Furthermore, step (1) simulates the specific process of the experiment:
[0017] 1.1 The selected shale core samples were washed, dried, and the nuclear magnetic resonance signals of the dry cores were scanned to establish the nuclear magnetic resonance matrix signal of the dry cores;
[0018] 1.2. Use fluorinated liquid to saturate the target core to establish the core oil saturation;
[0019] 1.3 After saturation is complete, the core is placed in a nuclear magnetic resonance analyzer to obtain the one-dimensional frequency code of the corresponding core.
[0020] 1.4. Fluorinated fluid was injected at constant pressure from the right end face of the core at the target reservoir pressure to simulate a constant reservoir pressure. At the same time, KCl solution was injected at constant pressure from the left end face of the core to simulate the invasion of fracturing fluid.
[0021] 1.5. Maintain the state of steps 1.3 and 1.4 and perform well shut-in operation, measuring the one-dimensional frequency-coded signal of nuclear magnetic resonance every fixed number of days;
[0022] 1.6. After well shut-in, simulate flowback and production, and measure the one-dimensional frequency-encoded nuclear magnetic resonance signal every fixed number of days. Use a cumulative volumetric flow meter to record the cumulative total oil production.
[0023] 1.7 Set up a waterless control experiment, omit the well-clogging operation, keep the rest of the process the same, and record the cumulative total oil production;
[0024] 1.8 After completion, shut down the system, remove the core, clean the device, and complete the simulation experiment.
[0025] Furthermore, in step (2), the water phase content in the core is quantitatively characterized using nuclear magnetic resonance signal intensity, and a conversion relationship between the one-dimensional frequency-coded total signal quantity and the core saturated water quantity is established; the specific process is as follows:
[0026] 2.1. Measure the mass and NMR matrix signal of the dried core after oil washing and drying;
[0027] 2.2 After the core is vacuum-saturated, it is weighed and the one-dimensional frequency code of the water-saturated core is measured;
[0028] 2.3 Centrifuge the core at speeds of 2000 r / min, 4000 r / min, and 8000 r / min for 1 h, weigh them, and measure the T2 spectrum after centrifugation;
[0029] 2.4 Calculate the mass difference before and after centrifugation and the difference of the one-dimensional frequency-coded cumulative signal, and establish the conversion relationship between nuclear magnetic resonance signal quantity and core water content.
[0030] Furthermore, step (2) yields a quantitative relationship between core water content and one-dimensional NMR frequency encoding:
[0031] (1);
[0032] In the formula, The water content of the core is expressed in mL. This represents the total amount of one-dimensional frequency-coded signals of the water phase in the core.
[0033] After converting the one-dimensional frequency-encoded signal into core water content, the water saturation at various locations in the core is calculated:
[0034] (2);
[0035] In the formula, This represents the water saturation level of the core sample. coordinates The water content in the pores at a given location when the pores are fully saturated with water, i.e., the pore volume, in mL; for Pore volume at [location], mL; for The one-dimensional frequency-coded signal quantity of nuclear magnetic resonance at a certain moment; Fully saturated water core One-dimensional frequency-coded signal quantity at the NMR location;
[0036] The water saturation variation curve of the core throughout the entire process is obtained using formula (2).
[0037] Furthermore, in step (2), the damage to the waterlock is quantitatively evaluated using a waterlock damage coefficient. The specific process is as follows: the waterlock damage coefficient is defined in the following form:
[0038] (3);
[0039] in, This refers to the water lock damage coefficient. The relative permeability of the oil phase when water lock has not occurred; The relative permeability of the oil phase under water-locking effect;
[0040] According to Darcy's Law:
[0041] (4);
[0042] in, Oil production rate; This represents the absolute permeability of the core. A represents the relative permeability of the oil phase; A represents the cross-sectional area of the core. For production pressure differential; Crude oil viscosity; The length of the core sample;
[0043] For cases without a waterlock:
[0044] (5);
[0045] Regarding waterlock situations:
[0046] (6);
[0047] in, The oil production rate without water lock is expressed in mL / min. The oil production rate during water lock, in mL / min;
[0048] Dividing the oil production rate under water-locked conditions by the oil production rate under non-water-locked conditions, i.e., formula (6) / (5), yields:
[0049] (7);
[0050] Substituting formula (7) into formula (3) yields:
[0051] (8);
[0052] The waterlock damage coefficient is calculated using the cumulative oil production. By integration, formula (8) is rewritten as the formula for the waterlock damage coefficient calculated using the cumulative oil production:
[0053] (9);
[0054] The water-lock damage coefficient of the core experiment was calculated using formula (9).
[0055] Furthermore, in step (3), the formula for determining the fitting accuracy is:
[0056] (10);
[0057] (11);
[0058] in, The loss function is used to fit the water saturation curve. This represents the number of water saturation points fitted. This represents the water saturation value of the matrix at a certain point during the well-steaming process in the numerical model. This represents the water saturation value of a certain point in the matrix during the core experiment's well-steaming process. This represents the average value of the matrix water saturation during the well-steaming process in the numerical model. This represents the average value of matrix water saturation during the core experiment's well-steaming process. This represents the water saturation value of a certain point in the matrix during the numerical model production process. This represents the water saturation value at a certain point in the matrix during core experiment production. This represents the average value of the matrix water saturation during the numerical model production process. This represents the average value of the matrix water saturation during the core experiment production process. The loss function fitted to the waterlock damage coefficient; The waterlock damage coefficient in the numerical model; The water-lock damage coefficient is the core experiment water-lock damage coefficient.
[0059] Furthermore, in step (3), when the fitted result is obtained... Less than or equal to 0.01 If the value is less than or equal to 0.001, the fitting accuracy is considered to meet the requirements, indicating that the numerical model can characterize the previous core experiment results.
[0060] The beneficial effects of this invention are:
[0061] 1. This method extends the water-locking damage measured by indoor core experiments to the reservoir scale, realizing the quantitative characterization of water-locking damage at the reservoir scale in the field.
[0062] 2. A dynamic evaluation method for water-lock damage in shale reservoirs over time was constructed, enabling accurate description of the evolution of water-lock in actual production processes and precise prediction of water-lock damage.
[0063] 3. A displacement experimental system was built that can simulate the entire shale production process and monitor the migration and distribution of fracturing fluid in the shale fracture-matrix system in real time.
[0064] 4. The established reservoir-scale model can be used to screen appropriate well-clogging time and economic production pressure differential, reduce water-locking damage, and improve the development efficiency of shale oil reservoirs. Attached Figure Description
[0065] Figure 1 This is a schematic diagram of the process of the present invention;
[0066] Figure 2 This invention provides a full-process simulation device for shale production.
[0067] Figure 3 This is the one-dimensional frequency encoding change curve of nuclear magnetic resonance during the core experiment well-steaming process in this invention;
[0068] Figure 4 This is a curve showing the change of one-dimensional frequency-coded nuclear magnetic resonance signal during the core experiment production process in this invention.
[0069] Figure 5 This is a quantitative relationship diagram between core water content and one-dimensional frequency-coded nuclear magnetic resonance signal in this invention;
[0070] Figure 6 This is the water saturation change curve during the core experiment well-steaming process in this invention;
[0071] Figure 7 This is the water saturation change curve during the core experiment production process in this invention;
[0072] Figure 8 This is a schematic diagram of the core numerical simulation model in this invention;
[0073] Figure 9 This is the fitting result of water saturation during the core experiment and core numerical model well shut-in process in this invention;
[0074] Figure 10 This is the fitting result of water saturation during the core experiment and core numerical model production process in this invention;
[0075] Figure 11 This is a schematic diagram illustrating the division of the reservoir matrix by different types of fractures in this invention;
[0076] Figure 12 This is the dynamic evolution curve of waterlock damage in the 100m scale model of this invention. Detailed Implementation
[0077] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 a part of the embodiments of the present invention, and not all of them. 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.
[0078] This invention addresses the challenge of water-locking experiments on shale cores failing to accurately represent actual reservoir production, by providing a method for predicting water-locking levels at the shale reservoir scale. Through a combination of a full-process shale production simulation device, nuclear magnetic resonance monitoring, and numerical simulation, the method achieves prediction of water-locking levels and optimization of production parameters at the shale reservoir scale.
[0079] This invention discloses a method for predicting water-locking at the shale reservoir scale, employing methods such as... Figure 1 The flowchart shown is as follows: Figure 1 As shown. Specifically,
[0080] (1) First, a core water-locking phenomenon simulation experiment was conducted on the target shale reservoir. The purpose of the experiment was to obtain the changes in the one-dimensional frequency-coded water phase signal and the cumulative oil production during the entire process of core well shut-in, flowback, and production on a laboratory timescale. The simulation experiment adopted the following methods: Figure 2 The core displacement system shown reproduces the entire process of well shut-in, flowback, and production after fracturing in a shale reservoir. To avoid interference from the oil phase signal, a fluorinated fluid with virtually no NMR signal was used instead of crude oil, and a 4% KCl solution was used as the fracturing fluid to reduce the impact of clay expansion on the experimental results. This ensures that the NMR signal only characterizes the migration and distribution of the fracturing fluid (aqueous phase) in the fracture-matrix system.
[0081] The core experiment procedure is as follows:
[0082] ① The selected shale core samples were washed, dried, and the nuclear magnetic resonance signals of the dry cores were scanned to establish the nuclear magnetic resonance basement signal of the dry cores.
[0083] ② Use fluorinated liquid to saturate the target core to establish the core oil saturation.
[0084] ③ After saturation is complete, the core is placed in a nuclear magnetic resonance analyzer to obtain the one-dimensional frequency code of the corresponding core.
[0085] ④ Open valves 6, 7, 8, and 9, and use pump 2 to inject fluorinated fluid at constant pressure from the right end face of the core at the target reservoir pressure, simulating that the reservoir pressure remains constant; open valves 2, 3, 4, and 5, and use pump 1 to inject KCl solution at constant pressure from the left end face of the core at the fracturing pressure, simulating the invasion of fracturing fluid.
[0086] ⑤ Maintain states ③ and ④ during the well-sealing operation (example) Figure 3 The well soaking time was set to 15 days, and then the one-dimensional frequency-coded nuclear magnetic resonance signal was measured every fixed number of days (every 5 days in this example). The signal results are as follows. Figure 3 As shown.
[0087] ⑥ After well shutting, close pump 1 and valves 2, 3, 4, and 5. Simulate flowback and production by opening valve 1 and setting the pressure of pump 3 connected to the back pressure valve to the target wellbore production pressure (7 MPa in the example). Figure 4 The production time is set to 15 days. Subsequently, the one-dimensional frequency-coded NMR signal is measured every fixed number of days (every 5 days in this example), and the signal results are as follows. Figure 4 As shown. Because the permeability of shale cores is too low, the oil production rate obtained from real-time monitoring has a large error. Therefore, a cumulative volumetric flow meter is used to record the cumulative total oil production.
[0088] ⑦ Set up a waterless control experiment, omit the well-clogging operation (i.e. simple displacement), keep the rest of the process the same, and record the cumulative total oil production.
[0089] ⑧ After completion, shut down the system, remove the core sample, and clean the device.
[0090] (2) The data obtained in the first step of the experiment were processed to obtain the water saturation change and water lock damage coefficient of the core throughout the entire production process.
[0091] Theoretically, under the condition that the test parameters, temperature, magnetic field strength, and instrument gain remain consistent, the amplitude of the NMR signal is related to the amplitude of the measured fluid. 1 The amount of H is positively correlated; for a relatively stable single fluid... 1 The amount of H is approximately proportional to the volume of the fluid. Therefore, we can use the nuclear magnetic resonance signal intensity to quantitatively characterize the water phase content in the core, thereby establishing a one-dimensional frequency-coded total signal quantity and a conversion relationship between the core saturated water quantity. The specific steps are as follows:
[0092] ① Measure the mass and NMR matrix signal of the dried core after washing and drying. ② After vacuum saturation of the core, weigh it and measure the one-dimensional frequency code of the water-saturated core.
[0093] ③ Centrifuge the core at speeds of 2000, 4000, and 8000 r / min for 1 h, weigh them, and measure the T2 spectrum after centrifugation.
[0094] ④ Calculate the mass difference before and after centrifugation and the difference in the one-dimensional frequency-encoded cumulative signal, and establish the conversion relationship between nuclear magnetic resonance signal quantity and core water content.
[0095] Following the steps above, we can obtain a calibration curve that exhibits excellent linearity (e.g., Figure 5 As shown in the example, the quantitative relationship between core water content and its one-dimensional NMR frequency encoding is obtained (using the following formula as an example):
[0096] (1);
[0097] In the formula, This refers to the water content of the core. This represents the total amount of one-dimensional frequency-coded signals of the water phase in the core.
[0098] After converting the one-dimensional frequency-encoded signal into core water content, we can then use this to calculate the water saturation at various locations in the core (using the following formula as an example):
[0099] ;
[0100] In the formula, This represents the water saturation level of the core sample. coordinates The water content in pores when they are fully saturated with water, i.e., pore volume; for The pore volume at that location; for The one-dimensional frequency-coded signal quantity of nuclear magnetic resonance at a certain moment; Fully saturated water core The one-dimensional frequency-coded signal quantity of nuclear magnetic resonance at that location.
[0101] Formula (2) can be used to... Figure 3 and Figure 4 Transform into Figure 6 and Figure 7 The curve showing the change in water saturation throughout the entire process of core samples.
[0102] To intuitively explore the impact of water lock on reservoir productivity, a quantitative evaluation of water lock damage is needed. Below, we quantify water lock damage using a water lock damage coefficient. Essentially, water lock damage manifests as a decrease in the effective permeability of the oil phase within the matrix after water intrusion. The water lock damage coefficient is typically defined as follows:
[0103] (3);
[0104] in, This refers to the water lock damage coefficient. The relative permeability of the oil phase when water lock has not occurred; The relative permeability of the oil phase under water-locking effect.
[0105] when When, it indicates no waterlock damage; when This indicates that the oil phase's permeability is almost completely lost, and the water lock damage is severe.
[0106] However, it is difficult to obtain the relative permeability of the oil phase in the core matrix in real time during core experiments. Therefore, it is impractical to calculate the water-locking damage coefficient of the core using formula (3), and a simple and feasible calculation method is needed. According to Darcy's law:
[0107] (4);
[0108] in, Oil production rate; This represents the absolute permeability of the core. A represents the relative permeability of the oil phase; A represents the cross-sectional area of the core. For production pressure differential; Crude oil viscosity; This represents the core length.
[0109] For the case without a waterlock, we have:
[0110] (5);
[0111] Regarding the water lock situation, we have:
[0112] (6);
[0113] in, The oil production rate without water lock; The oil production rate during water lock.
[0114] Dividing the oil production rate under water-locked conditions by the oil production rate under non-water-locked conditions, i.e., formula (6) / (5), we can obtain:
[0115] (7);
[0116] Substituting formula (7) into formula (3) yields:
[0117] (8);
[0118] Because the permeability of shale cores is too low, the oil production rate obtained from real-time monitoring has a large error. Therefore, the cumulative oil production is used to calculate the waterlock damage coefficient. By integration, formula (8) can be rewritten as the formula for the waterlock damage coefficient calculated using the cumulative oil production:
[0119] (9);
[0120] The water-lock damage coefficient of the core experiment can be calculated using formula (9). The calculated water-lock damage coefficient in this example is 0.70, indicating that the water-lock damage is relatively serious.
[0121] (3) A numerical model with the same size as the above core experiment was established using the CMG-IMEX simulator. By fitting the water saturation change and water lock damage coefficient of the whole production stage of the core, a numerical model that can characterize the above core experiment was obtained.
[0122] Establish the following in Cartesian coordinate system using the CMG-IMEX simulator: Figure 8 The core numerical model shown is divided into a matrix zone and a fracture zone, similar to the experimental core. Its dimensions are 0.1 m in the I direction and 0.025 m in the J and K directions, with a fracture width of 1 mm. All dimensions are consistent with the previous core experiments. Initially, the matrix zone is saturated with oil (corresponding to the fluorinated fluid in the core experiment), and the fracture zone is saturated with water (corresponding to the fracturing fluid in the core experiment). The model has a production well and an injection well at each end. Injection well 1 is used to inject fracturing fluid, and production well 2 is used to produce oil. Production well 1 and injection well 2 are used to ensure that the pressure at the left end of the model remains at the reservoir pressure during well shut-in and flowback production processes.
[0123] A specific simulation process example (the process settings are consistent with the core experiment in the previous part) is as follows: When simulating the fracturing fluid simmering process, the right injection well 1 is injected with water at a constant pressure of 40MPa for 15 days, and the left production well 1 is produced with oil at a constant pressure of 26MPa for 15 days, and then shut down; subsequently, when simulating the flowback production process, the left injection well 2 is injected with oil at a constant pressure of 26MPa for 15 days (simulating that there is always crude oil replenishment in the deep matrix), and the right production well 2 is produced with a constant pressure of 7MPa for 15 days.
[0124] The key parameters and boundary conditions of the model are consistent with those of the core experiments. The capillary force curves for the matrix portion of the model are obtained through mercury intrusion porosimetry, while no capillary force curves are set for the fracture portion, aiming to make the simulation more closely resemble the core experiments. The fitting process uses Corey-type functions to parameterize the relative permeability curves. Instead of directly inputting an entire relative permeability table point by point, several physically meaningful parameters are used to generate the entire oil-water relative permeability curve through mathematical functions. Then, in the CMG-CMOST history fitting, the endpoint relative permeability and curve shape parameters are continuously adjusted to make the target parameters of our fitting close to the experimental results, without having to modify the table data one by one.
[0125] The following formulas can be used to determine whether the fitting accuracy meets the requirements:
[0126] (10);
[0127] (11);
[0128] in, The loss function is used to fit the water saturation curve; This represents the number of water saturation points fitted. This represents the water saturation value of the matrix at a certain point during the well-steaming process in the numerical model. This represents the water saturation value of a certain point in the matrix during the core experiment's well-steaming process. This represents the average value of the matrix water saturation during the well-steaming process in the numerical model. This represents the average value of matrix water saturation during the core experiment's well-steaming process. This represents the water saturation value of a certain point in the matrix during the numerical model production process. This represents the water saturation value at a certain point in the matrix during core experiment production. This represents the average value of the matrix water saturation during the numerical model production process. This represents the average value of the matrix water saturation during the core experiment production process. The loss function fitted to the waterlock damage coefficient; The waterlock damage coefficient in the numerical model; The water-lock damage coefficient is the core experiment water-lock damage coefficient.
[0129] When obtained after fitting Less than or equal to 0.01 If the value is less than or equal to 0.001, the fitting accuracy is considered to meet the requirements, indicating that the numerical model can characterize the previous core experiment results.
[0130] Figure 9 , Figure 10 This is the result of fitting the water saturation curves of the well-sealing process and the production process, as illustrated in the previous example. The value is 0.008, which meets the requirements; in addition, the waterlock damage coefficients calculated by the experiment and simulation using formula (9) are 0.70 and 0.68, respectively. The value is 0.0004, which meets the requirements. Therefore, we have obtained a numerical model that can well characterize the previous core experiment results.
[0131] (4) Based on the numerical model obtained in step (3) that can characterize the core experiment results, the model is extended from the spatial and temporal scales to the reservoir scale, thereby obtaining a reservoir-scale water-lock dynamic evolution prediction model. After large-scale hydraulic fracturing in horizontal wells, actual shale reservoirs will contain different types of fractures, such as primary hydraulic fractures, secondary fractures, and natural fractures. These fractures will divide the reservoir matrix into rock blocks of different sizes, such as... Figure 11 As shown. Based on the calibrated core-scale model, a reservoir-scale water-lock prediction model divided by different fracture networks can be simulated by extending the model length in the I direction. The number of grids and grid width in the J and K directions of the reservoir-scale model are consistent with the core-scale model. Only the number of grids and grid width in the I direction are adjusted to achieve the extension from the core scale to the reservoir scale.
[0132] Taking the prediction of the dynamic evolution of matrix water lock in the main hydraulic fracture division as an example, the following example operations will be performed on the calibrated model. Taking the model extended to a 100 m scale in the I direction as an example, 990 0.1 m grids, 94 0.01 m grids, and 60 0.001 m grids can be set (because the water intrusion depth is only on the centimeter level, the grid can be locally refined only for the grids near the matrix-fracture surface), and the production time can be extended to the reservoir development time scale (e.g., Figure 11 The 1-year (100 m scale model prediction results are shown below) Figure 12 As shown in the figure. After the model is established, the water-locking damage coefficient under different production regimes can be predicted by changing parameters such as well-clogging time, production pressure differential, and production time.
[0133] Through the above steps, quantitative prediction of water-locking at the reservoir scale can be achieved from core experiments, and appropriate well-clogging time and economic production pressure differential can be selected accordingly to reduce water-locking damage and improve shale oil reservoir development efficiency.
[0134] 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 method for predicting the degree of water locking at the shale oil reservoir scale, characterized in that, Specifically, the following steps are included: Step (1) Shale core simulation experiment; Using a core displacement system, a water-locking phenomenon simulation experiment was conducted on the core of the target shale reservoir to simulate the entire process of well stagnation, flowback and production after fracturing of the shale reservoir. The laboratory timescale was used to obtain the changes in the one-dimensional frequency-coded water phase signal and the cumulative oil production during the entire process of well stagnation, flowback and production. Step (2) Experimental data processing; The experimental data obtained in step (1) are processed to obtain the water saturation change curve of the core throughout the entire production stage, and the water-locking damage coefficient is used to quantify the water-locking damage. Step (3) Establish a core data model and fit the experimental results; A numerical model consistent with the core experiment size was established using the CMG-IMEX simulator. By fitting the water saturation changes and water-locking damage coefficients during the entire core production process, a numerical model characterizing the core simulation experiment was obtained. Step (4) extends the fitted numerical model to the reservoir scale; The model is extended from spatial and temporal scales to the reservoir scale to obtain a reservoir-scale water-lock dynamic evolution prediction model. Based on the calibrated core-scale model, the reservoir-scale water-lock prediction model divided by different fracture networks is simulated by extending the model length in the I direction. The number of grids and grid width in the J and K directions of the reservoir-scale model are consistent with the core-scale model. Only the number of grids and grid width in the I direction are adjusted to achieve the extension from the core scale to the reservoir scale.
2. The method for predicting water-locking at the shale oil reservoir scale as described in claim 1, characterized in that, Step (1) Specific process of shale core simulation experiment: 1.1 The selected shale core samples were washed and dried, and the nuclear magnetic resonance signals of the dried cores were scanned to establish the nuclear magnetic resonance matrix signal of the dried cores; 1.
2. Use fluorinated liquid to saturate the target core to establish the core oil saturation; 1.3 After saturation is complete, the core is placed in a nuclear magnetic resonance analyzer to obtain the one-dimensional frequency code of the corresponding core. 1.
4. Fluorinated fluid was injected at constant pressure from the right end face of the core at the target reservoir pressure to simulate a constant reservoir pressure. At the same time, KCl solution was injected at constant pressure from the left end face of the core to simulate the invasion of fracturing fluid. 1.
5. Maintain the state of steps 1.3 and 1.4 and perform well shut-in operation, measuring the one-dimensional frequency-coded signal of nuclear magnetic resonance every fixed number of days; 1.
6. After well shut-in, simulate flowback and production, and measure the one-dimensional frequency-encoded nuclear magnetic resonance signal every fixed number of days. Use a cumulative volumetric flow meter to record the cumulative total oil production. 1.7 Set up a waterless control experiment, omit the well-clogging operation, keep the rest of the process the same, and record the cumulative total oil production; 1.8 After completion, shut down the system, remove the core, clean the device, and complete the simulation experiment.
3. The method for predicting water-locking at the shale oil reservoir scale as described in claim 2, characterized in that, In step (2), the water phase content in the core is quantitatively characterized using nuclear magnetic resonance signal intensity, and a conversion relationship between the one-dimensional frequency-coded total signal quantity and the core saturated water quantity is established; the specific process is as follows: 2.
1. Measure the mass and NMR matrix signal of the dried core after oil washing and drying; 2.2 After the core is vacuum-saturated, it is weighed and the one-dimensional frequency code of the water-saturated core is measured; 2.3 Centrifuge the core at speeds of 2000 r / min, 4000 r / min, and 8000 r / min for 1 h, weigh them, and measure the T2 spectrum after centrifugation; 2.4 Calculate the mass difference before and after centrifugation and the difference of the one-dimensional frequency-coded cumulative signal, and establish the conversion relationship between nuclear magnetic resonance signal quantity and core water content.
4. The method for predicting water-locking at the shale oil reservoir scale as described in claim 3, characterized in that, Step (2) yields the quantitative relationship between core water content and one-dimensional NMR frequency coding: （1）； In the formula, This refers to the water content of the core. The total amount of one-dimensional frequency-coded signal of the water phase in the core; After converting the one-dimensional frequency-encoded signal into core water content, the water saturation at various locations in the core is calculated: （2）； In the formula, This represents the water saturation level of the core sample. coordinates Moisture content at the location; for The water content in pores when they are fully saturated with water, i.e., pore volume; for The one-dimensional frequency-coded signal quantity of nuclear magnetic resonance at a certain moment; Fully saturated water core One-dimensional frequency-coded signal quantity at the NMR location; The water saturation variation curve of the core throughout the entire process is obtained using formula (2).
5. The method for predicting water-locking at the shale oil reservoir scale as described in claim 4, characterized in that, In step (2), the damage to the waterlock is quantitatively evaluated using a waterlock damage coefficient. The specific process is as follows: The waterlock damage coefficient is defined in the following form: （3）； in, This refers to the water lock damage coefficient. The relative permeability of the oil phase when water lock has not occurred; The relative permeability of the oil phase under water-locking effect; According to Darcy's Law: （4）； in, Oil production rate; This represents the absolute permeability of the core. A represents the relative permeability of the oil phase; A represents the cross-sectional area of the core. For production pressure differential; Crude oil viscosity; The length of the core sample; For cases without a waterlock: （5）； Regarding waterlock situations: （6）； in, The oil production rate without water lock; The oil production rate during water lock; Dividing the oil production rate under water-locked conditions by the oil production rate under non-water-locked conditions, i.e., formula (6) / (5), yields: （7）； Substituting formula (7) into formula (3) yields: （8）； The waterlock damage coefficient is calculated using the cumulative oil production. By integration, formula (8) is rewritten as the formula for the waterlock damage coefficient calculated using the cumulative oil production: （9）； The water-lock damage coefficient of the core experiment was calculated using formula (9).
6. The method for predicting water-locking at the shale oil reservoir scale as described in claim 5, characterized in that, In step (3), the formula for determining the fitting accuracy is: （10）； （11）； in, The loss function is used to fit the water saturation curve; This represents the number of water saturation points fitted. This represents the water saturation value of the matrix at a certain point during the well-steaming process in the numerical model. This represents the water saturation value of a certain point in the matrix during the core experiment's well-steaming process. This represents the average value of the matrix water saturation during the well-steaming process in the numerical model. This represents the average value of matrix water saturation during the core experiment's well-steaming process. This represents the water saturation value of a certain point in the matrix during the numerical model production process. This represents the water saturation value at a certain point in the matrix during core experiment production. This represents the average value of the matrix water saturation during the numerical model production process. This represents the average value of the matrix water saturation during the core experiment production process. The loss function fitted to the waterlock damage coefficient; The waterlock damage coefficient in the numerical model; The water-lock damage coefficient is the core experiment water-lock damage coefficient.
7. The method for predicting water-locking at the shale oil reservoir scale as described in claim 6, characterized in that, In step (3), when the fitted result is obtained Less than or equal to 0.01 If the value is less than or equal to 0.001, the fitting accuracy is considered to meet the requirements, indicating that the numerical model represents the previous core experiment results.