Recovery efficiency stability analysis method for fault block oil reservoir and application of recovery efficiency stability analysis method

Abstract

The invention relates to the field of oil exploitation, and particularly discloses a recovery efficiency stability analysis method for a fault block oil reservoir and application of the recovery efficiency stability analysis method. Comprising the following steps that S1, a well area adopting injection-production coupling mining in an oil field is selected for data collection; s2, selecting influence factors of the current injection-production contradiction, and carrying out a flat plate sand filling experiment and numerical simulation; s3, proposing an improved scheme according to an experimental result, and comparing the change conditions of the recovery ratio before and after the improved scheme is executed by means of the reliability analysis model; according to the method, the defects in the aspect of engineering verification are overcome on the basis of traditional flat plate sand filling experiments and numerical simulation research results. The invention provides a scheme analogy mathematical analysis method which is high in matching performance and high in responsivity. The problems of long development scheme verification period, weak parameter contrast and the like are solved, and the systematic working steps of experiment, data simulation, scheme proposing and mathematical demonstration of oil reservoir development are perfected.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of oil exploitation, and specifically discloses a recovery stability analysis method for fault block reservoirs and application thereof. BACKGROUND

[0002] The complex fault block reservoirs in the central and eastern China are mainly continental sedimentary reservoirs, and there are many oil layers with serious heterogeneity. Without deploying infill well network in new wells, most old oilfields improve the recovery of remaining oil by studying injection-production coupling (Liao, Changlin, et al. "Well Testing Analysis Methodology and Application for Complex Fault-Block reservoirs in the Exploration Stage." Gas&Oil Technology Showcase and Conference. OnePetro, 2023 and Wang, Qunyi, et al. "Study on the influence of different key parameters on EOR in LB block." E3S Web of Conferences. 2023, 375:01053). Therefore, it is of great practical significance to clarify the interlayer, intra-layer and plane heterogeneity problems and discuss the flow fixation problem of the original liquid flow direction for the development adjustment of old oil areas.

[0003] Currently, the research on injection-production coupling for the exploitation of remaining oil mainly focuses on mechanism, physical experiments, and oilfield applications. LSun (2023) conducted sandpack experiments for CO2 miscible flooding in tight reservoirs, proposed the correlation between TPG and CO2 concentration, and further fitted a mathematical model of miscible front migration. The most important parameter affecting CO2 breakthrough is injection rate (Sun, Lianting, et al. "A mathematical model of CO2 miscible front migration in tight reservoirs with injection-production coupling technology." Geoenergy Science and Engineering 221 (2023): 211376.). X Chen (2023) established a comprehensive energy consumption calculation model for the injection-oil reservoir-production stages to reduce the energy consumption of high-water-content oilfield development. The particle swarm optimization algorithm was used to propose a solution (Chen, Xianlei, et al. "Energy Consumption Reduction and Sustainable Development for Oil&Gas Transport and Storage Engineering." Energies 16.4 (2023): 1775.). H Zhai (2023) took the HDR oil reservoir geothermal development as an example and proposed a new type of multi-lateral horizontal well network accounting model. The sensitivity of the results was analyzed by changing the well pattern and injection-production parameters, and the heat production potential was evaluated (Zhai, Haizhen, et al. "Parametric study of the geothermal exploitation performance from a HDR reservoir through multilateral horizontal wells: The Qiabuqia geothermal area, Gonghe Basin." Energy (2023): 127370.). L Liu (2022) studied the deformation mechanism and law of natural / artificial fractures and matrix pores during the fracturing and injection-production processes in unconventional tight reservoirs, focusing on the relationship between injection energy supplement and the life cycle of oil and gas development.A dynamic model of multi-medium geometry, physical properties, and permeability was proposed (Liu, Lifeng, et al. "Hydro-mechanical coupling numerical simulation method of multi-scale pores and fractures in tight reservoir." IOP Conference Series: Earth and Environmental Science. 2022, 983(1): 012060.). SZheng (2022) proposed a fully coupled reservoir-fracture-wellbore model, which realized the modeling of fluid flow, solid mechanics, energy balance, fracture propagation, and particle filtration in the reservoir, fracture, and wellbore domains (Zheng, Shuang, and Mukul Sharma. "Coupling a Geomechanical Reservoir and Fracturing Simulator with a Wellbore Model for Horizontal Injection Wells." International Journal for Multiscale Computational Engineering 20.3 (2022.).). H Kesarwani (2022) conducted experimental analysis of α-MnO 2 nanoparticle additives for surfactant polymer flooding oil technology, resulting in a 70% increase in recovery rate (Kesarwani, Himanshu, et al. "Application of α-MnO 2 nanoparticles for residual oil mobilization through surfactant polymer flooding." Environmental Science and Pollution Research 29.29 (2022): 44255-44270.). A Aleidan (2017) explored the distribution characteristics of the remaining oil by comparing the information of the main oil layer and the remaining oil (filling time, hydrocarbon content, chemical characteristics) (Aleidan, Ahmed, et al. "Residual-oil zone: paleo-oil characterization and fundamental analysis." SPE Reservoir Evaluation&Engineering 20.02 (2017): 260-268.).N Gupta (2022) used field production data to conduct mathematical analysis for the case of Wasson oilfield and proposed a reliable, geoscience-driven ROZ development prediction technology (Gupta, Neha, et al. "Residual Oil Zone Recovery Evaluation and Forecast Methodology: A Wasson Field Case Study." SPE Improved Oil Recovery Conference. OnePetro, 2022.).

[0004] However, most of the existing achievements are obtained through a single research program, only discussing the inference of mechanism, the establishment of model and the phenomenon of development, lacking the combination of laboratory experiments and field development, and there is no timeliness evaluation method for development program (Nan, Jin-hao. "Study and Application of Resource Potential Evaluation in Complex Fault Block Reservoirs of Hailaer Basin." Proceedings of the International Field Exploration and Development Conference 2021. Singapore: Springer Nature Singapore, 2022: 2520-2529.).

[0005] In view of this, the present application is proposed. SUMMARY

[0006] The purpose of the present application is to provide a recovery stability analysis method for fault block reservoirs, which carries out quantitative research by means of plate sand filling experiment and numerical simulation, aims to explore the main controlling factors affecting recovery without changing the well pattern deployment, and proposes and verifies the development adjustment scheme, gives the engineering verification method of the scheme, and helps to explore the internal reasons of injection-production coupling disorder in the research area.

[0007] In order to solve the above technical problems and achieve the above purposes, the present application provides the following technical solutions.

[0008] In a first aspect, the present application provides a recovery stability analysis method for fault block reservoirs, which comprises the following steps:

[0009] S1. Selecting a well area in an oilfield using injection-production coupling mining to collect data;

[0010] S2. Select the influencing factors of the current injection-production contradiction, and carry out flat plate sand filling experiment and numerical simulation;

[0011] S3. Propose an improvement scheme through experimental results, and compare the change of recovery rate before and after the implementation of the improvement scheme by means of the reliability analysis model;

[0012] S4. Make recovery rate prediction through the DGM(1, 1) model.

[0013] In an optional implementation, the data in the step S1 includes one or more of the following data: oil layer burial depth, stratum dip angle, reservoir average porosity, permeability, ground crude oil density, ground crude oil viscosity, original stratum pressure, saturation pressure, pressure coefficient.

[0014] In an optional implementation, the well area in the step S1 is a complex fault block reservoir with medium porosity and medium permeability.

[0015] In an optional implementation, the influencing factors in the step S2 include permeability, viscosity and fluid production rate.

[0016] In an optional implementation, the flat plate sand filling experiment in the step S2 includes the following steps:

[0017] 1) Perform the experiment under different permeabilities: respectively measure the recovery rate under different permeabilities of the flat plate physical model experiment bench, and draw the recovery rate-time change curve of each well;

[0018] 2) Perform the experiment under different crude oil viscosities: respectively measure the recovery rate under different viscosities, and draw the recovery rate-time change curve of each well;

[0019] 3) Perform the experiment under different fluid production rates: respectively measure the recovery rate under different fluid production rates, and draw the recovery rate-time change curve of each well.

[0020] In an optional implementation, the numerical simulation in the step S2 includes the following steps: use the petrel software to perform the simulation of the influence of different crude oil viscosities, different fluid production rates and different permeabilities on the recovery rate, and obtain the curve graph and regression equation of the influence law of different factors on water drive oil.

[0021] In an optional implementation, the engineering verification in the step S3 includes the following steps:

[0022] i. Data screening: according to the requirement of the distribution analysis arbitrary deletion function module, input the sample data containing the start time in the initial variable and the end variable; the start time in this column depends on the deletion mode of the data, and the column containing the failure mode is indicated in the failure mode column;

[0023] ii. Data fitting and goodness-of-fit test: goodness-of-fit test of the full data set is performed using maximum likelihood, chi-square test and least square function;

[0024] iii. Hypothesis testing and reliability calculation: the failure distribution function F(t) and reliability function R(t) of the two-parameter Weibull distribution are calculated:

[0025]

[0026]

[0027] wherein: η is a scale parameter; β is a shape parameter; γ is a threshold parameter; t represents a location parameter constant value.

[0028] In an optional embodiment, the recovery factor prediction in step S4 is calculated using the grey system DGM(1,1) model in a function framework that changes over time.

[0029] In an optional embodiment, the recovery factor prediction by the DGM(1,1) model comprises the following steps:

[0030] ① Let the time series X (0) have n observations, X (0) = {x (0) (1), x (0) (2), x (0) (3)...x (0) (n)}, generate a new sequence by one-time accumulation:

[0031] X (1) = {x (1) (1), x (1) (2), x (1) (3)...x (1) (n)}

[0032] ② Generate the DGM(1,1) model differential equation:

[0033] x (1) (k+1) = β1x (1) (k) + β2

[0034] wherein

[0035]

[0036] ③ Calculate the parameter values using the least square method:

[0037] If X is the parameter column, and X

[0038]

[0039] Then the discrete grey prediction model X (1) (k+1)=β1,X (1) The least squares estimation parameter sequence of (k)+β2 satisfies:

[0040]

[0041] ④ Take X (1) (1) = X (0) (1), then the recursive function is:

[0042] or

[0043] ⑤ Discrete solution prediction model:

[0044]

[0045] ⑥ Residual test, calculated according to the prediction model. And Cumulative subtraction generation Then calculate the original sequence X. (0) (i) with Absolute error sequence and relative error sequence:

[0046]

[0047] In the formula: i = 1, 2, ..., n

[0048]

[0049] In the formula: i = 1, 2, ..., n.

[0050] Secondly, the present invention provides the application of the analytical method described in any of the foregoing embodiments in complex fault-block reservoirs.

[0051] As can be seen from the above technical solutions, the present invention has the following beneficial effects:

[0052] This invention addresses the shortcomings in engineering verification based on traditional plate-filling experiments and numerical simulations. It proposes a mathematical analysis method for scheme analogy with strong matching and high responsiveness. This solves problems such as long verification cycles and weak parameter comparability in reservoir development, and improves the systematic work steps of "experiment → data simulation → scheme proposal → mathematical verification". The method described in this invention can explore the main controlling factors affecting recovery rate without changing the well network deployment, and propose and verify development adjustment schemes, providing engineering verification methods for the schemes. It also helps to explore the intrinsic causes of injection-production coupling disorder in the study area. Attached Figure Description

[0053] In order to more clearly illustrate the technical solutions in the specific embodiments of the present application or the prior art, the drawings required to be used in the description of the specific embodiments or the prior art will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0054] Figure 1 The top structure map of Qian 41 in the west slope zone of the Tan Kou protrusion in Example 1 of the present application;

[0055] Figure 2 The experimental logic diagram of Example 1 of the present application;

[0056] Figure 3 The experimental device and flow chart of Example 1 of the present application, wherein a is the experimental device, b is the inverted nine-spot well pattern well layout, and c is the displacement simulation flow chart;

[0057] Figure 4 The planar grid map (plane) of the flat plate mechanism model in Example 1 of the present application;

[0058] Figure 5 The profile grid map (3D) of the flat plate mechanism model in Example 1 of the present application;

[0059] Figure 6 The relationship curve of the recovery rate of each well with time under the premise of permeability 1000 mD in Example 2 of the present application;

[0060] Figure 7 The relationship curve of the recovery rate of each well with time under the premise of permeability 2000 mD in Example 2 of the present application;

[0061] Figure 8 The relationship curve of the recovery rate of each well with time under the premise of viscosity 90 mPa·s in Example 2 of the present application;

[0062] Figure 9 The relationship curve of the recovery rate of each well with time under the premise of viscosity 180 mPa·s in Example 2 of the present application;

[0063] Figure 10 The relationship curve of the recovery rate of each well with time under the premise of viscosity 360 mPa·s in Example 2 of the present application;

[0064] Figure 11 The relationship curve of the recovery rate of each well with time under the premise of liquid production rate 1 mL / min in Example 2 of the present application;

[0065] Figure 12The recovery rate of each well with the change of time under the condition of the liquid production rate of 2 mL / min in the embodiment 2 of the present application is shown in the following figure:

[0066] Figure 13 The recovery rate of each well with the change of time under the condition of the liquid production rate of 5 mL / min in the embodiment 2 of the present application is shown in the following figure:

[0067] Figure 14 The curve of the influence of the viscosity of crude oil on the water flooding effect in the embodiment 3 of the present application is shown in the following figure:

[0068] Figure 15 The curve of the influence of the liquid production rate on the water flooding effect in the embodiment 3 of the present application is shown in the following figure:

[0069] Figure 16 The curve of the influence of the model permeability on the water flooding effect in the embodiment 3 of the present application is shown in the following figure:

[0070] Figure 17 The exponential distribution probability graph of the scheme comparison with the confidence degree of 95% in the embodiment 4 of the present application is shown in the following figure. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme in the embodiments of the present application will be described clearly and completely below in combination with the drawings of the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments of the present application.

[0072] Some embodiments of the present application will be described in detail below in combination with the drawings. In the case of no conflict, the embodiments described below and the features in the embodiments can be combined with each other.

[0073] The following embodiments take the X well area of the complex fault block reservoir in Tankou as the research object, systematically describe the water injection development status of the well area, and quantitatively research by means of the flat plate sand filling experiment and numerical simulation, so as to explore the main control factors affecting the recovery rate without changing the well pattern deployment, propose and verify the development adjustment scheme, give the engineering verification method of the scheme, and help explore the internal reasons of the injection-production coupling disorder in the research area.

[0074] The general situation of the research problem in the embodiments of the present application is as follows: the first step is to determine the X well area of Tankou as the research object, and preset the experimental direction through the existing development data; the second step is to determine the permeability, viscosity and liquid production rate as the main factors affecting the current injection-production contradiction, and to carry out the flat plate sand filling experiment and numerical simulation research; the third step is to propose the development adjustment scheme through the experimental results, and to carry out engineering verification by means of the reliability analysis model; the fourth step is to select the data of the last several observation points for engineering verification for prediction.

[0075] Embodiment 1

[0076] Data were collected from well areas in the oilfield that employ injection-production coupling, and a three-factor flat sand-filling experiment was conducted.

[0077] 1.1. Overview

[0078] Tankou Oilfield is a typical complex fault-block reservoir type in eastern China, and its regional structure belongs to the Zhongtan Fault Zone in the northern part of the Qianjiang Depression of the Jianghan Basin (e.g., Figure 1 As shown in the image, Well X is located in the eastern part of the Tankou Oilfield. It contains multiple oil-bearing formations, mainly concentrated in the Qian 41, 41 Lower, 40, 40 Lower, 42, and 43 oil groups. The oil layers are buried at depths ranging from 660m to 2945m, with dip angles of 45° to 60°. The average reservoir porosity is 23.8%, and the permeability is 534.2 × 10⁻⁶. -3 μm 2 The density of crude oil at ground level is 0.863-0.920 g / cm³. 3 The surface crude oil viscosity ranges from 21.2 to 63.8 mPa·s, the original formation pressure ranges from 12.6 to 29.5 MPa, the saturation pressure ranges from 0.70 to 2.2 MPa, and the pressure coefficient is 1.12. It belongs to a complex fault-block reservoir with medium porosity and medium permeability. Its main geological characteristics are: extremely well-developed low-sequence faults, small fault blocks, large differences in the thickness and permeability of oil-bearing sublayers, and strong interlayer heterogeneity; complex oil-water relationships and large differences in natural energy. The area generally has a reverse nine-point well network, but due to the confidentiality of geological data, the well locations are not marked.

[0079] The X-well area of ​​the Tankou Oilfield has a total of 38 oil wells and 7 water wells, with a daily oil production of 77.6 tons, an average daily production of 2.0 tons per well, a comprehensive water cut of 53.0%, and a daily water injection of 139 m³. 3 To date, the cumulative oil production is 29.7 × 10⁻⁶. 4 The geological reserves have an oil recovery rate of 0.79% (based on recalculated reserves), which is moderately low; the current injection-production ratio is 0.85, and the cumulative water production is 43 × 10⁻⁶. 4 m 3 Cumulative water injection: 16.8 × 10 4 m 3 The cumulative injection ratio is 0.28, indicating severe formation deficit and low geological reserve recovery (8.9%). Currently, the rated recovery rate is 15.04%, suggesting significant potential for development and production adjustments. However, due to the complexity and unique characteristics of its geological structure, fluid properties, and oil-water system, it is necessary to conduct water-drive flat plate model experiments on complex fault-block reservoirs. This will allow for the development and adjustment of old oilfields where engineering must conform to geology, and geology must also consider engineering principles, optimizing current development parameters to achieve stable production. The logic of this study is as follows (see...). Figure 2 (As shown).

[0080] 1.2 Flat Plate Sand Filling Experiment

[0081] Three-dimensional physical simulation can better approach real formation conditions, reflect the vertical and horizontal heterogeneity changes of reservoirs in the process of water flooding, and realize the adjustment of well pattern, so as to achieve the purpose of simulating reservoir development. According to the similarity criterion, this experiment carries out three-dimensional physical simulation water flooding sand filling experiment by considering different permeability, crude oil viscosity, liquid production rate and other influencing factors, explores the main controlling factors affecting recovery, and provides a basis for improving the reservoir development effect of X well area in Tankou Oilfield. Since the well pattern deployment of the study area has been determined, we do not discuss the influence of well spacing density here, and set the experimental conditions to be the most close to the field conditions of the inverted nine-spot well pattern. The bound water saturation fitting formula of the work area is The experimental device and flow chart are shown in Figure 3

[0082] Materials and equipment: constant speed displacement pump (flow rate range: 0.001-200 mL / min; pressure resistance: 5 MPa), piston type intermediate container, pressure gauge, flat plate model (size: 30x30x5 cm, by adjusting the well pattern arrangement on the model body, various well pattern combination modes such as four-point well pattern, five-point well pattern, seven-point well pattern and nine-point well pattern can be realized), measuring cylinder, six-way valve, etc.

[0083] Experimental samples: formation water sample, formation crude oil sample, quartz sand of different mesh.

[0084] Experimental steps under different permeability: ① Assemble the experimental bench, and use quartz sand of different mesh to fill the flat plate model with permeability of x and well spacing of 12 cm. ② Inject the filled model with formation water, then use the saturated 180 mPa·s formation crude oil to simulate the inverted nine-spot well pattern, and carry out water flooding experiment at a flow rate of 2 mL / min. ③ End the experiment when the comprehensive water cut of the production well reaches 98%. ④ Measure the recovery rate under the conditions of permeability of 1000 mD and 2000 mD of the flat plate model experimental bench, and draw the recovery rate-time curve of each well.

[0085] Experimental steps under different crude oil viscosity: ① Assemble the experimental bench, and use quartz sand of different mesh to fill the flat plate model with permeability of 2000 mD and well spacing of 12 cm. ② First, inject the filled model with formation water, then displace the formation water with formation crude oil with viscosity of x, and finally adjust the bound water saturation to about 39% by using the fitting formula. ③ Connect the injection pipeline by using the inverted nine-spot well pattern connection method, and inject the formation water at a constant speed of 2 mL / min. ④ End the experiment when the comprehensive water cut of the production well reaches 98%. ⑤ Measure the recovery rate under the conditions of viscosity of 90 mPa·s, 180 mPa·s and 360 mPa·s, and draw the recovery rate-time curve of each well.

[0086] ​Experimental steps under different liquid withdrawal rates: ① Assemble the experimental bench, fill the plate model with quartz sand of different mesh numbers to have a permeability of 2000 mD and a well spacing of 12 cm. ② Inject the prepared model with formation water, then use saturated 180 mPa·s formation crude oil to simulate the inverted nine-spot well pattern. Perform water flooding experiment at the flow rate of x. ③ End the experiment when the comprehensive water cut of the production well reaches 98%. ④ Measure the recovery rate under three conditions of liquid withdrawal rates of 1 mL / min, 2 mL / min and 5 mL / min, and draw the recovery rate-time curve of each well.

[0087] Numerical simulation method: Use petrel software to simulate the influence of different crude oil viscosities, different liquid withdrawal rates and different permeabilities on recovery rate. Obtain the curve graph and regression equation of the influence law of different factors on water flooding. The mechanism model adopts an angle point grid system, the grid on the plane is divided into 30 in I direction and 30 in J direction, the plane grid size is 1 cm x 1 cm, the vertical direction is divided into 1 simulation layer according to the variable depth, the single layer grid thickness is 5 cm, and the total number of nodes of the model is 30 x 30 x 1 = 900. There are 16 wells in the model, which are set to nine-point method through well switching. The crude oil viscosity is set to 90, 180, 270 and 360 mPa·s. The permeability is set to 500, 1000, 1500, 2000, 2500 and 3000 mD. The liquid withdrawal rate is set to 1, 2, 3, 4, 5 mL / min (see Figure 4 and Figure 5 ).

[0088] Example 2

[0089] Numerical analysis

[0090] 2.1 Permeability influence

[0091] Through experimental determination, under the premise of permeability of 1000 mD, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 953.3 mL. The cumulative oil production is 406.5 mL, and the final recovery rate is calculated to be 42.64%; under the premise of permeability of 2000 mD, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 1103.35 mL. The cumulative oil production is 489.9 mL, and the final recovery rate is calculated to be 44.40%. The recovery rate-time curve is drawn respectively (see Figure 6 and Figure 7 ).

[0092] 2.2 Viscosity influence

[0093] Through experimental determination, under the premise of crude oil viscosity of 90 mPa·s, the experimental simulation pore volume is 1830 cm 3, the total oil content is 1108 mL. The cumulative oil production is 516.90 mL, and the calculated ultimate recovery is 46.65%; under the premise of crude oil viscosity of 180 mPa·s, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 1103.35 mL. The cumulative oil production is 489.9 mL, and the calculated ultimate recovery is 44.40%; under the premise of crude oil viscosity of 360 mPa·s, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 1094.45 mL. The cumulative oil production is 469.30 mL, and the calculated ultimate recovery is 42.88%. The recovery rate-time curve is drawn respectively (see Figure 8 、 Figure 9 and Figure 10 ).

[0094] 2.3 Effect of liquid production rate

[0095] Through experimental determination, under the premise of liquid production rate of 1 mL / min, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 1100 mL. The cumulative oil production is 463.90 mL, and the calculated ultimate recovery is 42.17%; under the premise of liquid production rate of 2 mL / min, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 1103.35 mL. The cumulative oil production is 489.9 mL, and the calculated ultimate recovery is 44.40%; under the premise of liquid production rate of 5 mL / min, the experimental simulation pore volume is 1830 cm 3 , the total oil content is 1102 mL. The cumulative oil production is 528.1 mL, and the calculated ultimate recovery is 47.92%. The recovery rate-time curve is drawn respectively (see Figure 11 、 Figure 12 and Figure 13 ).

[0096] 2.4 Summary of liquid production and oil production data

[0097] Through the determination data screening of different experimental categories, the last group of single well production data at the end of water flooding experiment is compared (Table 1). It can be seen that the liquid production of the oil production well near the injection well is high, but the oil production is low. However, the average oil production, liquid production and recovery rate under different conditions still need to be compared comprehensively.

[0098] Table 1 Water flooding experiment results

[0099]

[0100]

[0101] Example 3

[0102] 3.1 Numerical Simulation

[0103] Numerical simulation can further extend the application scope of the conclusions from plate experiments. Data fitting curves can yield reliable relationships, which can be easily accessed during the mining process.

[0104] Comparison of numerical simulation and experimental results regarding viscosity-related factors shows that the inverted recovery rate is basically close to the experimental recovery rate, meeting the research requirements. The recovery rate decreases exponentially with increasing crude oil viscosity. The numerical simulation inversion fitting equation is: y = 65.451x -0.0695 ,like Figure 14 As shown.

[0105] Comparison of numerical simulation and experimental results regarding the factors influencing fluid collection rate shows that the inverted recovery rate is basically close to the experimental recovery rate, meeting the research requirements. With increasing fluid collection rate, the recovery rate gradually increases, exhibiting a logarithmic upward trend. The numerical simulation inversion fitting equation is: y = 3.0698ln(x) + 43.503, as... Figure 15 As shown.

[0106] Comparison of numerical simulation and experimental results regarding the factors influencing permeability shows that the inverted recovery rate is basically close to the experimental recovery rate, meeting the research requirements. With increasing model permeability, the recovery rate gradually increases, exhibiting a power-law upward trend. The numerical simulation inversion fitting equation is: y = 18.898x 0.1173 ,like Figure 16 As shown.

[0107] In summary, the numerical simulation inversion fitting curve basically matches the experimental data curve of the flat sand filling test, with only a small difference in R-value. Therefore, the formula is valid.

[0108] 3.2 Analysis of Numerical Simulation Results

[0109] The following information can be obtained through the flat plate sand filling experiment.

[0110] 3.2.1 Under the premise of different permeability, crude oil viscosity, and fluid production rate, wells #1, #3, #5, and #7, which are farther from injection well #0, have higher final recovery rates than wells #2, #4, #6, and #8, which are closer. The recovery rate ranking within the two groups is: #1 > #3 > #7 > #5; #2 > #8 > #6 > #4. Experimental measurements of the three influencing factors show that 3000 min is the optimal stable production period. When the fluid production rate reaches 1 mL / min or the crude oil viscosity is 360 mPa·s, the stable production period can be extended to 5000 min.

[0111] 3.2.2, The effect of permeability on the experiment ( Figure 6 , 7) under 1000 mD and 2000 mD, the development curves of the two groups began to separate at 119 min and 132 min. The curves showed a rapid upward trend from 0 to 250 min. The initial recovery rate under 1000 mD increased more than that under 2000 mD, which may be because the flow rate of 2 mL / min can fully overcome the capillary force in the pore at the initial stage of injection and production, and the hydraulic shock in the environment of smaller permeability can promote oil displacement in a limited time.

[0112] 3.2.3, viscosity influence experiment in the experiment Figure 8 , 9 , 10) under 90 mPa·s and 180 mPa·s, the development curves of the two groups began to separate at about 100 min. The development curves of the two groups began to separate at 700 min under 360 mPa·s. Under 90 mPa·s, the curve of 1# well showed a clear stepwise upward trend at 1600-1800 min. The curves of 3#, 5#, and 7# wells also showed similar trends. This may be due to the influence of the plastic state of crude oil. The viscous flow temperature of crude oil under 90 mPa·s is more active, which leads to the conversion of the viscoelastic transition zone of the material properties and the combined effect of hydraulic flushing, resulting in fluctuation of recovery rate. Under the premise of high viscosity of 360 mPa·s, the molecular force of oil droplets is greater, and after aggregation, it is more easily displaced and produced by water.

[0113] 3.2.4, in the experiment of liquid production rate influence Figure 11 , 12 , 13) under 1 mL / min, 2 mL / min, and 5 mL / min, the development curves of the two groups began to separate at 50 min, 40 min, and 30 min, respectively. Under 5 mL / min, the development curves of each well showed a balanced separation trend, and the initial curve had the largest slope. This may be because the displacement force generated at high flow rate is stronger, leading to more complete displacement of crude oil in the pore. 5, under the same permeability, well pattern, and liquid production rate, the recovery rate decreases with the increase of crude oil viscosity, showing a power decrease trend; under the same permeability, crude oil viscosity, and well pattern, the recovery rate increases with the increase of liquid production rate, showing a power increase trend; under the same crude oil viscosity, well pattern, and liquid production rate, the recovery rate of 1000 mD is 1.76% lower than that of the 2000 mD plate model.

[0114] Numerical simulation can make up for the lack of quantitative determination of the plate sand filling experiment, and the following information is obtained. 1, under the influence of viscosity, the recovery rate decreases with the increase of crude oil viscosity, showing a power decrease trend. The fitting relationship is: y = 65.451x -0.06952、Under the premise of liquid production rate, with the increase of liquid production rate, the recovery rate gradually increases, showing a logarithmic upward trend. The fitting relationship is: y = 3.0698ln(x) + 43.503. 3、Under the influence of permeability, with the increase of model permeability, the recovery rate gradually increases, showing a power rising trend. The fitting relationship is: y = 18.898x 0.1173 .

[0115] 3.3 The following engineering guidelines can be drawn:

[0116] 3.3.1 Under the premise of different categories of influencing factors, 2000mD permeability, 90mPa·s viscosity, 5mL / min liquid production rate can obtain the highest recovery rate. Therefore, the water injection rate is adjusted to 5mL / min.

[0117] 3.3.2 The liquid production rate of the oil well near the water injection well is high, but the oil production rate is low. This may be due to the disorder of vertical injection-production coupling, so the ground switch and downhole separator should be used to roughly divide the multi-layer reservoir into independent injection-production units. According to the physical properties of sandstone reservoir, the layer with good physical properties is injected first, and the layer with poor physical properties is produced. The injection-production coupling mechanism is used to adjust the injection rate, timing, cycle and oil production pressure difference of the upper and lower layers for production.

[0118] 3.3.3 Under the premise of 2000mD permeability, 180mPa·s viscosity and 2mL / min liquid production rate, the average liquid production rate, average oil production rate and recovery rate are equal. This may be due to the disorder of horizontal injection-production coupling, so the auxiliary alternative scheme of not moving the pipe string alternately should be adopted. During production, the formation fluid enters the switch body from the liquid hole, flows into the casing through the check valve, and is lifted to the ground by the pump.

[0119] 3.3.4 During the production process, the fitting formula obtained by numerical simulation can be used to compare the rationality of the parameters in this stage to identify the problems in this stage and predict the production trend in the next production period. Realize real-time adjustment of development plan.

[0120] 3.5 Based on the experimental results analysis above, combined with the current development situation of the research area, the development adjustment scheme can be further proposed:

[0121] 3.5.1 Due to the poor effect of early water drive, the liquid production rate is adjusted to 0.5mL / min. This may lead to a decrease in water injection swept area, but the degree of advance will increase.

[0122] 3.5.2, Due to the high viscosity of crude oil in the study area, it is decided to take the measure of mixing water and adding chemicals. Specifically, according to the size of single well liquid production, 2% of wax remover and viscosity reducer are added, and the dosing system is to add once every 2 days, with 2% of the dose added every week.

[0123] 3.5.3, Based on the current situation of water flooding and injection-production contradiction in the longitudinal direction of the study area, it is necessary to take measures to block high permeability production layers and combine low permeability production layers to alleviate the imbalance of injection and production caused by reservoir heterogeneity.

[0124] 3.5.4, Based on the tongue-in situation in the lateral direction of the study area, it is necessary to take the measures of 0-3h pulse type large displacement dosing (3m 3 / h) injection, and 6h shut-in reaction, followed by speed control oil production.

[0125] 3.5.5, Based on the characteristics of X well area in Tan Kou, take chloride ion monitoring test system to calculate the degree of water flooded tongue-in, dynamically adjust the water injection parameters. In addition, use deep pumping matching technology to meet the needs of oil well production, take coupling injection-production as the main means of water injection adjustment, increase the swept area of remaining oil, and finally achieve the purpose of improving recovery.

[0126] Example 4

[0127] 4.1 Engineering verification

[0128] The conventional oil production plan verification usually focuses on recovery rate and oil well maintenance frequency, and observes for 1 year or more, with long verification period and high economic cost. Therefore, in order to verify the feasibility of the adjustment plan, accelerated experiment is needed. Here we select the reliability analysis method described in quality engineering, focus on the internal and external causes of weak links, find out the rules, and give improvement measures and the impact on system reliability after improvement.

[0129] The engineering verification method is to use minitab software to perform reliability analysis with reliability as the judgment index, and the model steps are as follows:

[0130] ① Data screening

[0131] According to the requirement of distribution analysis arbitrary deletion function module, in the initial variable and the end variable, at most 50 columns of sample data containing start time are input. The start time in this column depends on the deletion mode of data. The column containing failure mode is input. To indicate the right-censored observation value in the failure mode column.

[0132] ② Data fitting and goodness-of-fit test

[0133] The failure mode data column will be subject to Weibull distribution model, exponential distribution model, extreme value distribution model and normal distribution model. Usually use maximum likelihood, chi-square test and least square function to test the goodness of fit of the whole data column. The chi-square test formula is:

[0134]

[0135] Where: χ 2 is the chi-square test; f oi is the observed frequency of the i-th group; f ei is the expected frequency of the i-th group; k is the number of data groups.

[0136] For exponential distribution, the maximum likelihood estimate of parameter λ is:

[0137]

[0138] In this formula, t represents the scale

[0139] For the maximum likelihood estimate of μ and σ 2 in normal distribution:

[0140]

[0141]

[0142] In this formula, x i and x respectively represent the parameter position

[0143] For lognormal distribution, the maximum likelihood estimate of μ and σ 2 can be obtained according to the parameter estimate of normal distribution, respectively:

[0144]

[0145]

[0146] For two-parameter Weibull distribution, the maximum likelihood estimate of its parameters is solved by the following transcendental equation:

[0147]

[0148] Among the preliminary distribution models, the parameters to be estimated are the parameter λ of exponential distribution, the mean μ and standard deviation σ of normal distribution, the log mean μ and log standard deviation σ of lognormal distribution, and the scale parameter η and shape parameter β of two-parameter Weibull distribution.

[0149] ③ Hypothesis testing and reliability calculation

[0150] The two-parameter Weibull distribution is usually preferred in the calculation. So taking the Weibull distribution as an example, write the failure distribution function F(t) and the reliability function R(t) of the two-parameter Weibull distribution:

[0151]

[0152]

[0153] In the formula: η is the scale parameter; β is the shape parameter; γ is the threshold parameter; t represents the position parameter value.

[0154] According to the statistical data of the recovery rate of the existing 38 oil wells in the study area, combined with the conclusions of the plate sand filling experiment, the moderate recovery rate under the premise of 2000 mD permeability, 180 mPa·s viscosity and 2 mL / min liquid production rate is obtained as the average value. Then compare the actual calibrated recovery rate of 15.04% in the study area. It is determined that there is an error between the plate sand filling experiment and the actual working condition, so the error is weakened by reducing it by 3 times, and the threshold value of 14.8% is obtained. It is determined that the single well recovery rate is less than 4.8% is a failure, and the failure frequency of every 3 weeks interval of 53 weeks before and after the implementation of the adjustment scheme in the study area X well is counted (Table 2), and the minitab software is used for reliability analysis under the premise of arbitrary deletion. Compare the reliability before and after the implementation of the scheme.

[0155] Table 2 Failure sample data statistics table

[0156] Start (old), week End (old), week Frequency (old), times Start (new), week End (new), week Frequency (new), times 3 0 3 1 3 6 1 3 6 0 6 9 23 6 9 3 9 12 17 9 12 11 12 15 20 12 15 12 15 18 3 15 18 8 18 21 10 18 21 9 21 24 19 21 24 5 24 27 22 24 27 4 27 30 2 27 30 7 30 33 7 30 33 8 33 36 9 33 36 0 36 39 16 36 39 2 39 41 3 39 41 3 41 44 23 41 44 4 44 47 11 44 47 1 47 50 4 47 50 12 50 53 2 50 53 9

[0157] The goodness-of-fit test of the entire data column is carried out by using the maximum likelihood function, and it is concluded that the data column obeys the exponential distribution. Further parameter distribution analysis is carried out to obtain the exponential distribution probability graph of the scheme comparison with a confidence level of 95% (see Figure 17 ).

[0158] Table 3 Exponential distribution percentile before and after the implementation of the scheme

[0159]

[0160]

[0161] From the above operation results, the mean estimation value before the implementation of the scheme is 25.7081, and the life of 99% of the oil wells is greater than 0.258375 under the premise of the preset recovery threshold; the mean estimation value after the implementation of the scheme is 27.5591, and the life of 99% of the oil wells is greater than 0.276978 under the premise of the preset recovery threshold. Comparison shows that the comprehensive oil well life after the implementation of the scheme is greater than before. Although the life technical data is not significantly high, the engineering practice after the implementation of the composite new scheme can only slightly improve the recovery rate. It proves that the scheme is effective.

[0162] 4.2 Development scheme prediction

[0163] After observing the implementation of the new scheme, the recovery rate data of the terminal observation point is combined with the DGM(1, 1) model to calculate at least 4 groups of data and meet the requirement that the modeling data cannot be 0. Select 2, 3, 4, 1, 12, 9 data strings, and artificially eliminate the large fluctuation point 12.

[0164] The prediction method selects the gray system DGM(1, 1) model, which is specifically operated under a function framework that changes with time. This model requires at least 4 groups of time series data. First, weaken the random disturbance factor, perform original data accumulation conversion, then operate the exponential fitting curve of the least square method to realize function prediction, and finally use residual detection to detect the result reliability. The model steps are as follows:

[0165] ① Let the time series X (0) have n observation values, X (0) = {x (0) (1), x (0) (2), x (0) (3)...x (0) (n)}, generate a new sequence by once accumulation:

[0166] X (1) = {x (1) (1), x (1) (2), x (1) (3)...x (1) (n)} (10)

[0167] ② Generate the differential equation of the DGM(1, 1) model:

[0168] x (1) (k+1) = β1x (1) (k) + β2 (11)

[0169] Where

[0170]

[0171] β1 and β2 represent the development coefficient and the grey action amount respectively

[0172] ③ Calculate the parameter value by using the least square method:

[0173] If is the parameter sequence, and

[0174]

[0175] The discrete grey prediction model X (1) (k+1) = β1, X (1) (k) + β2 is the least square estimation parameter sequence of:

[0176]

[0177] In the formula, T represents the transpose, and the expressions of B and Y are shown in formula 13.

[0178] ④ Take X (1) (1) = X (0) (1), then the recursive function is:

[0179] Or

[0180] In the formula, k represents the time recursive function

[0181] ⑤ Discrete solution of the prediction model:

[0182]

[0183] ⑥ Residual test, calculate according to the prediction model and cumulative decrease is generated Then calculate the absolute error sequence and the relative error sequence of X (0) (i) and :

[0184]

[0185] In the formula, i = 1, 2,..., n.

[0186]

[0187] In the formula, i = 1, 2,..., n.

[0188] The calculation process is as follows:

[0189] ① Original sequence initialization: 2, 3, 4, 1, 9.

[0190] 2.0000, 5.0000, 9.0000, 10.0000, 19.0000.

[0191] 3. Calculate the gray model development coefficient β1 and the gray action β2: β1 = 1.3537; β2 = 1.9512.

[0192] 4. Simulated value calculation: 2.0000, 2.6585, 3.5988, 4.8715, 6.5943.

[0193] 5. Residual error: 21.0532; average relative error: 34.9868%; further 2-step prediction: 8.93, 12.08.

[0194] In summary, the average relative error is low, and the model is reliable. The prediction conclusion (the recovery rate of the next two observation points is 8.93, 12.08) can be compared with the observation data at the next time node to verify the perfection of the overall workflow. Timely warning of the failure node in the entire observation process facilitates the systematic adjustment of the development plan.

[0195] 4.3 Conclusion

[0196] From the above examples, the examples of the present application take the X well area of Tankou Oilfield as an example, and carry out research by selecting a flat plate sand filling experiment at the injection-production coupling link, reflect the heterogeneity change in the vertical and horizontal directions of the oil reservoir in the water drive oil process, simulate the reservoir development, and obtain the recovery rate relative time change law under the influence factors of different permeability, crude oil viscosity and liquid production rate. The accuracy of the experiment is verified by numerical simulation, which makes up for the narrow reference range of the results caused by the limited boundary of the experiment. The development plan details of the X well area of Tankou Oilfield are optimized in engineering.

[0197] From the examples, we find that the liquid production of the oil production well near the injection well is high, but the oil production is low. This may be caused by the vertical injection-production coupling disorder of the oil reservoir. The average liquid production, average oil production and recovery rate under the premise of 2000mD permeability, 180mPa·s viscosity and 2mL / min liquid production rate are equal, which may be caused by the horizontal injection-production coupling disorder of the oil reservoir. The best stable production time obtained by the experiment is 3000min, and the maximum recovery rate obtained under the premise of 5mL / min liquid production rate is 47.92%. In terms of geology, when the premise conditions are the same, the recovery rate of the 1000mD permeability is 1.76% lower than that of the 2000mD permeability.

[0198] Compared with the prior art, the method has higher pertinence, lower experimental cost and shorter research process. Compared with the determination of the development influencing factors such as the water injection mode, water injection time, water injection cycle, injection speed and injection viscosity in the prior art, the method focuses more on the combination of theory and engineering practice, and the numerical simulation makes up for the deficiency in the real experiment, and extends the application degree of the results. The next step of revising the development plan cannot simply take the recovery ratio as the standard, but needs to comprehensively consider the influence of the two factors of the crude oil properties and the stable production period. The development adjustment work idea of the old oilfield is constructed, which is engineering compliant to geology and geology giving consideration to engineering.

[0199] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for analyzing the recovery stability of fault-block oil reservoirs, characterized in that, Includes the following steps: S1. Select well areas in the oilfield that use injection-production coupling for data collection; S2. Select the influencing factors of the current injection-production conflict, and conduct flat plate sand filling experiments and numerical simulations; S3. Based on the experimental results, propose an improved scheme and use a reliability analysis model to compare the changes in recovery rate before and after the implementation of the improved scheme; S4. Predict the recovery rate using the DGM(1,1) model.

2. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The data in step S1 includes one or more of the following: oil layer burial depth, formation dip angle, average reservoir porosity, permeability, surface crude oil density, surface crude oil viscosity, original formation pressure, saturation pressure, and pressure coefficient.

3. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The well area in step S1 is a complex fault-block oil reservoir with medium porosity and medium permeability.

4. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The influencing factors in step S2 include permeability, viscosity, and fluid collection rate.

5. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The plate sand-filling experiment in step S2 includes the following steps: 1) Conduct experiments at different permeabilities: Measure the recovery rate of each well at different permeabilities on the flat plate physical model experimental platform, and plot the curves showing the relationship between the recovery rate of each well and time. 2) Conduct experiments at different crude oil viscosities: Measure the recovery rate under different viscosity conditions and plot the curves showing the relationship between the recovery rate of each well and time; 3) Use at different fluid production rates: Measure the recovery rate under different fluid production rates and plot the curves showing the relationship between the recovery rate of each well and time.

6. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The numerical simulation in step S2 includes the following steps: using petrel software, performing simulations of the effects of different crude oil viscosities, different fluid production rates, and different permeabilities on the recovery rate, and obtaining curves and regression equations showing the influence of different factors on waterflooding.

7. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The engineering verification in step S3 includes the following steps: i. Data Filtering: In accordance with the requirements of the arbitrary censoring function module of distribution analysis, input sample data containing the start time in the initial variable and the ending variable; the start time in this column depends on the data censoring method, and input a column containing failure modes, to represent the right-censored observations in the failure mode column; ii. Data Fitting and Goodness-of-Fit Test: Use maximum likelihood, chi-square test and least squares function to test the goodness-of-fit of the entire data series; iii. Hypothesis Testing and Reliability Calculation: Calculate the failure distribution function F(t) and reliability function R(t) of the two-parameter Weibull distribution: In the formula: η is the scale parameter; β is the shape parameter; γ is the threshold parameter; t represents the position parameter constant.

8. The method for recovery stability analysis of fault-block reservoirs according to claim 1, characterized in that, The recovery rate prediction in step S4 is performed using the grey system DGM(1,1) model within a time-varying functional framework.

9. The method for recovery stability analysis of fault-block oil reservoirs according to claim 1, characterized in that, The recovery rate prediction using the DGM(1,1) model includes the following steps: ① Let time series X (0) There are n observations, X (0) ={x (0) (1),x (0) (2),x (0) (3)...x (0) A new sequence is generated by accumulating (n)}: X (1) ={x (1) (1),x (1) (2),x (1) (3)...x (1) (n)} ② Generate the differential equations of the DGM(1,1) model: x (1) (k+1)=β1x (1) (k)+β2 in In the formula, β1 and β2 represent the development coefficient and the grey effect, respectively; Preferably, β1 = 1.3537, β2 = 1.9512; ③ Calculate the parameter values ​​using the least squares method: like For parameter columns, and Then the discrete grey prediction model X (1) (k+1)=β1,X (1) The least squares estimation parameter sequence of (k)+β2 satisfies: In the formula, T represents transpose; ④ Take X (1) (1) = X (0) (1), then the recursive function is: or In the formula, k represents the time recursion function; ⑤ Discrete solution prediction model: ⑥ Residual test, calculated according to the prediction model. And Cumulative subtraction generation Then calculate the original sequence X. (0) (i) with Absolute error sequence and relative error sequence: In the formula: i = 1, 2, ..., n In the formula: i = 1, 2, ..., n.

10. The application of the recovery stability analysis method for fault-block reservoirs as described in any one of claims 1-9 in complex fault-block reservoirs.

Images