A polymer flooding well infill optimization method based on improved Archimedes algorithm

Abstract

This invention relates to a polymer flooding well location infill optimization method based on an improved Archimedes algorithm, belonging to the field of reservoir development technology. The steps include: establishing a polymer flooding numerical simulation model; determining the number and location range of infill wells for production and injection wells; establishing a polymer flooding well location infill calculation model; using the Halton sequence-improved Archimedes algorithm to optimize polymer flooding well location infill; and determining the optimal polymer flooding well location infill scheme. This invention, based on a reliable polymer flooding numerical simulation model and using the Halton sequence-improved Archimedes optimization algorithm, solves the problem of poor randomness and easy getting trapped in local optima when generating initial well location infill schemes in the original algorithm. It improves the efficiency and accuracy of polymer flooding well location infill optimization, overcomes the limitations of existing methods for polymer flooding well location infill optimization, and provides an effective method for polymer flooding well location infill optimization.

Description

Technical Field

[0001] The invention relates to a polymer drive well location infill optimization method based on an improved Archimedean algorithm, and belongs to the technical field of oil reservoir exploitation. Background Art

[0002] Polymer flooding is an important means to improve the recovery rate of offshore crude oil, and has achieved good results in increasing oil production and controlling water in offshore oil fields. At present, polymer flooding has been successfully implemented in 44 wells in multiple oil fields, with a reserve of 110 million cubic meters, a cumulative increase of 7.91 million cubic meters of oil, and an average increase of 7.1% in the degree of recovery. By injecting suitable polymers, polymer flooding technology can change the physical and chemical properties of the reservoir, reduce the oil-water interfacial tension, improve the fluidity of crude oil, and thus improve oil recovery efficiency and recovery rate. By injecting appropriate polymers into the injection well, the fluidity of crude oil and the formation conditions can be improved, and the mining efficiency can be increased. The application of polymer flooding technology can improve the economic benefits of oil fields. By improving the recovery rate, the production cycle of the oil field can be extended, the amount of remaining oil can be reduced, and the output and income can be increased. In addition, polymer flooding technology can also reduce the maintenance and operating costs of oil wells and improve production efficiency and benefits.

[0003] Domestic and foreign research and practice have found that well infill can change the flow direction of polymers, increase the utilization efficiency of remaining oil, reduce the water content of polymer flooding, and thus improve the later development effect of polymer flooding. When formulating a well infill plan, it is necessary to determine the number of infill wells and their locations. However, due to many factors such as reservoir boundaries, remaining oil distribution, and development constraints, it is difficult to accurately and efficiently determine the optimal well infill plan. At present, the vector well network method, empirical formula method, conceptual model index prediction method, and reservoir numerical simulation method are usually used at home and abroad to optimize well infill. Among them, the vector well network method, empirical formula method, and conceptual model index prediction method have large optimization errors, simple applicable scenarios, and poor adaptability to complex polymer flooding well infill optimization processes; the reservoir numerical simulation method requires a lot of data and consumes a lot of time, manpower, and material resources. Therefore, there is still a lack of effective polymer flooding well infill optimization methods. How to quickly, accurately, and conveniently optimize polymer flooding well infill is a key issue that needs to be solved urgently, which seriously restricts the widespread application of polymer flooding technology.

[0004] In recent years, intelligent optimization algorithms have been gradually applied in the field of oil and gas field development, effectively solving many complex optimization problems and improving optimization efficiency. Among them, compared with conventional intelligent optimization algorithms such as BP neural networks, the Archimedes algorithm has stronger local search capabilities and faster convergence speed. However, the algorithm has poor randomness when generating the initial well location encryption scheme and is prone to fall into the local optimal solution. The Halton sequence can improve the randomness of the initial well location encryption scheme. Therefore, the Halton sequence is used to improve the Archimedes algorithm and optimize the well location encryption scheme, thereby forming a polymer flooding well location encryption optimization method based on the Halton sequence to improve the Archimedes algorithm, which provides an effective method for polymer flooding well location encryption optimization. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a polymer flooding well location infill optimization method based on the Halton sequence improved Archimedean algorithm, which can quickly, accurately and conveniently determine the number of infill wells and their locations, form a polymer flooding well location infill optimization plan, and improve the polymer flooding later development effect. Based on the polymer flooding oil field, the well location infill optimization is performed through an intelligent optimization algorithm.

[0006] The technical solution of the present invention is as follows:

[0007] A polymer flooding well location infill optimization method based on Halton sequence improved Archimedean algorithm includes the following steps:

[0008] (1) Based on reservoir geological parameters, oil, water and rock property parameters, polymer property parameters and reservoir development data, and considering the polymer enhanced oil recovery mechanism, a polymer flooding numerical simulation model was established using reservoir numerical simulation software (CMG-STARS), laying the model foundation for subsequent well infill optimization.

[0009] (2) Based on the distribution pattern of residual oil after water flooding and polymer flooding, existing well locations, reservoir boundaries, and minimum well spacing, preliminarily determine the number of infill wells and the range of well locations for production wells and injection wells. Based on conventional experience, within the constraints of existing well locations, minimum well spacing, and reservoir boundaries, infill wells are distributed in areas with high residual oil, i.e., reservoir boundaries, with a minimum of 0 wells and a maximum of the number of wells that can be distributed under the above constraints.

[0010] (3) Based on the polymer flooding development parameters, a polymer flooding well infill calculation model was established to calculate the net present value as the optimization target of polymer flooding well infill.

[0011] (4) The Halton sequence was used to improve the Archimedean algorithm to optimize the polymer flooding well location infill and determine the optimal polymer flooding well location infill scheme.

[0012] (5) Compare the development indicators of the polymer flooding well infilling scheme before well infilling, the reservoir numerical simulation method, the Archimedes algorithm, and the polymer flooding well infilling scheme determined by the Halton sequence improved Archimedes algorithm, and evaluate the feasibility of this method in determining the optimal polymer flooding well infilling scheme.

[0013] Preferably, according to the present invention, the reservoir geological parameters in step (1) include a reservoir top structure map, stratigraphic layering data, a porosity distribution map, a permeability distribution map, an oil-bearing area map, and a reservoir effective thickness map; the oil, water and rock property parameters include rock compressibility, oil-water relative permeability curve, oil and water density and viscosity; the polymer property parameters include polymer viscosity-concentration relationship data, dynamic adsorption data, degradation data, inaccessible pore volume, resistance coefficient and residual resistance coefficient; the reservoir development data include the bottom hole flow pressure, liquid production rate, oil production rate, water content of the production wells during water flooding and polymer flooding, and the bottom hole flow pressure and injection volume of the injection wells.

[0014] Preferably, according to the present invention, the polymer flooding enhanced oil recovery mechanism in step (1) includes the polymer effect on viscosity mechanism, adsorption mechanism, permeability reduction mechanism, shear rate effect on system viscosity mechanism and inaccessible pore volume mechanism.

[0015] Further preferably, the polymer property parameters in step (1) are obtained by indoor experimental measurement, but due to the scale effect, they cannot be directly applied to the polymer flooding numerical simulation model to perform reservoir scale numerical simulation. Therefore, the one-dimensional core polymer flooding cumulative oil production and water content are calculated and fitted by reservoir numerical simulation software (CMG-STARS). In this process, the polymer property parameters are adjusted until the recovery factor and water content fitting accuracy are greater than 90%, so that the polymer property parameters are suitable for reservoir scale polymer flooding numerical simulation.

[0016] Preferably, in the step (2), the distribution law of the remaining oil after water flooding and polymer flooding is determined by historical fitting through reservoir numerical simulation software (CMG-STARS); the existing well locations include the locations of existing production wells and injection wells; the reservoir boundary includes a planar reservoir boundary and a vertical reservoir boundary, the planar reservoir boundary is determined by an oil-bearing area map, and the vertical reservoir boundary is determined by an effective reservoir thickness map; the minimum well spacing is determined by the minimum oil drainage radius of the existing well locations in the reservoir.

[0017] Preferably, according to the present invention, the polymer flooding development parameters in step (3) include cumulative oil production, the number of infilled injection wells, the number of infilled production wells, the polymer flooding time after well infilling, polymer injection costs, crude oil sales price, drilling costs, depreciation rate and crude oil commodity rate.

[0018] According to the preferred embodiment of the present invention, the polymer flooding well location infill calculation model in step (3) is:

[0019] (1)

[0020] In the formula, NPV is the net present value, billion yuan; for t Operating cash flow in 2010, billion yuan; i is the depreciation rate, t 0 is the polymer flooding production year after well infilling, year; T 0 is the polymer flooding time after well infilling, year;

[0021] in The calculation formula is as follows:

[0022] (2)

[0023] In the formula, INCOME The crude oil revenue from polymer flooding after well infilling, 100 million yuan; OPEX The total investment for polymer flooding after well infilling is RMB 100 million;

[0024] in INCOME The calculation formula is as follows:

[0025] (3)

[0026] Where: Indicates t Cumulative oil production in 2010, m 3 ; Indicates t Cumulative oil production in 0-1 years, m 3 ; P oil Indicates the crude oil sales price, yuan / m 3 ; λ is the crude oil commodity rate;

[0027] OPEX The calculation formula is as follows:

[0028] (4)

[0029] Where: n i and n p They represent the number of injection wells and production wells, respectively; cost d represents the drilling cost of infill wells, billion yuan; cost p It represents the annual injection cost of each infill injection well, billion yuan.

[0030] Preferably, according to the present invention, the step of using Halton sequence to improve Archimedean optimization algorithm to perform polymer flooding well infill optimization in step (4) is:

[0031] (4-1) Generate initial well location encryption plan

[0032] According to the number of production wells and injection wells initially determined in step (2), the total number of wells and well locations is n parameters, set the N Initial well encryption scheme, the number of iterations is t , the maximum number of iterations is t max ;

[0033] Convert decimal integer t Convert to b A string of numbers in base 1 , and then place the number symmetrically to the right of the decimal point to obtain the number shown in formula (5):

[0034] (5)

[0035] in For Van der Corput sequence t by b is the bottom value, m for The number of decimal places, for t Convert to b The first number after the base number l digits;

[0036] For different bases b and the number of iterations t Substituting into formula (5) yields n Van der Corput sequences, and arrange them to obtain the Halton sequence, as shown in formula (6);

[0037] (6)

[0038] in, is the Halton sequence t OK, b 1. b 2… b n Starting from 1 n mutually prime integers;

[0039] Use formula (7) to initialize all well densification scheme parameters, volume, density and acceleration; all well densification scheme parameters include the number of densified wells for production wells and injection wells and the range of well locations;

[0040] (7)

[0041] In the formula ，d i ( t ), v i ( t ), a i ( t ), Respectively represent i Well Densification Plan t The density, volume, acceleration, and well densification scheme parameters (number of production wells and injection wells and their range) of the iteration. rand is a random number between 0 and 1. u 0. l 0 represents the upper and lower limits of acceleration, , They represent the number of infill wells and the upper and lower limits of the well location range respectively;

[0042] (4-2) Calculation of fitness of well density scheme

[0043] The first i The parameters of the well densification scheme are input into the established polymer flooding numerical simulation model for calculation to obtain the cumulative oil production corresponding to the parameters of the well densification scheme. Then, combined with the number of densified wells of production wells and injection wells in the parameters of the well densification scheme, the well densification scheme is calculated using formulas (1)-(4). NPV , the obtained NPV The negative number of is taken as the fitness value of the well encryption scheme;

[0044] (8)

[0045] In the formula, fitness i ( t ) is the i Well Densification Plan t The fitness value of the iteration, NPV i ( t ) is the i Well Densification Plan t The net present value of the iteration, 100 million yuan;

[0046] Compare tThe fitness value of each well densification scheme in the iteration is set as the parameters, volume, density and acceleration of the well densification scheme with the smallest fitness value. , v best ( t ), d best ( t )and a best ( t );

[0047] (4-3) Update of volume and density

[0048] Using formula (9), the volume and density in the well densification scheme are updated:

[0049] (9)

[0050] In the formula, d i ( t +1) indicates the i Well Densification Plan t +1 iteration density, v i ( t +1) indicates the i Well Densification Plan t +1 iteration volume;

[0051] In order to improve the efficiency of well density optimization, the calculation t The transfer factor of the iteration TF ( t ), its value range is 0~1, and its calculation method is shown in formula (10);

[0052] (10)

[0053] In the formula, t max Indicates the maximum number of iterations;

[0054] (4-4) Acceleration and well density scheme parameter update

[0055] ① When TF(t) When ≤0.5, the algorithm is in the global exploration stage, and the acceleration update method is shown in formula (11);

[0056] (11)

[0057] In the formula, Indicates i Well Densification Plan t+1 The acceleration of the iteration,d mr ( t ), and v mr ( t ) indicates the t The density, acceleration, and volume of a well infill solution randomly determined in the iteration;

[0058] No. i The parameters of the well density scheme are updated according to formula (12);

[0059] (12)

[0060] In the formula, express i Well Densification Plan t +1 iteration of well encryption scheme parameters, C1 is a constant, Indicates t The well encryption scheme parameters randomly selected in the iteration, D ( t +1) is the t +1 iteration density factor, For the t +1 iteration normalized acceleration;

[0061] The calculation method is as follows:

[0062] (13)

[0063] in, u a , l a They are The upper and lower limits of , Respectively t +The maximum and minimum values ​​of the acceleration of the well densification scheme in 1 iteration;

[0064] D ( t +1) The calculation method is as follows:

[0065] (14)

[0066] ② When TF(t) When >0.5, the algorithm is in the local exploration stage, and the acceleration update method is shown in formula (15);

[0067] (15)

[0068] No.i The parameters of the well density scheme are updated according to formula (16);

[0069] (16)

[0070] Among them, C2 and C3 are constants. TF ( t +1) t +1 iteration transfer factor value, F is the random direction variable, expressed by formula (17);

[0071] (17)

[0072] Among them, C4 is a constant;

[0073] (4-5) Fitness update

[0074] After the well location densification scheme parameters are updated, the fitness values ​​of each well location densification scheme are recalculated using step (4-2);

[0075] (4-6) Algorithm solution

[0076] Repeat steps (4-2) to (4-5) and record NPV and t If the relationship t In the iteration, all well encryption schemes have the same fitness value or t = t max , the algorithm ends, at this time is the optimal well location encryption scheme parameter.

[0077] Preferably, according to the present invention, the development indicators in step (5) are cumulative oil production, net present value, water content, recovery degree, polymer flow line distribution and residual oil saturation distribution.

[0078] Based on the above algorithm, the number of production wells, injection wells and the coordinates of each well location can be automatically output to facilitate on-site application.

[0079] The beneficial effects of the present invention are:

[0080] 1. Based on reservoir geological parameters, oil, water and rock property parameters, polymer property parameters and reservoir development data, considering the polymer enhanced oil recovery mechanism, a polymer flooding numerical simulation model is established to achieve accurate simulation of the oilfield polymer flooding process.

[0081] 2. The present invention improves the Archimedean optimization algorithm through the Halton sequence, solves the problem that the original algorithm has poor randomness when generating the initial well location encryption plan and is prone to falling into the local optimal solution, improves the efficiency and accuracy of polymer flooding well location encryption optimization, and breaks through the limitations of existing methods for polymer flooding well location encryption optimization.

[0082] 3. Based on the characteristics of polymer flooding in oil reservoirs, polymer flooding development and economic parameters, the present invention proposes an economic calculation model for polymer flooding well infilling, which can conveniently and accurately realize the optimization process of polymer flooding well infilling, and provides a practical goal and basis for the optimization of polymer flooding well infilling in oil reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 It is a schematic diagram of production data of production wells;

[0084] Figure 2 It is a schematic diagram of injection well production data;

[0085] Figure 3 It is the result diagram of cumulative oil production and water content of one-dimensional core polymer flooding experiment calculated and fitted by reservoir numerical simulation software;

[0086] Figure 4 Cumulative oil production of polymer flooding in each production year;

[0087] Figure 5 It is the remaining oil distribution map determined after history fitting of water flooding and polymer flooding;

[0088] Figure 6 It is based on the Archimedean Algorithm (AOA) NPV Graph changes with the number of iterations;

[0089] Figure 7 It is based on Halton sequence improved Archimedean algorithm (Halotn-AOA) NPV Graph changes with the number of iterations;

[0090] Figure 8 is the cumulative oil production of each scheme and NPV Comparison chart;

[0091] Fig. 9 This is a comparison chart of the recovery degree and water content of each scheme;

[0092] Fig.10 It is a comparison diagram of polymer streamline distribution of each scheme;

[0093] Fig.11 It is a comparison chart of the remaining oil saturation distribution of each scheme. DETAILED DESCRIPTION

[0094] The present invention will be further described below by way of embodiments in conjunction with the accompanying drawings, but is not limited thereto.

[0095] Embodiment 1:

[0096] A polymer flooding well infill optimization method based on an improved Archimedes algorithm comprises the following steps:

[0097] (1) Based on reservoir geological parameters, oil, water and rock property parameters, polymer property parameters and reservoir development data, and considering the polymer enhanced oil recovery mechanism, a polymer flooding numerical simulation model was established using reservoir numerical simulation software (CMG-STARS), laying the model foundation for subsequent well infill optimization.

[0098] The reservoir geological parameters in step (1) include the reservoir top structure map, stratigraphic layering data, porosity distribution map, permeability distribution map, oil-bearing area map, and reservoir effective thickness map. The oil, water, and rock property parameters include rock compressibility, oil-water relative permeability curve, oil and water density and viscosity. The polymer property parameters include polymer viscosity-concentration relationship data, dynamic adsorption data, degradation data, inaccessible pore volume, resistance coefficient, and residual resistance coefficient. The reservoir development data include the bottom hole flow pressure, liquid production rate, oil production rate, water content of the production wells during water flooding and polymer flooding, and the bottom hole flow pressure and injection volume of the injection wells.

[0099] In this embodiment, the rock compression coefficient is 0.00036 kPa -1 The oil-water relative permeability curve, polymer viscosity-concentration relationship data, dynamic adsorption data, degradation data, inaccessible pore volume, resistance coefficient and residual resistance coefficient are shown in Tables 1 to 5. The bottom hole pressure, liquid production rate, oil production rate and water content of the production wells during water flooding and polymer flooding are shown in Tables 1 to 5. Figure 1 As shown in Figure 2, the bottom hole pressure and injection rate of the injection well are as follows: Figure 2 .

[0100] Table 1 Oil-water relative permeability curve

[0101]

[0102] Table 2 Polymer viscosity-concentration relationship data

[0103]

[0104] Table 3 Dynamic adsorption data

[0105]

[0106] Table 4 Degradation data

[0107]

[0108] Table 5 Inaccessible pore volume, resistance coefficient and residual resistance coefficient

[0109]

[0110] The mechanism of polymer flooding to improve oil recovery in step (1) includes the mechanism of polymer influence on viscosity, adsorption mechanism, permeability reduction mechanism, shear rate influence on system viscosity and inaccessible pore volume mechanism. In this embodiment, the mechanism of polymer flooding to improve oil recovery includes the mechanism of polymer influence on viscosity, adsorption mechanism, permeability reduction mechanism and inaccessible pore volume mechanism.

[0111] The polymer property parameters in step (1) are obtained from indoor experimental measurements. However, due to the scale effect, they cannot be directly applied to the reservoir polymer flooding numerical simulation model. Therefore, the one-dimensional core polymer flooding experimental recovery and water content are calculated and fitted using the reservoir numerical simulation software (CMG-STARS). In this process, the polymer viscosity-concentration relationship data, dynamic adsorption data, degradation data, inaccessible pore volume, resistance coefficient and residual resistance coefficient are adjusted until the recovery and water content fitting accuracy is greater than 90%.

[0112] In this example, the cumulative oil production and water content of the one-dimensional core polymer flooding experiment were calculated and fitted by reservoir numerical simulation software. Figure 3 The fitting accuracy of recovery factor and water cut is 98.51% and 92.45%, respectively, which makes the polymer property parameters suitable for reservoir-scale polymer flooding numerical simulation.

[0113] (2) Based on the distribution pattern of residual oil after water flooding and polymer flooding, existing well locations, reservoir boundaries, and minimum well spacing, the number of production wells and injection wells to be added and their location range are preliminarily determined.

[0114] In step (2), the distribution law of the remaining oil in the reservoir after water flooding and polymer flooding is determined by historical fitting through reservoir numerical simulation software (CMG-STARS); the existing well locations include the locations of existing production wells and injection wells; the reservoir boundary includes the horizontal reservoir boundary and the vertical reservoir boundary, the horizontal reservoir boundary is determined by the oil-bearing area map, and the vertical reservoir boundary is determined by the reservoir effective thickness map; the minimum well spacing is determined by the minimum oil drainage radius of the existing well locations in the reservoir.

[0115] In this embodiment, the remaining oil distribution law determined by water flooding and polymer flooding history fitting using reservoir numerical simulation software is as follows: Figure 5As shown in the figure; the existing well locations include the existing production wells and injection well locations; the horizontal reservoir boundary is determined by the oil-bearing area map, and the vertical reservoir boundary is determined by the reservoir effective thickness map; the minimum well spacing is determined by the minimum oil leakage radius of the existing well locations in the reservoir, which is 250m. It is preliminarily determined that the number of infilled production wells and injection wells ranges from 0 to 3; the infilled well location range is the grid range of the polymer flooding numerical simulation model, and the grid ranges in the X and Y directions are 0 to 18 and 0 to 16 respectively.

[0116] (3) Based on the polymer flooding development parameters, a polymer flooding well infill calculation model was established to calculate the net present value as the optimization target of polymer flooding well infill.

[0117] The reservoir polymer flooding development parameters in step (3) include cumulative oil production, the number of infilled injection wells, the number of infilled production wells, the polymer flooding time after well infilling, polymer injection costs, crude oil sales price, drilling costs, depreciation rate and crude oil commodity rate.

[0118] The economic calculation model for polymer flooding well infill is:

[0119] (1)

[0120] In the formula, NPV is the net present value, billion yuan; for t Operating cash flow in 2010, billion yuan; i is the depreciation rate; t 0 is the polymer flooding production year after well infilling, year; T 0 is the polymer flooding time after well infilling, years.

[0121] in The calculation formula is as follows:

[0122] (2)

[0123] In the formula, INCOME The crude oil revenue from polymer flooding of reservoirs after well infilling, 100 million yuan; OPEX The total investment for polymer flooding in oil reservoirs after well infilling is RMB 100 million.

[0124] in INCOME The calculation formula is as follows:

[0125] (3)

[0126] Where: Indicates t Cumulative oil production in 2010, m 3 ; Indicates t Cumulative oil production in 0-1 years, m 3 ;P oil Indicates the crude oil sales price, yuan / m 3 ; λ is the commodity rate of crude oil.

[0127] OPEX The calculation formula is as follows:

[0128] (4)

[0129] Where: n i and n p They represent the number of injection wells and production wells, respectively; cost d represents the drilling cost of infill wells, billion yuan; cost p represents the annual injection cost of each infill injection well, billion yuan; in this embodiment, cost d 30 million yuan, cost p 50 million yuan per year, n i and n p are 0 and 2 respectively, T 0 For 8 years, P oil 3000 yuan / m 3 , λ The total oil production of polymer flooding in each production year is Figure 4 As shown, it is calculated by formula (4) OPEX =1.05 billion yuan. Calculated by formula (3): INCOME =201.6 million yuan, calculated by formula (1) NPV It is 4.46 billion yuan.

[0130] (4) The Halton sequence was used to improve the Archimedean algorithm to optimize the polymer flooding well location infill and determine the optimal polymer flooding well location infill scheme.

[0131] The steps of optimizing the well location infill of polymer flooding in oil reservoirs using the Archimedean optimization algorithm improved by Halton sequence are as follows:

[0132] (4-1) Generate initial well location encryption plan

[0133] According to the number of production wells and injection wells and their location range determined in step (2), the total number is n parameters, set the generated N Initial well encryption scheme, the number of iterations is tIn this embodiment, the number of production wells and injection wells determined in step (2) ranges from 0 to 3. n 6~14, set to generate 10 initial well encryption schemes, the maximum number of iterations t max is 5.

[0134] Convert decimal integer t Convert to b A string of numbers in base 1 , and then place the number symmetrically to the right of the decimal point to obtain the number shown in formula (5):

[0135] (5)

[0136] in For Van der Corput sequence t by b is the bottom value, m for The number of decimal places, for t Convert to b The first number after the base number l In this embodiment, t is 1, b The value is 2, which converts the decimal integer 1 into a string of binary numbers. , and then place the number symmetrically to the right of the decimal point to get is 0.1.

[0137] For different bases b and the number of iterations t Substituting into formula (5) yields n Van der Corput sequences are arranged to obtain the Halton sequence, as shown in formula (6).

[0138] (6)

[0139] in, is the Halton sequence t OK, b 1. b 2… b n Starting from 1 n mutually prime integers.

[0140] In this embodiment, the base b The number of iterations are 1, 2, 3, 5, 7, 11, 13 and 17 respectively. t for 1， Substituting into formula (5) yields nVan der Corput sequence, and then get , The result is [0.1, 0.11, 0.12, 0.20,0.11, 0.10, 0.11, 0.11].

[0141] Formula (7) is used to initialize all well infill scheme parameters (number of infilled production wells and injection wells and their range), volume, density and acceleration.

[0142] (7)

[0143] In the formula ，d i ( t ), v i ( t ), a i ( t ), Respectively represent i Well Densification Plan t The density, volume, acceleration, and well densification scheme parameters (including the number of densified production wells and injection wells and the range of well locations) of the iteration, rand is a random number between 0 and 1, u 0. l 0 represents the upper and lower limits of acceleration, , Respectively represent the number of encrypted wells and the upper and lower limits of the well location range; in this embodiment, the number of iterations t When it is 1, the upper and lower limits of acceleration u 0. l 0 is 1 and 0 respectively, the upper and lower limits of the number of encrypted wells and the well location range , As shown in Table 6, the density of the first iteration of the first well encryption scheme is d 1(1) is 0.4914, the volume of the first iteration of the first well density scheme v 1(1) is 0.3367, the well location encryption scheme parameter of the first iteration of the first well location encryption scheme See Table 7 for the acceleration of the first iteration of the first well densification scheme. a 1(1) is 0.1.

[0144] Table 6 Scheme Upper and lower limits

[0145]

[0146] Table 7 Well location encryption scheme parameters

[0147]

[0148] (4-2) Calculation of fitness of well density scheme

[0149] The first i The parameters of the well densification scheme are input into the established reservoir polymer flooding numerical simulation model for calculation to obtain the cumulative oil production corresponding to the parameters of the well densification scheme. Then, the densification number of production wells and injection wells in the parameters of the well densification scheme is combined with formulas (1) to (4) to calculate the well densification scheme. NPV , the obtained NPV The negative number of is used as the fitness value of the well encryption scheme.

[0150] (8)

[0151] In the formula, fitness i ( t ) is the i Well Densification Plan t The fitness value of the iteration, NPV i ( t ) is the i Well Densification Plan t In this embodiment, the first well densification scheme parameters are input into the established polymer flooding numerical simulation model for calculation to obtain the cumulative oil production corresponding to the well densification scheme parameters. Then, combined with the densification well numbers of production wells and injection wells in the well densification scheme parameters, the net present value of the well densification scheme is calculated using formulas (1)-(4). NPV is 45.46, the obtained NPV The negative number of is used as the fitness value of the well encryption scheme, and the fitness is -45.46.

[0152] Compare t The fitness value of each well densification scheme in the iteration, and the parameters, volume, density and acceleration of the well densification scheme with the smallest fitness value are set as , v best ( t ), d best ( t )and a best ( t ); In this embodiment See Table 8, v best (1) is 0.5484, dbest (1) is 0.5584, a best (1) is 0.20.

[0153] Table 8 Well location encryption scheme parameters

[0154]

[0155] (4-3) Update of volume and density

[0156] Using formula (9), the volume and density in the well densification scheme are updated:

[0157] (9)

[0158] In the formula, d i ( t +1) indicates the i Well Densification Plan t +1 iteration density, v i ( t +1) indicates the i Well Densification Plan t +1 iteration volume; in this example t When it is 1, the density of the second iteration of the first well encryption scheme d 1(2) and volume v 1(2) are 0.49 and 0.34 respectively.

[0159] In order to improve the search efficiency of the optimal solution, calculate the t The transfer factor of the iteration TF ( t ), its value range is 0~1, and the calculation method is shown in formula (10)

[0160] (10)

[0161] in, t max Indicates the maximum number of iterations of the population; in this embodiment, t max The value is 5. t is 1, TF(1) It is 0.3716.

[0162] (4-4) Acceleration and well density scheme parameter update

[0163] ① When TF When ≤0.5, the algorithm is in the global exploration stage, and the acceleration update method is shown in formula (11).

[0164] (11)

[0165] In the formula, Indicates i Well Densification Plan t+1 The acceleration of the iteration, d mr ( t ), and v mr ( t ) indicates the t The density, acceleration and volume of a well location densification scheme randomly determined in the iteration; the density of a well location densification scheme randomly determined in the first iteration in this embodiment d mr (1) Acceleration a mr (1) and volume v mr (1) are 0.8959, 0.8071, 0.05 and 6.10 respectively. The acceleration of the second iteration of the first well densification scheme a 1( 2 ) is 6.10.

[0166] No. i The parameters of the well encryption scheme are updated according to formula (12).

[0167] (12)

[0168] In the formula, express i Well Densification Plan t +1 iteration of well encryption scheme parameters, C1 is a constant, Indicates t The well encryption scheme parameters randomly selected in the iteration, D ( t +1) is the t +1 iteration density factor, For the t +1 iteration normalized acceleration. In this embodiment, the constant C1 is 2. 2 Well encryption scheme parameters for the iteration See Table 9 for the parameters of the well density scheme randomly selected in the first iteration. See Table 10.

[0169] Table 9 Parameters of well densification scheme

[0170]

[0171] Table 10 Parameters of well density scheme

[0172]

[0173] The calculation method is as follows:

[0174] (13)

[0175] in, u a , l a They are the normalized acceleration The upper and lower limits of , Respectively t +1 The maximum and minimum values ​​of the acceleration of the well encryption scheme in the iteration; the upper and lower bounds of the normalized acceleration in this embodiment u a and l a are 0.9 and 0.1 respectively, a i-norm (2) is 0.8817.

[0176] D ( t +1) The calculation method is as follows:

[0177] (14)

[0178] In this embodiment D (1) is 2.6812, D (2) is 1.4211.

[0179] ② When TF When >0.5, the algorithm is in the local exploration stage, and the acceleration update method is shown in formula (15).

[0180] (15)

[0181] No. i The parameters of the well encryption scheme are updated according to formula (16).

[0182] (16)

[0183] Among them, C2 and C3 are constants. TF ( t +1) t +1 iteration transfer factor value, F is the random direction variable, expressed by formula (17);

[0184] (17)

[0185] Among them, C4 is a constant.

[0186] (4-5) Fitness update

[0187] After the well density scheme parameters are updated, the fitness values ​​of each well density scheme are recalculated using formula (8) in step (4-2). In this embodiment, the recalculated fitness values ​​of each well density scheme are shown in Table 11.

[0188] Table 11 Fitness values ​​of various well densification schemes

[0189]

[0190] (4-6) Algorithm solution

[0191] Repeat steps (4-2) to (4-5) and record NPV and t If the relationship t In the iteration, all well encryption schemes have the same fitness value or t = t max , the algorithm ends, at this time is the optimal well encryption scheme parameter. In this embodiment, t = t max =5, the algorithm ends, at this time are the optimal well density scheme parameters, as shown in Table 12. NPV and t The relationship is as Figure 7 shown.

[0192] Table 12 Parameters of the optimal well density scheme

[0193]

[0194] (5) Compare the development indicators of the polymer flooding well infilling scheme before well infilling, the reservoir numerical simulation method, the Archimedes algorithm, and the polymer flooding well infilling scheme determined by the Halton sequence improved Archimedes algorithm, and evaluate the feasibility of this method in determining the optimal polymer flooding well infilling scheme.

[0195] The medium development indicators refer to cumulative oil production, net present value, water content, recovery degree, polymer flow line distribution and remaining oil saturation distribution.

[0196] In this embodiment, the development indicators (cumulative oil production, net present value, water content, recovery degree, polymer streamline distribution and remaining oil distribution) of the polymer flooding well infilling scheme determined by the polymer flooding scheme before well infilling (Scheme 1), the reservoir numerical simulation method (Scheme 2), the Archimedes algorithm (Scheme 3) and the Halton sequence improved Archimedes algorithm (Scheme 4) are compared. Figure 6~Figure 11 As shown in the figure, compared with Scheme 1, Scheme 2 and Scheme 3, the cumulative oil production of Scheme 4 increased by 17,800 cubic meters, 10,200 cubic meters and 7,000 cubic meters respectively ( Figure 8 ), the net present value increased by 419.6 million yuan, 344 million yuan and 104 million yuan respectively ( Figure 8 ), the recovery rate increased by 5.2%, 2.9% and 1.4% respectively ( Fig. 9 ), the minimum moisture content decreased by 3.2%, 2.8% and 1.9% respectively ( Fig. 9 ), polymer flooding streamlines are more evenly distributed and the swept area is larger ( Fig.10 ), the residual oil saturation is lower ( Fig.11 ). As can be seen from the above, the improved Archimedes optimization algorithm based on Halton sequence can quickly and accurately realize the optimization of polymer flooding well locations, which is superior to the existing reservoir numerical simulation methods and Archimedes optimization algorithm.

Claims

1. A polymer flooding well infill optimization method based on an improved Archimedes algorithm, characterized in that: The steps include: (1) Based on reservoir geological parameters, oil, water and rock property parameters, polymer property parameters and reservoir development data, considering the polymer enhanced oil recovery mechanism, a polymer flooding numerical simulation model was established using reservoir numerical simulation software; (2) Determine the number of production wells and injection wells and their location range based on the distribution of remaining oil after water flooding and polymer flooding, existing well locations, reservoir boundaries, and minimum well spacing; (3) Based on the polymer flooding development parameters, a polymer flooding well infill calculation model was established to calculate the net present value as the optimization target of polymer flooding well infill; (4) Using Halton sequence to improve Archimedean algorithm to optimize polymer flooding well infilling and determine the optimal polymer flooding well infilling scheme; In step (4), the steps of using Halton sequence to improve Archimedean optimization algorithm to perform polymer flooding well infill optimization are as follows: (4-1) Generate initial well location encryption plan According to the number of production wells and injection wells and the range of well locations determined in step (2), a total of n parameters are set to generate N initial well location densification schemes, the number of iterations is t, and the maximum number of iterations is t max ; Convert the decimal integer t into a string of numbers in base b a1(t)a2(t)a3(t)…a m (t), and then place the number symmetrically to the right of the decimal point to obtain the number shown in formula (5): Φ b (t)=0.a m (t)a m-1 (t)…a l (t) (5) where Φ b (t) is the value of t of the Van der Corput sequence with base b, and m is Φ b The number of decimal places of (t), a l (t) is the first digit of t converted to base b; Substituting different bases b and iteration times t into formula (5) generates n Van der Corput sequences, which are arranged to obtain the Halton sequence, as shown in formula (6); in, is the tth row of the Halton sequence, b1, b2…b n are n mutually prime integers starting from 1; Use formula (7) to initialize all well densification scheme parameters, volume, density and acceleration; all well densification scheme parameters include the number of densified wells for production wells and injection wells and the range of well locations; Where, d i (t), v i (t), a i (t) They represent the density, volume, acceleration, and well density scheme parameters of the tth iteration of the i-th well density scheme, rand is a random number between 0 and 1, u0 and l0 represent the upper and lower limits of acceleration, respectively. They represent the number of infill wells and the upper and lower limits of the well location range respectively; (4-2) Calculation of the fitness of the well density scheme Input the parameters of the i-th well densification scheme into the established polymer flooding numerical simulation model for calculation, and obtain the cumulative oil production corresponding to the parameters of the well densification scheme. Then, combine the densification numbers of production wells and injection wells in the parameters of the well densification scheme, and use formulas (1)-(4) to calculate the NPV of the well densification scheme, and use the negative number of the obtained NPV as the fitness value of the well densification scheme. fitness i (t)=-NPV i (t),i=1,2,3,…,N (8) Where, fitness i (t) is the fitness value of the tth iteration of the i-th well encryption scheme, NPV i (t) is the net present value of the tth iteration of the i-th well infill scheme, 100 million yuan; Compare the fitness values ​​of each well densification scheme in the tth iteration, and set the parameters, volume, density and acceleration of the well densification scheme with the smallest fitness value as v best (t), d best (t) and a best (t); (4-3) Update of volume and density Using formula (9), the volume and density in the well densification scheme are updated: Where, d i (t+1) represents the density of the t+1th iteration of the i-th well encryption scheme, v i (t+1) represents the volume of the t+1th iteration of the i-th well densification scheme; Calculate the transfer factor TF(t) of the t-th iteration, whose value range is 0 to 1. The calculation method is shown in formula (10); Where, t max Indicates the maximum number of iterations; (4-4) Acceleration and well density scheme parameter update ① When TF(t)≤0.5, the algorithm is in the global exploration stage, and the acceleration update method is shown in formula (11); In the formula, a i (t+1) represents the acceleration of the t+1th iteration of the i-th well densification scheme, d mr (t), a mr (t) and v mr (t) represents the density, acceleration and volume of a well densification scheme randomly determined in the tth iteration; The parameters of the i-th well encryption scheme are updated according to formula (12); In the formula, represents the parameters of the well encryption scheme for the t+1th iteration of the i-well encryption scheme, C1 is a constant, represents the parameters of the well density scheme randomly selected in the tth iteration, D(t+1) is the density factor of the t+1th iteration, and a i-norm (t+1) is the normalized acceleration of the t+1th iteration; a i-norm (t+1) is calculated as follows: Among them, u a , l a They are respectively i-norm The upper and lower limits of (t+1); max(a(t+1)) and min(a(t+1)) are the maximum and minimum values ​​of the acceleration of the well encryption scheme in the t+1th iteration respectively; D(t+1) is calculated as follows: ② When TF(t)>0.5, the algorithm is in the local exploration stage, and the acceleration update method is shown in formula (15); The parameters of the i-th well encryption scheme are updated according to formula (16); Where C2 and C3 are constants, TF(t+1) is the transfer factor value of the t+1 iteration, and F is the random direction variable, which is expressed by formula (17); Among them, C4 is a constant; (4-5) Fitness update After the well location encryption scheme parameters are updated, the fitness values ​​of each well location encryption scheme are recalculated using step (4-2); (4-6) Algorithm solution Repeat steps (4-2) to (4-5) and record the relationship between NPV and t. If all well densification schemes have the same fitness value in the tth iteration or t = t max , the algorithm ends, at this time is the optimal well location encryption scheme parameter.

2. The polymer flooding well location infill optimization method based on the improved Archimedes algorithm according to claim 1 is characterized in that: The reservoir geological parameters in step (1) include a reservoir top structure map, stratigraphic layering data, a porosity distribution map, a permeability distribution map, an oil-bearing area map, and a reservoir effective thickness map; the oil, water, and rock property parameters include rock compressibility, oil-water relative permeability curve, oil and water density and viscosity; the polymer property parameters include polymer viscosity-concentration relationship data, dynamic adsorption data, degradation data, inaccessible pore volume, resistance coefficient, and residual resistance coefficient; the reservoir development data include the bottom hole flow pressure, liquid production rate, oil production rate, water content of the production wells during water flooding and polymer flooding, and the bottom hole flow pressure and injection volume of the injection wells.

3. The polymer flooding well location infill optimization method based on the improved Archimedean algorithm according to claim 1 is characterized in that: The polymer flooding enhanced oil recovery mechanism in step (1) includes the polymer effect on viscosity, adsorption mechanism, permeability reduction mechanism, shear rate effect on system viscosity and inaccessible pore volume mechanism.

4. The polymer flooding well location infill optimization method based on the improved Archimedean algorithm according to claim 2 is characterized in that: In the step (1), the polymer property parameters are obtained by indoor experimental measurement, and the one-dimensional core polymer flooding cumulative oil production and water content are calculated and fitted by reservoir numerical simulation software. In this process, the polymer property parameters are adjusted until the recovery factor and water content fitting accuracy are greater than 90%.

5. The polymer flooding well infill optimization method based on the improved Archimedean algorithm according to claim 1 is characterized in that: In the step (2), the distribution law of the remaining oil after water flooding and polymer flooding is determined by historical fitting through reservoir numerical simulation software; the existing well locations include the locations of existing production wells and injection wells; the reservoir boundary includes the plane reservoir boundary and the vertical reservoir boundary, the plane reservoir boundary is determined by the oil-bearing area map, and the vertical reservoir boundary is determined by the reservoir effective thickness map; the minimum well spacing is determined by the minimum oil drainage radius of the existing well locations in the reservoir.

6. The polymer flooding well location infill optimization method based on the improved Archimedean algorithm according to claim 1 is characterized in that: The polymer flooding development parameters in step (3) include cumulative oil production, number of injection wells, number of production wells, polymer flooding time after well densification, polymer injection cost, crude oil sales price, drilling cost, depreciation rate and crude oil commodity rate.

7. The polymer flooding well location infill optimization method based on the improved Archimedean algorithm according to claim 1 is characterized in that: The polymer flooding well location infill calculation model in step (3) is: In the formula, NPV is the net present value, billion yuan; is the operating cash flow in year t0, RMB 100 million; i is the depreciation rate, t0 is the polymer flooding production year after well infilling, years; T0 is the polymer flooding time after well infilling, years; in The calculation formula is as follows: Where, INCOME is the crude oil revenue from polymer flooding after well infilling, billion yuan; OPEX is the total investment in polymer flooding after well infilling, billion yuan; The calculation formula of INCOME is as follows: Where; represents the cumulative oil production in year t0, m 3 ; represents the cumulative oil production in year t0-1, m 3 ; P oil Indicates the crude oil sales price, yuan / m 3 ;λ is the crude oil commodity rate; The calculation formula for OPEX is as follows: OPEX=(n i +n p )·cost d +n i ·T0·cost p (4) Where n i and n p Respectively represent the number of injection wells and production wells, cost d represents the drilling cost of the infill wells, billion yuan; cost p It represents the annual injection cost of each infill injection well, billion yuan.

8. The polymer flooding well location infill optimization method based on the improved Archimedean algorithm according to claim 1 is characterized in that: The method also includes step (5), comparing the polymer flooding scheme before well infilling, the reservoir numerical simulation method, the Archimedes algorithm and the development index of the polymer flooding well infilling scheme determined by the Halton sequence improved Archimedes algorithm, and evaluating the feasibility of determining the optimal polymer flooding well infilling scheme by this method.

9. The polymer flooding well location infill optimization method based on the improved Archimedean algorithm according to claim 8, characterized in that: The development indicators in step (5) are cumulative oil production, net present value, water content, recovery degree, polymer flow line distribution and residual oil saturation distribution.

Images