Well location optimization method and system based on genetic algorithm and production potential map

Abstract

The invention discloses a well location optimization method and system based on a genetic algorithm and a production potential map, and relates to the technical field of oil and gas reservoir well pattern optimization, the key points of the technical scheme are that an initial population is determined based on the production potential map, genetic algorithm operations such as selective crossover are carried out by taking an accumulated petroleum yield as a fitness function, and an evolved population is obtained; and outputting the optimal well location through multiple iterations of the genetic algorithm. According to the method, a good optimization result can be rapidly obtained, the difference between random operation optimization results is small, namely, the risk of the optimization result is reduced, and the method has certain applicability and reliability for automatic well position optimization of an actual oil reservoir. The method has important significance in improving the production effect of a single well, adjusting the structure of an injection-production well pattern and the like, and is wide in application range, high in operability and easy to popularize in different blocks.

Description

Technical Field

[0001] The present invention relates to the technical field of oil and gas reservoir well pattern optimization, and more specifically, to a well location optimization method and system based on a genetic algorithm and a production potential diagram. Background Art

[0002] With the continuous development of oil exploration technology, more and more oil reservoirs are being developed. However, how to efficiently determine the well location and well network layout is a very important issue for oil field development. Traditional well location optimization methods usually only consider some static factors, such as geological structure, rock properties and other factors. However, these factors are difficult to capture the characteristics of dynamic changes in oil fields, making it difficult for traditional methods to optimize well locations and well network layouts, and the efficiency is low.

[0003] In this regard, production potential maps are also used in this field. Based on geological models, they are used to predict indicators such as production capacity and water absorption radius after oil field development through simulation and other methods, and visualize them as images. The main principle of production potential maps is to predict the impact of well networks and oil production plans on oil field production through numerical simulation of oil reservoirs, thereby providing a scientific basis for oil field well location optimization. As a highly nonlinear problem, it is difficult for reservoir engineers to obtain the most effective deployment plan when optimizing well locations based on static factors such as geological structure and rock properties and production potential maps.

[0004] In recent years, with the rapid development of artificial intelligence technology, intelligent optimization methods such as genetic algorithms have gradually been applied to various industries. Genetic algorithm is an optimization algorithm that simulates natural evolution. Through continuous gene crossover, gene mutation and other operations, it gradually optimizes the population and gradually approaches the optimal solution. It has the characteristics of strong global search ability and fast calculation speed. Combining genetic algorithm with production potential map can more reasonably determine the well location and well network layout, and improve the efficiency of oilfield development.

[0005] To this end, the present invention hopes to provide a well location optimization method and system based on genetic algorithm and production potential map. Summary of the invention

[0006] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a well location optimization method and system based on genetic algorithms and production potential maps, which can quickly obtain better optimization results for well location optimization based on genetic algorithms and production potential maps, and the differences between random operation optimization results are small, that is, the risk of optimization results is reduced, and it has certain applicability and reliability for automatic well location optimization of actual oil reservoirs. It is of great significance to improve the production effect of a single well, adjust the injection and production well network structure, etc., and has a wide range of applications, strong operability, and is easy to promote in different blocks.

[0007] To achieve the above object, the present invention provides the following technical solution: a well location optimization method based on genetic algorithm and production potential map, comprising the following steps:

[0008] S1. Divide the grids, with i, j, and k representing the number of grids in the X, Y, and Z directions of the oil and gas reservoir, respectively. Calculate the production potential value of each grid to construct a production potential map.

[0009] S2. Determine the initial population. The steps are as follows:

[0010] The location of the first production well P1 is determined by the grid position where the maximum production potential value is located in the production potential map.

[0011] According to the distance d between the remaining grid (i, j, k) and the grid where P1 is located PijkP1 , update the remaining production potential value in the production potential map. The location of the second production well P2 is determined by the grid where the largest remaining production potential value in the production potential map is located. The updated production potential value is recorded as J P2,i,j,k , J P2,i,j,k The calculation formula is as follows:

[0012] J P2,i,j,k =J P1,i,j,k ·ln(d PijkP1 -d PijkP1min +1)

[0013] In the formula, J P1,i,j,k To determine the production potential of the grid (i, j, k) at the location of P1. P2,i,j,k To determine the production potential of the grid (i, j, k) at the location of P2. PijkP1 is the distance between the remaining grid (i, j, k) and the grid where well P1 is located. PijkP1min is the shortest distance between the remaining grids (i, j, k) and the grid where well P1 is located.

[0014] Similarly, the position of the mth production well Pm is determined by the grid position of the largest remaining potential value in the production potential map. The positions from P1 to Pm are the initial population.

[0015] S3, taking the cumulative oil production COP as the fitness function, the COP formula is as follows:

[0016]

[0017] Where T is the total production time in years. is the oil production in year n, in m 3 .

[0018] According to the result of the fitness function, the well position with high fitness is selected. The selected individuals are crossovered to obtain a new generation of population. The chromosome used for crossover contains multiple binary numbers, each of which represents a variable representing the well direction or well position.

[0019] S4, calculate the fitness of individuals in the new generation population, and replace the chromosomes with lower fitness with chromosomes with higher fitness in the new generation population under the condition that the number of populations remains unchanged. Determine whether the termination condition is met. If so, output the optimal well position. If not, perform S3 based on the current population until the termination condition is met. The termination condition is that a given maximum number of iterations is reached or the individual difference in the population is less than a given threshold.

[0020] The present invention is further configured as follows: in S3, the selected individuals are crossovered and then mutated to obtain a new generation population.

[0021] The present invention is further configured as follows: the production potential calculation formula in S1 is as follows:

[0022] J i,j,k (t) = [S o,i,j,k (t)-S or ].[P o,i,j,k (t)-P min ].LnK i,j,k .Lnr i,j,k .h woc,i,j,k .φ i,j, k .h goc,i,j,k

[0023] Where t represents the time and is a given value. i,j,k (t) is the production potential of the grid (i, j, k) at time t, in MPa·ln(10 -3 μm 2 )·ln(m). S o,i,j,k (t) is the original oil saturation at the grid (i, j, k) at time t. Sor is the residual oil saturation at the grid (i, j, k). o,i,j,k (t) is the oil pressure at the grid (i, j, k) at time t, in MPa. Pmin is the minimum bottom hole pressure, in MPa. K i,j,k is the permeability at the grid (i,j,k), in units of 10 -3 μm 2 . r i,j,k is the horizontal distance from the grid (i, j, k) to the nearest boundary, in meters. i,j,k is the porosity at the grid (i,j,k). woc,i,j,k is the vertical distance from the grid (i, j, k) to the nearest oil-water interface, in meters.goc,i,j,k It is the vertical distance from the grid (i, j, k) to the nearest oil and gas interface, in meters.

[0024] The present invention is further configured to: set well-to-well constraint conditions:

[0025] Minimum distance between wells R:

[0026] The well coordinates of all wells are positive integers: x∈Z+, y∈Z+

[0027] All wells must be deployed in a valid grid: x L ≤x≤x U ,y L ≤y≤y U

[0028] Among them, x and y represent the position of the grid on the horizontal plane, respectively. c and d are different well location numbers. L represents the lower limit, U represents the upper limit, and x represents the lower limit. L 、x U ,y L and U All are given values.

[0029] Set S5: Use the inter-well constraint conditions to match the optimal well location. If the optimal well location meets the inter-well constraint conditions, then output the optimal well location. If the optimal well location does not meet the inter-well constraint conditions, then randomly select a binary number in the chromosome of the optimal well location and modify it to a new value. Repeat S3 and S4 until the optimal well location meets the inter-well constraint conditions.

[0030] The present invention is further configured such that: P1 to Pm in S2 all satisfy the inter-well constraint conditions.

[0031] The present invention is further configured such that: if the optimal well position is outside the effective grid, the grid position in the optimal well position chromosome is modified to a grid position closest to the boundary.

[0032] The present invention is further configured such that: when selecting according to the result of the fitness function in S3, the algorithm used is a roulette wheel selection algorithm.

[0033] The present invention is further configured such that the crossover in S3 adopts a single-point crossover.

[0034] The present invention is further configured as follows: the variables of the well position direction include the wellhead inclination angle and the azimuth angle.

[0035] The present invention is further configured as follows: the variables of the well position include the position of the well on the X-axis and the position of the well on the Y-axis.

[0036] The present invention also provides a well location optimization system based on genetic algorithm and production potential map: it is installed on a computer or mobile phone, and its algorithm for well location optimization adopts the above-mentioned well location optimization method based on genetic algorithm and production potential map.

[0037] In summary, the present invention has the following beneficial effects compared with the prior art: the present invention determines the initial population based on the production potential map, takes the cumulative oil production as the fitness function, performs genetic algorithm operations such as selection and crossover, obtains the evolved population, and outputs the optimal well location through multiple iterations of the genetic algorithm. The present invention can quickly obtain better optimization results, and the differences between the random operation optimization results are small, that is, the risk of the optimization results is reduced, and it has certain applicability and reliability for the automatic well location optimization of actual oil reservoirs. It is of great significance to improve the production effect of a single well, adjust the injection and production well network structure, etc. It has a wide range of applications, strong operability, and is easy to promote in different blocks. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 Flow chart of the well location optimization method based on genetic algorithm and production potential map.

[0039] Figure 2 This is a map of the oil and gas area at the top of the oil group in the study area.

[0040] Figure 3 is the optimal well location output. DETAILED DESCRIPTION

[0041] The technical solution of the present invention will be clearly described below in conjunction with the accompanying drawings. Obviously, the described embodiments are not all embodiments of the present invention, and all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the invention.

[0042] Example

[0043] like Figure 1 FIG. 1 is a flow chart of well location optimization in a preferred embodiment of the present invention, a method for well location optimization based on a genetic algorithm and a production potential map, comprising the following steps:

[0044] S1. Divide the grids, with i, j, and k representing the number of grids in the X, Y, and Z axes of the oil and gas reservoir, respectively. Calculate the production potential value of each grid to construct a production potential map. The production potential calculation formula is as follows:

[0045] J i,j,k (t) = [S o,i,j,k (t)-S or ].[P o,i,j,k (t)-P min ].LnK i,j,k.Lnr i,j,k .h woc,i,j,k .φ i,j, k .h goc,i,j,k

[0046] Where t represents the time and is a given value. i,j,k (t) is the production potential of the grid (i, j, k) at time t, in MPa·ln(10 -3 μm 2 )·ln(m). S o,i,j,k (t) is the original oil saturation at the grid (i, j, k) at time t. Sor is the residual oil saturation at the grid (i, j, k). o,i,j,k (t) is the oil pressure at the grid (i, j, k) at time t, in MPa. Pmin is the minimum bottom hole pressure, in MPa. K i,j,k is the permeability at the grid (i,j,k), in units of 10 -3 μm 2 . r i,j,k is the horizontal distance from the grid (i, j, k) to the nearest boundary, in meters. i,j,k is the porosity at the grid (i,j,k). woc,i,j,k is the vertical distance from the grid (i, j, k) to the nearest oil-water interface, in meters. goc,i,j,k It is the vertical distance from the grid (i, j, k) to the nearest oil and gas interface, in meters.

[0047] S2. Determine the initial population. The steps are as follows:

[0048] The location of the first production well P1 is determined by the grid position where the maximum production potential value is located in the production potential map.

[0049] According to the distance d between the remaining grid (i, j, k) and the grid where P1 is located PijkP1 , update the remaining production potential value in the production potential map. The location of the second production well P2 is determined by the grid where the largest remaining production potential value in the production potential map is located. The updated production potential value is recorded as J P2,i,j,k , J P2,i,j,k The calculation formula is as follows:

[0050] J P2,i,j,k =J P1,i,j,k ·ln(d PijkP1 -d PijkP1min +1)

[0051] In the formula, J P1,i,j,k To determine the production potential of the grid (i, j, k) at the location of P1. P2,i,j,kTo determine the production potential of the grid (i, j, k) at the location of P2. PijkP1 is the distance between the remaining grid (i, j, k) and the grid where well P1 is located, in meters. PijkP1min It is the shortest distance between the remaining grids (i, j, k) and the grid where well P1 is located, in meters.

[0052] Similarly, the position of the mth production well Pm is determined by the grid position of the largest remaining potential value in the production potential map. The positions from P1 to Pm are the initial population.

[0053] S3, taking the cumulative oil production COP as the fitness function, the COP formula is as follows:

[0054]

[0055] Where T is the total production time in years. is the oil production in year n, in m 3 .

[0056] According to the result of the fitness function, a well position with high fitness is selected. In this embodiment, a roulette wheel selection algorithm is used for selection. The higher the fitness of the individual well position, the greater the probability of being selected. Then the selected individuals are crossed to obtain a new generation of population. The chromosomes used for the crossover contain multiple binary numbers, each of which represents a variable representing the direction of the well position or the position of the well position. In this embodiment, the variables of the well position direction include two variables: the wellhead inclination angle and the azimuth angle, and the variables of the well position include two variables: the position of the well position on the X-axis and the position of the well position on the Y-axis.

[0057] The crossover in this embodiment is a single-point crossover, that is, a crossover point is randomly selected, the two chromosomes are separated at the point, and then some binary numbers of the two segments are exchanged to generate two new chromosomes.

[0058] Specifically, mutation can be performed after crossover, that is, a part of the binary numbers in the chromosome of an individual is randomly selected for modification, thereby generating a new chromosome, and the mutated population is used as the new generation population for S4. Introducing mutation can enable the algorithm to have local random search capabilities and maintain population diversity.

[0059] S4, calculate the fitness of individuals in the new generation population, and replace the chromosomes with lower fitness with chromosomes with higher fitness in the new generation population under the condition that the number of populations remains unchanged. Determine whether the termination condition is met. If so, output the optimal well position. If not, perform S3 based on the current population until the termination condition is met. The termination condition is that a given maximum number of iterations is reached or the individual difference in the population is less than a given threshold.

[0060] Specifically, this embodiment also sets inter-well constraint conditions:

[0061] Minimum distance between wells R:

[0062] The well coordinates of all wells are positive integers: x∈Z+, y∈Z+

[0063] All wells must be deployed in a valid grid: x L ≤x≤x U ,y L ≤y≤y U

[0064] Among them, x and y represent the position of the grid on the horizontal plane, respectively. c and d are different well location numbers. L represents the lower limit, U represents the upper limit, and x represents the lower limit. L 、x U ,y L and U All are given values.

[0065] On this basis, setting S5 is added: match the optimal well location with the well constraint conditions. If the optimal well location meets the well constraint conditions, the optimal well location is output. If the optimal well location does not meet the well constraint conditions, a binary number is randomly selected from the chromosome of the optimal well location and modified to a new value. Repeat S3 and S4 until the optimal well location meets the well constraint conditions.

[0066] Specifically, if the optimal well location is outside the valid grid, the grid position in the chromosome of the optimal well location is modified to the grid position closest to the boundary. For optimization problems with boundary constraints, directly applying genetic algorithms may cause some well locations in the population to leave the feasible region of the search space. For these infeasible solutions, coercive measures are introduced; that is, if the well locations in the population are located outside the invalid grid or boundary, these well locations are pulled back to the grid closest to the boundary.

[0067] Specifically, P1 to Pm in S2 in this embodiment all satisfy the above-mentioned well-to-well constraint conditions. This embodiment also provides a well location optimization system based on genetic algorithm and production potential map: it is installed on a computer or mobile phone, and its algorithm for well location optimization adopts the above-mentioned well location optimization method based on genetic algorithm and production potential map.

[0068] like Figure 2As shown in the figure, this is the study area for applying the well location optimization method based on genetic algorithm and production potential map. The overall oil structure in the study area is a fault-anticline structure controlled by near east-west and northeast faults. The top surface structure of each small layer in the Zhongyou Formation is similar from bottom to top and has a certain inheritance. After 15 years of production, the existing 6 vertical wells in the study area have a cumulative production of 657 million cubic meters and a recovery rate of 25.6%. The remaining reserves in the areas with sparse production flow lines and poor reserve control are mainly divided into two categories, namely those blocked by faults and those with high structures. There are three remaining potential parts: ① The remaining gas reserves are 380 million cubic meters, ② The remaining gas reserves are 240 million cubic meters. ③ The remaining gas reserves are 260 million cubic meters.

[0069] like Figure 3 As shown in the figure, when only 5 wells were put into production in the study area, the well location optimization method based on genetic algorithm and production potential map was used to deploy new wells in the remaining potential parts of the above three locations. The purpose of deployment was to improve the reserve control and utilization degree and improve the oil group development well network.

[0070] (1) Well 1: Designed as a horizontal well, the drilling azimuth is 101°, the well depth is 5471.95m obliquely and 4516.5m vertically, i.e. the vertical thickness above the gas-water interface is 13.5m. The designed displacement before the target is 350m, and the horizontal section length is 700m.

[0071] (2) Well 2: Designed as a horizontal well, the drilling azimuth is 200°, the well depth is 5301.35 in the oblique direction and 4518 m in the vertical direction, i.e. the vertical thickness above the gas-water interface is 11 m. The designed displacement before the target is 350 m, and the horizontal section length is 550 m.

[0072] (3) Well 3: Designed as a horizontal well, the drilling azimuth is 50°, the well depth is 5223.94m obliquely and 4515.5m vertically, i.e. the vertical thickness above the gas-water interface is 13m. The designed displacement length before the target is 260m, and the horizontal section is 550m.

[0073] In summary, this embodiment determines the initial population based on the production potential map, takes the cumulative oil production as the fitness function, performs genetic algorithm operations such as selection and crossover, obtains the evolved population, and outputs the optimal well location through multiple iterations of the genetic algorithm. Using the well location optimization method and system based on the genetic algorithm and the production potential map provided by this embodiment, a good optimization result can be quickly obtained, and the difference between the random operation optimization results is small, that is, the risk of the optimization result is reduced, and it has certain applicability and reliability for the automatic well location optimization of actual oil reservoirs. It is of great significance to improve the production effect of a single well, adjust the injection and production well network structure, etc. At the same time, the method and system have a wide range of applications, strong operability, and are easy to promote in different blocks.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A well location optimization method based on genetic algorithm and production potential map, characterized by: The steps include: S1. Divide the grids, with i, j, and k representing the grid numbers in the X, Y, and Z axis directions of the oil and gas reservoir, respectively, and calculate the production potential value of each grid to construct a production potential map; S2. Determine the initial population. The steps are as follows: The location of the first production well P1 is determined by the grid location of the maximum production potential value in the production potential map; According to the distance d between the remaining grid (i, j, k) and the grid where P1 is located PijkP1 , update the remaining production potential value in the production potential map. The location of the second production well P2 is determined by the grid where the largest remaining production potential value in the production potential map is located. The updated production potential value is recorded as J P2,i,j,k , J P2,i,j,k The calculation formula is as follows: J P2,i,j,k =J P1,i,j,k ·ln(d PijkP1 -d PijkP1min +1) In the formula, J P1,i,j,k To determine the production potential of the grid (i, j, k) at the P1 position; J P2,i,j,k To determine the production potential of the grid (i, j, k) at the P2 position; d PijkP1 is the distance between the remaining grid (i, j, k) and the grid where well P1 is located; d PijkP1min is the shortest distance between the remaining grids (i, j, k) and the grid where well P1 is located; Similarly, the position of the mth production well Pm is determined by the grid position of the largest remaining potential value in the production potential map; the positions from P1 to Pm are the initial population; S3, taking the cumulative oil production COP as the fitness function, the COP formula is as follows: Where T is the total production time in years; is the oil production in year n, in m 3 ; According to the result of the fitness function, a well position with high fitness is selected; the selected individuals are crossovered to obtain a new generation of population; the chromosome used for the crossover contains multiple binary numbers, each of which represents a variable representing the direction or position of the well position; S4, calculate the fitness of individuals in the new generation population, and replace the chromosomes with lower fitness with chromosomes with higher fitness in the new generation population under the condition that the number of populations remains unchanged; determine whether the termination condition is met, and if so, output the optimal well position; if not, perform S3 based on the current population until the termination condition is met; the termination condition is that a given maximum number of iterations is reached or the individual difference in the population is less than a given threshold.

2. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1, characterized in that: In S3, the selected individuals are crossovered and mutated to obtain a new generation of population.

3. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1 or 2, characterized in that: The production potential calculation formula in S1 is as follows: J i,j,k (t)=[S o,i,j,k (t)-S or ].[P o,i,j,k (t)-P min ].LnK i,j,k .Lnr i,j,k .h woc,i,j,k .φ i,j, k .h goc,i,j,k Where, t represents the time and is a given value; J i,j,k (t) is the production potential at the grid (i, j, k) at time t; S o,i,j,k (t) is the original oil saturation at the grid (i, j, k) at time t; Sor is the residual oil saturation at the grid (i, j, k); P o,i,j,k (t) is the oil pressure at the grid (i, j, k) at time t; Pmin is the minimum bottom hole pressure; K i,j,k is the permeability at the grid (i, j, k); r i,j,k is the horizontal distance from the grid (i, j, k) to the nearest boundary; φ i,j,k is the porosity at the grid (i, j, k); h woc,i,j,k is the vertical distance from the grid (i, j, k) to the nearest oil-water interface; h goc,i,j,k is the vertical distance from the grid (i, j, k) to the nearest oil and gas interface.

4. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1 or 2, characterized in that: Set inter-well constraints: Minimum distance between wells R: The well coordinates of all wells are positive integers: x∈Z+, y∈Z+ All wells must be deployed in a valid grid: x L ≤x≤x U ,y L ≤y≤y U Among them, x and y represent the positions of the grid on the X and Y axes on the horizontal plane, respectively; c and d are different well location numbers; L represents the lower limit, U represents the upper limit, and x represents the lower limit. L 、x U ,y L and U All are given values; Setting S5: Use the inter-well constraint conditions to match the optimal well location. If the optimal well location meets the inter-well constraint conditions, output the optimal well location. If the optimal well location does not meet the inter-well constraint conditions, randomly select a binary number from the chromosome of the optimal well location and modify it to a new value. Repeat S3 and S4 until the optimal well location meets the inter-well constraint conditions.

5. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 4, characterized in that: In S2, P1 to Pm all satisfy the inter-well constraints.

6. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 4, characterized in that: If the optimal well position is outside the valid grid, the grid position in the optimal well position chromosome is modified to the grid position closest to the boundary.

7. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1 or 2, characterized in that: When selecting according to the result of the fitness function in S3, the algorithm used is the roulette wheel selection algorithm.

8. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1 or 2, characterized in that: The crossover in S3 adopts single-point crossover.

9. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1 or 2, characterized in that: Variables of well location direction include wellhead inclination and azimuth.

10. The method for optimizing well locations based on genetic algorithm and production potential map according to claim 1 or 2, characterized in that: The variables of the well location include the well location on the X-axis and the well location on the Y-axis.

11. A well location optimization system based on genetic algorithm and production potential map, characterized by: It is installed on a computer or a mobile phone, and the algorithm used for well location optimization adopts the well location optimization method based on genetic algorithm and production potential map as described in any one of claims 1-10.

Images