A DTS Monitoring Inversion Interpretation Method for Shale Gas Horizontal Wells Based on PSO Algorithm

Abstract

The present invention discloses a DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm, comprising the following steps: setting the PSO algorithm parameters and fitness function; initializing the positions and velocities of the particles, and calculating the individual extreme values and the global extreme value gbestt; updating the velocities and positions of the particles, and calculating the new individual extreme values and the global extreme value gbestt + 1; comparing the new individual extreme value and the global extreme value gbestt + 1 with the current individual extreme value and the global extreme value gbestt respectively, and retaining the better individual extreme values and global extreme values; performing iterative calculations until the fitness function satisfies the convergence condition, and outputting the inversion solution of the m×n-dimensional target parameters to be inverted for the shale gas horizontal well; inputting the obtained inversion solution of the m×n-dimensional target parameters to be inverted into the temperature forward prediction model, and calculating the production contribution and production profile of each cluster of fractures. By inverting the measured DTS data, the present invention can quantitatively interpret the m×n-dimensional fracture parameters, the production contribution and production profile of each cluster of fractures for the shale gas horizontal well.

Description

Technical Field

[0001] The present invention relates to a DTS (Distributed Temperature Sensing) monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO (Particle Swarm Optimization) algorithm, belonging to the technical field of oil and gas reservoir development. Background Art

[0002] At present, shale gas, as a clean unconventional natural gas resource, has become the focus of domestic natural gas resource exploration and development. In order to improve the productivity of shale gas wells, the current main method is to develop shale gas reservoirs by combining horizontal wells with staged volume fracturing stimulation measures. Therefore, the effective transformation of the reservoir directly determines the productivity of shale gas horizontal wells. However, after the fracturing of shale gas horizontal wells is put into production, problems such as unknown production profiles, unclear fracture contributions, and unknown effective fracture parameters are generally faced, making it difficult to ensure the accuracy, effectiveness, and rationality of fracturing transformation, which greatly affects the development efficiency of shale gas reservoirs. How to quantitatively evaluate the production profile of shale gas horizontal wells, the production contribution of each effective artificial fracture, and the characteristic parameters has become the key to solving the above technical problems.

[0003] Although it is difficult to directly measure the production profile and the flow rates of each stage of fractures in a fractured horizontal well, it is much easier to measure the temperature profile of a fractured horizontal well. With the continuous development and application of temperature measurement technologies, especially distributed fiber optic temperature measurement (DTS) technology in the oil field, the temperature profile measurement technology for horizontal wells has become relatively mature. Using technologies such as DTS, it is possible to realize real-time monitoring of the temperature profile of the entire horizontal well section and provide accurate and continuous temperature profile data of the fractured horizontal well.

[0004] At present, the research of domestic and foreign scholars on the interpretation of distributed optical fiber monitoring is mostly carried out for conventional horizontal wells, while there is less research on the inversion and interpretation of DTS data of shale gas horizontal wells. At present, the inversion and interpretation of horizontal well temperature profile data at home and abroad are mainly realized based on two algorithms, namely L-M (Levenberg-Marquart) and MCMC (Markov Chain Monte Carlo). The temperature profile of shale gas horizontal wells is affected by many factors such as the characteristic parameters of artificial fractures and the permeability distribution of the fracture network in the transformation area. Usually, the artificial fracture parameters (such as the length of artificial main fractures and conductivity) and the permeability distribution of the fracture network of shale gas horizontal wells are unknown. Therefore, when inversely interpreting the production profile of shale gas horizontal wells through DTS big data, there are multiple (m) unknown parameters to be inverted, and each unknown parameter to be inverted is of high dimension (dimension n = the number of effective artificial fractures). Therefore, when inversely interpreting DTS data of shale gas horizontal wells, in essence, the unknown parameters in the well to be inversely interpreted are an m×n-dimensional matrix variable. The conventional L-M and MCMC algorithms can at most realize the adaptive inversion of one-dimensional vectors based on DTS big data, and cannot be used to solve the key scientific problem of realizing the adaptive inversion of m×n-dimensional unknown parameters of shale gas horizontal wells based on DTS data. Because of this, the quantitative interpretation of the production profile and artificial fracture parameters of shale gas horizontal wells is still a technical problem in the industry.

[0005] In view of this, aiming at the inversion problem of DTS data of shale gas horizontal wells, a DTS data inversion model of shale gas horizontal wells is established based on the particle swarm optimization (PSO) artificial intelligence algorithm to realize the adaptive inversion of m×n-dimensional unknown parameters of shale gas horizontal wells based on DTS data, so as to realize the quantitative interpretation of the production profile, the half-length of effective artificial fractures, conductivity and the permeability distribution of the fracture network in the transformation area of shale gas horizontal wells, in order to provide the most direct basis for the realization of accurate fracturing and production optimization of shale gas horizontal wells, and provide a new technical means for the quantitative evaluation of the fracturing transformation effect of shale gas horizontal wells, so as to promote the efficient and economic development of shale gas reservoirs in China. Summary of the Invention

[0006] The present invention mainly overcomes the deficiencies in the prior art and provides a DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm, providing a new technical means for quantitatively evaluating the fracturing transformation effect of fractured horizontal wells in low-permeability gas reservoirs.

[0007] The technical solution provided by the present invention to solve the above technical problems is: a DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm, including the following steps:

[0008] S1. Set the PSO algorithm parameters according to the measured DTS data of the target shale gas horizontal well: population P, scale I, learning factors c1 and c2, maximum iteration number T * , and use the fitting evaluation objective function as the fitness function;

[0009] S2. Randomly initialize the positions and velocities of I particles in the population P, and set the position of each particle as the value of the target parameter to be inverted (m×n dimensional matrix), and set the velocity of each particle as the update amount of the target parameter to be inverted (m×n dimensional matrix). Substitute the position of each particle into the temperature forward prediction model, and calculate the fitness value of each particle through the fitness function to determine the individual extreme value of each particle and the global extreme value gbest of the population P t ;

[0010] S3. Update the velocities and positions of each particle through the particle velocity update formula and the particle position update formula to obtain the new velocities and positions of each particle. Then substitute the position of each particle into the temperature forward prediction model, and calculate the fitness value of each particle again through the fitness function to determine the individual extreme value of the new generation of particles and the global extreme value gbest of the new generation of population P t+1 ;

[0011] S4. Compare the individual extreme value and the global extreme value gbest t+1 obtained in S3 with the individual extreme value and the global extreme value gbest t obtained in step S2. Keep the better ones and eliminate the worse ones to complete the update of the individual extreme value and the global extreme value;

[0012] S5. Repeat steps S2 to S4 until one of the termination conditions of the algorithm is met, and output the inversion solution of the m×n dimensional target parameter to be inverted for the shale gas horizontal well;

[0013] S6. Input the inversion solution of the obtained m×n dimensional target parameter to be inverted into the temperature forward prediction model, and calculate the production contribution of each cluster of fractures and the horizontal well production profile of the target shale gas horizontal well.

[0014] A further technical solution is that the fitting evaluation objective function:

[0015]

[0016] In the formula, [Xinver m×n is the target parameter to be inverted (an m×n dimensional matrix); is the measured DTS temperature profile data (a 1×n dimensional vector); is the value of the temperature profile obtained by inverse simulation after inputting [[X inver m×n into the temperature prediction model (a 1×n dimensional vector).

[0017] A further technical solution is that the forward temperature prediction model includes a reservoir seepage model, a reservoir thermodynamics model, a fracture seepage model, a fracture thermodynamics model, a wellbore flow model, and a wellbore temperature model.

[0018] A further technical solution is that the individual extreme value of each particle in step S2 and the global extreme value gbest of the population P t are as follows:

[0019]

[0020]

[0021] In the formula, is the individual extreme value of the (j,k) dimension found by the i-th particle in the t-th iteration; is the optimal solution of the (j,k) dimension found by the entire population in the t-th iteration; i = 1, 2, 3...I; j = 1, 2, 3...m; k = 1, 2, 3...n; t = 1, 2, 3...T * .

[0022] A further technical solution is that the particle velocity update formula and the particle position update formula in step S3 are as follows:

[0023]

[0024]

[0025]

[0026] In the formula, is the velocity of the (j,k) dimension of the i-th particle in the (t + 1)-th iteration; is the position of the (j,k) dimension of the i-th particle in the (t + 1)-th iteration; is the position of the (j,k) dimension of the i-th particle in the t-th iteration; c1 and c2 are the learning factors of the individual extreme value and the global extreme value respectively; r1 and r2 are the influence degree perturbation factors of the individual extreme value and the global extreme value respectively; w is the inertia weight parameter; w max is the maximum value of the inertia weight parameter; w​​min is the minimum value of the inertia weight parameter.

[0027] A further technical solution is that the termination conditions satisfied by the algorithm in step S5 are as follows:

[0028] ① The number of iterations exceeds T * ;

[0029] ② The currently updated global extreme value gbest t+1 makes the fitness function satisfy:

[0030]

[0031] In the formula, [X inver m×n is the target parameter to be inverted (m×n dimensional matrix); is the measured DTS temperature profile data (1×n dimensional vector); is the temperature profile value obtained by inversion simulation (1×n dimensional vector) after inputting [X inver m×n into the temperature prediction model; ε T is the acceptable inversion error accuracy.

[0032] The present invention has the following beneficial effects:

[0033] 1. By using the particle swarm optimization (PSO) artificial intelligence algorithm, an inversion model for distributed fiber optic temperature monitoring (DTS) data of shale gas horizontal wells is established. By inverting the measured temperature profile data, quantitative interpretation of the m×n dimensional fracture parameters, the production contribution of each cluster of fractures, and the production profile of the horizontal well in shale gas can be realized.

[0034] 2. It is very difficult to directly obtain the fracture parameters of shale gas horizontal wells and the flow contribution of each fracture by using conventional testing methods. The present invention provides an inversion interpretation model and method for DTS monitoring of shale gas horizontal wells, which can help technicians in this field clarify the effective artificial fracture parameters formed after fracturing of shale gas horizontal wells and the production contribution of each cluster of fractures, and then realize quantitative evaluation of the fracturing transformation effect, providing technical support for promoting the efficient and economic development of shale gas resources in China;

[0035] 3. The present invention can, but is not limited to, predict the temperature profile, pressure profile, flow profile, flow rate of each level of fractures, reservoir temperature field distribution, and reservoir pressure field distribution of shale gas horizontal wells. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a schematic diagram of the inversion interpretation process for DTS monitoring of shale gas horizontal wells;

[0037] Figure 2 ​​Schematic diagram of effective artificial fracture identification and diagnosis results based on DTS data;

[0038] Figure 3 Schematic diagram of the inversion and interpretation results of the half-length of effective artificial fractures;

[0039] Figure 4 Schematic diagram of the inversion and interpretation results of the conductivity of effective artificial fractures;

[0040] Figure 5 Schematic diagram of the inversion and interpretation results of the fracture network permeability in the stimulated area corresponding to effective artificial fractures;

[0041] Figure 6 Schematic diagram of the interpretation results of the production contribution of each cluster of fractures. Detailed implementation manner

[0042] Next, the technical solutions of the present invention will be clearly and completely described in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0043] As Figure 1 shown, a DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm of the present invention uses a shale gas horizontal well as the target well and the half-length of artificial fractures, conductivity, and permeability distribution in the stimulated area as the inversion target parameters, and describes the specific steps of using the described method for DTS monitoring inversion and interpretation of shale gas horizontal wells;

[0044] (1) Based on Figure 2 the measured DTS data of the target shale gas horizontal well shown, perform effective artificial fracture identification and diagnosis, set the PSO algorithm parameters: population P, scale I, learning factors c1 and c2, maximum iteration number T * , and use the fitting evaluation objective function as the fitness function;

[0045]

[0046] In the formula, [X inver m×n is the target parameter to be inverted (m×n-dimensional matrix); is the measured DTS temperature profile data (1×n-dimensional vector); is the temperature profile value (1×n-dimensional vector) inversely simulated after inputting [X inver m×n into the temperature prediction model.

[0047] ​​(2) Randomly initialize the positions and velocities of I particles in population P, and set the position of each particle as the value of the target parameter to be inverted (m×n dimensional matrix), and set the velocity of each particle as the update amount of the target parameter to be inverted (m×n dimensional matrix). Substitute the position of each particle into the temperature forward prediction model, and calculate the fitness value of each particle through the fitness function to determine the individual extreme value of each particle and the global extreme value gbest of population P t ;

[0048] (3) Update the velocity and position of each particle through the particle velocity update formula and the particle position update formula to obtain the new velocity and position of each particle. Then substitute the position of each particle into the temperature forward prediction model, and calculate the fitness value of each particle again through the fitness function to determine the individual extreme value of the new generation of particles and the global extreme value gbest of the new generation of population P t+1 ;

[0049] (4) Compare the individual extreme values and the global extreme value gbest t+1 obtained in step (3) with the individual extreme values and the global extreme value gbest t obtained in step (2). Keep the better ones and eliminate the worse ones to complete the update of the individual extreme values and the global extreme value;

[0050] (5) Repeat steps (2) to (4) until one of the termination conditions of the algorithm is met, and output the inversion solution of the m×n dimensional target parameter to be inverted for the shale gas horizontal well, including but not limited to the inversion solution of the effective artificial fracture half-length as shown in Figure 3 , the inversion solution of the effective artificial fracture conductivity as shown in Figure 4 , and the inversion solution of the fracture network permeability in the transformed area corresponding to the effective artificial fracture as shown in Figure 5 ;

[0051] (6) Input the inversion solution of the obtained m×n dimensional target parameter to be inverted into the temperature forward prediction model to calculate the production contribution interpretation result of each cluster of fractures of the target shale gas horizontal well, as shown in Figure 6 .

[0052] The temperature forward prediction model described above is a comprehensive temperature prediction model for shale gas horizontal wells, including:

[0053] Reservoir seepage model:

[0054] Gas phase:

[0055]

[0056] Aqueous phase:

[0057]

[0058] Reservoir thermal model:

[0059]

[0060] Artificial fracture seepage model: Gas phase:

[0061]

[0062] Aqueous phase:

[0063]

[0064] Artificial fracture thermal model:

[0065]

[0066] Wellbore flow model:

[0067]

[0068] Wellbore thermal model:

[0069]

[0070] Where:

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] U T,I = γ(ρvC p ) T,I + (1 - γ)U T

[0080] In the formula: A—the surface area of the wellbore, m 2 ; c1—the learning factor of the individual extreme value; c2—the learning factor of the global extreme value; C g —the gas compressibility factor, MPa -1 ; C p —the fluid heat capacity, J / (kg·K); C ps —the heat capacity of the reservoir rock, J / (kg·℃); D—the well depth, m; f—the wellbore friction coefficient; I—the number of particles in the population; K fh —the vertical permeability of the fracture network in the stimulation area, mD; K fv —the horizontal permeability of the fracture network in the stimulation area, mD. K JT —the Joule-Thompson coefficient, ℃ / MPa; K T —the rock thermal conductivity, J / (m·s·℃); K TF —the fracture thermal conductivity, J / (m·s·℃); K Tfh —the vertical permeability of the fracture network in the stimulation area, J / (m·s·℃); K Tfv —the horizontal permeability of the fracture network in the stimulation area, J / (m·s·℃); K Tnet —the comprehensive thermal conductivity of the fracture network in the stimulation area, J / (m·s·℃); K net —the comprehensive permeability of the fracture network in the stimulation area, mD; k Fx —the permeability of the fracture in the x direction, mD; k Fy —the permeability of the fracture in the y direction, mD; k Fz —the permeability of the fracture in the z direction, mD; k Frgx —the gas relative permeability of the fracture in the x direction; k Frgy —the gas relative permeability of the fracture in the y direction; k Frgz —the gas relative permeability of the fracture in the z direction; k Frwx —the water relative permeability of the fracture in the x direction; k rwy —the water relative permeability of the fracture in the y direction; k rwz —the water relative permeability of the fracture in the z direction; k rgx —the gas relative permeability of the reservoir in the x direction; k rgy —the gas relative permeability of the reservoir in the y direction; k rgz —the gas relative permeability of the reservoir in the z direction; k rwx —the water relative permeability of the reservoir in the x direction; k rwy —the water relative permeability of the reservoir in the y direction; k rwz —the water relative permeability of the reservoir in the z direction; k x —the permeability of the reservoir in the x direction, mD; k y —the permeability of the reservoir in the y direction, mD; k z— Permeability in the z - direction of the reservoir, mD; n — Number of effective artificial fractures; p — Reservoir pressure, MPa; p F — Pressure in the artificial fracture, MPa; p wb — Pressure in the wellbore, MPa; q Fg — Flow velocity of the gas phase in the fracture, m / s; q Fw — Flow velocity of the aqueous phase in the fracture, m / s;. q wb — Rate of heat transfer from the rock per unit volume in the cementing section to the wellbore, J / (m 3 ·s); — Heat transfer rate from the reservoir in the cementing section to the wellbore, J / s; r1 — Influence degree perturbation factor of individual extreme value; r2 — Influence degree perturbation factor of global extreme value; r eff — Equivalent wellbore diameter, m; R inw — Inner diameter of the wellbore, m; S — Saturation; t prod — Production time, days; t — Number of iterations in the inversion calculation; T — Temperature, °C; T * — Number of particle iterative optimizations in the entire population of the inversion algorithm; T F — Temperature in the fracture, °C; T I — Inlet fluid temperature, °C; — Measured value of the temperature profile (vector), °C; T res — Reservoir temperature, °C; T wb — Wellbore temperature, °C; — Temperature profile obtained from the inversion simulation (vector), °C; U T — Overall heat transfer coefficient, W / (m 2 ·K); — Velocity of the i - th particle in the (j,k) - dimension at the t - th iteration; v I — Flow velocity of the inflowing fluid, m / s; v wb — Flow velocity of the fluid in the wellbore, m / s; v wb,m — Velocity of the mixed fluid in the wellbore, m / s; W — Width of the artificial main fracture, m; w — Inertia weight parameter; — Individual extreme value of the i - th particle in the (j,k) - dimension found at the t - th iteration; — Optimal solution of the entire population in the (j,k) - dimension found at the t - th iteration; X — Spacing of the artificial main fractures, m; [x inver m×n — Parameters to be inverted (m×n - dimensional matrix, where m is the number of parameters to be inverted and n is the dimension of each parameter to be inverted = number of effective artificial fractures); — Position of the i - th particle in the (j,k) - dimension at the t - th iteration; Z — Gas deviation factor; — Porosity; μ g ​— Gas viscosity, mPa·s; ψ— Pseudo-pressure function, MPa 2 / mP·s; ψ F — Fracture pseudo-pressure function, MPa 2 / mP·s; β— Coefficient of thermal expansion, 1 / ℃; γ— Degree of wellbore opening; λ— Thermal conductivity of fluid, W / (m·℃); ρ— Fluid density in reservoir, kg / m 3 ; — Average heat capacity of reservoir, J / (kg·℃); ρ I — Density of inflowing fluid, kg / m 3 ; ρ wb — Fluid density in wellbore, kg / m 3 ; ρ wb,m — Density of mixed fluid in wellbore, kg / m 3 ; ρ s — Density of reservoir rock, kg / m 3 ; θ— Horizontal wellbore dip angle, °; ε T — Acceptable inversion temperature error accuracy, ℃; Subscript α— Coupling the reservoir seepage model, reservoir thermodynamics model, fracture seepage model, fracture thermodynamics model, wellbore flow model and wellbore thermodynamics model through the heat source / sink term constitutes the temperature forward prediction model, which is used to simulate the temperature profile of shale gas horizontal wells during the inversion iteration process of measured DTS data.

[0081] As mentioned above, it is not any form of limitation to the present invention. Although the present invention has been disclosed through the above embodiments, it is not intended to limit the present invention. Any person skilled in the art, without departing from the scope of the technical solution of the present invention, may make some changes or modifications to the above-disclosed technical content to form equivalent embodiments of equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. A DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm, characterized in that It includes the following steps: S1. Set the PSO algorithm parameters according to the measured DTS data of the target shale gas horizontal well: population P , scale I , learning factors c1 and c2, maximum number of iterations , and use the fitting evaluation objective function as the fitness function; S2. Randomly initialize the population P in I the positions and velocities of the particles. Set the position of each particle as the value of the target parameter to be inverted and set the velocity of each particle as the update amount of the target parameter to be inverted . Substitute the position of each particle into the temperature forward prediction model and calculate the fitness value of each particle through the fitness function to determine the individual extreme value of each particle and the global extreme value P of the population ; S3. Update the velocity and position of each particle through the particle velocity update formula and the particle position update formula to obtain the new velocity of each particle and position . Then substitute the position of each particle into the temperature forward prediction model, and calculate the fitness value of each particle again through the fitness function to determine the individual extreme value of the new generation of particles and the global extreme value of the new generation of population P ; ; S4. Take the individual extreme value obtained in S3 and the global extreme value Compare them with the individual extreme value obtained in step S2 and the global extreme value Keep the better ones and eliminate the worse ones to complete the update of the individual extreme value and the global extreme value; S5. Repeat steps S2 to S4 until one of the termination conditions of the algorithm is satisfied, and output the inversion solution of the m×n-dimensional target parameters to be inverted for the shale gas horizontal well; S6. Input the inversion solution of the m×n-dimensional target parameters to be inverted obtained into the temperature forward prediction model, and calculate the production contribution of each cluster of fractures and the production profile of the horizontal well of the target shale gas.

2. The shale gas horizontal well DTS monitoring inversion and interpretation method based on the PSO algorithm according to claim 1, characterized in that, The fitting evaluation objective function: （1） In the formula, is the target parameter to be inverted; is the measured DTS temperature profile data; is the temperature profile value obtained by inverse simulation after inputting 3. A DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm according to claim 1, characterized in that, The temperature forward prediction model includes a reservoir seepage model, a reservoir thermodynamics model, a fracture seepage model, a fracture thermodynamics model, a wellbore flow model, and a wellbore temperature model.

4. A DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm according to claim 1, characterized in that The individual extreme value of each particle in the step S2 and the population P of the global extreme value are as follows: （2） （3） Wherein, is the individual extreme value of the j, k dimension found by the i th particle in the t th iteration; is the optimal solution of the j, k dimension found by the entire population in the t th iteration; i = 1, 2, 3… I ; j = 1, 2, 3… m ; k = 1, 2, 3… n ; t = 1, 2, 3… .

5. A DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm according to claim 1, characterized in that The particle velocity update formula and the particle position update formula in step S3 are as follows: （4） （5） （6） In the formula, is the velocity of the i -th particle in the (j, k)-th dimension at the t +1-th iteration; is the position of the i -th particle in the (j, k)-th dimension at the t +1-th iteration; is the position of the i -th particle in the (j, k)-th dimension at the t -th iteration; c1 and c2 are the learning factors of the individual extreme value and the global extreme value respectively; r1 and r2 are the influence degree perturbation factors of the individual extreme value and the global extreme value respectively; w is the inertia weight parameter.

6. The DTS monitoring inversion and interpretation method for shale gas horizontal wells based on the PSO algorithm according to claim 1, characterized in that, The termination conditions satisfied by the algorithm in step S5 are as follows: ① The number of iterations exceeds ; ② The currently updated global extreme value Make the fitness function satisfy: （7） In the formula, is the acceptable inversion error accuracy.

Images