Reverse osmosis membrane element and process multi-scale intelligent design method for high-flux membrane

Abstract

The invention discloses a reverse osmosis membrane element and process multi-scale intelligent design method for a high-flux membrane. The method comprises the following steps: establishing a three-dimensional multi-physical field coupling CFD model of a reverse osmosis membrane element; on the sub-millimeter scale, optimizing the geometric parameters of the water inlet separation net unit by adopting a hybrid intelligent optimization algorithm combining Bayesian optimization with pattern search; according to an optimization result, representing hydrodynamics and mass transfer characteristics through three-dimensional multi-physical field simulation; establishing a meter-scale industrial system model, and performing optimization design on process parameters and membrane parameters by adopting a genetic algorithm; according to the invention, fluid resistance caused by improvement of mass transfer capability can be inhibited, and energy consumption is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of water treatment equipment, and particularly relates to a multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes. Background Art

[0002] In the prior art, most of the flow channels of reverse osmosis membranes are reticulated diamond-shaped channels, and the structure is as Figure 1 shown. During operation, the water to be treated enters the flow channel from the water inlet end. After being treated by the reverse osmosis membrane, the concentrated water containing impurities is discharged from the water production end through the concentrated water flow channel. The reticulated diamond-shaped spacer changes the hydrodynamic conditions on the membrane surface, reduces the intercepted substances near the membrane surface, and improves the mass transfer rate, thereby reducing the rate of membrane fouling. However, the ability of the reticulated diamond-shaped spacer structure to increase the mass transfer coefficient is limited, and as the mass transfer ability increases, the pressure drop in the channel will rapidly rise, which will require more energy consumption. Summary of the Invention

[0003] Object of the Invention: The object of the present invention is to provide a multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes to solve the problems in the background art.

[0004] Technical Solution: A multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes according to the present invention includes the following steps: (1) Establish a three-dimensional multi-physical field coupled CFD model of the reverse osmosis membrane element; (2) On the sub-millimeter scale, use a hybrid intelligent optimization algorithm combining Bayesian optimization and pattern search to optimize the geometric parameters of the inlet spacer unit; (3) According to the optimization results, characterize the hydrodynamic and mass transfer characteristics through three-dimensional multi-physical field simulation; (4) Establish an industrial system model at the meter scale, and use a genetic algorithm to optimize the design of process parameters and membrane parameters.

[0005] Furthermore, the reverse osmosis membrane element includes: a reverse osmosis unit and a water production central pipe; a plurality of pure water diversion ports are provided on the pipe wall of the water production central pipe, and the reverse osmosis unit is wound around the circumference of the water production central pipe; the reverse osmosis unit includes an inlet spacer, a reverse osmosis membrane sheet, and a water production spacer stacked in sequence, wherein the inlet spacer is composed of a plurality of closely arranged V-shaped sub-units.

[0006] Furthermore, the cross-sectional shape of the V-shaped sub-unit is defined by a parabola function, a cosine function, or an absolute value function; the V-shaped sub-unit generates eddy currents in the inlet channel to improve the mass transfer coefficient and inhibit concentration polarization.

[0007] Furthermore, the three-dimensional multi-physical field coupled CFD model is constructed based on the Navier-Stokes equation and convection-diffusion.

[0008] Furthermore, the hybrid intelligent optimization algorithm is as follows: Bayesian optimization is used as the global search to generate the initial optimal solution; starting from the initial optimal solution, the pattern search algorithm is used for local fine optimization. Furthermore, the Bayesian optimization uses the expected improvement function EI as the acquisition function; the pattern search algorithm is as follows: a positive definite extended set is used to define the search direction; the solution is updated by iteratively performing search and polling operations, and the process terminates when the grid size is less than the threshold.

[0009] Furthermore, the optimization objective function in step (2) is: ; where APLR represents the ratio of the axial pressure drop per meter between the designed membrane module and the commercial membrane module, ; and respectively represent the mass transfer coefficients of the designed membrane module and the commercial membrane module; is used to control the trade-off between fluid resistance and mass transfer ability, serves as a penalty factor to penalize the case of excessive pressure drop.

[0010] Furthermore, the meter-scale industrial system model is described by one-dimensional differential-algebraic equations.

[0011] Furthermore, step (4) is as follows: The one-dimensional differential-algebraic equation formula is as follows: ; where k is the number of stages of the seawater reverse osmosis system, Q is the flow rate, is the transmembrane pressure, is the water flux, is the salinity in the retentate bulk, is the salinity in the permeate bulk, is the salinity at the membrane wall, is the dimensionless length, is the membrane area of the k-th stage, is the number of membrane elements in each pressure vessel in the k-th stage, and are the lengths of the membrane sheets in the vertical and parallel directions to the feed direction, i.e., the x and y axes directions, in the optimization module, is the water permeability coefficient, is the reflection coefficient, is the osmotic pressure coefficient, is the intrinsic rejection rate of the membrane; The flow rate is expressed as where is the kThe number of pressure vessels at level is the number of feed spacer sheets for each membrane element, and are the feed channel height and feed channel porosity, respectively; The membrane area at the k-th level is expressed as , where is the length of the membrane sheet parallel to the feed y-axis in the commercial module, is the membrane area of the commercial membrane element; The salinity in the permeate bulk and at the membrane wall is expressed as , where is the salt permeability coefficient, is the average unit mass transfer coefficient at the membrane wall; Then the process and membrane parameters for optimal design at the meter-scale include: the water permeability coefficient and the salt permeability coefficient , the number of pressure vessels at each level , the number of membrane elements in each pressure vessel , the transmembrane pressure at the inlet of each level , the number of spacer sheets for each membrane element ; and the optimization objective function for the optimization design of the process and membrane parameters is: ; including the annualized cost of the membrane and the power consumption cost per cubic meter of fresh water, where is the total membrane area required, is the cost per unit area of the membrane, is the annual amortization factor, is the annual operating time, is the permeate flow rate, is the energy cost per kilowatt-hour, is the specific energy consumption.

[0012] An electronic device according to the present invention includes a memory and a processor, the memory stores a computer program, and when the processor executes the program, the steps of any one of the methods are implemented.

[0013] Beneficial effects: Compared with the prior art, the present invention has the following remarkable advantages: In the design of the inlet screen structure at the sub-millimeter scale, the present invention adopts a V-shaped structure instead of a mesh structure, and uses the proposed hybrid intelligent optimization algorithm combining Bayesian and pattern search to optimize the geometric parameters of the membrane element. The designed structure generates eddy currents in the cross-section of the inlet channel, which can not only greatly improve the mass transfer ability of the reverse osmosis membrane element and improve the water production efficiency, but also significantly weaken the influence of concentration polarization and membrane fouling while suppressing the fluid resistance brought about by the increased mass transfer ability, reducing the energy consumption. At the same time, based on the three-dimensional multi-physics field simulation, the present invention constructs a system-level model at the meter scale and further optimizes the process and membrane parameters through the genetic algorithm. Description of the Drawings

[0014] Figure 1 It is the flow channel distribution diagram of the inlet screen in the prior art; Figure 2 It is the schematic structural diagram of the reverse osmosis membrane element of the present invention; Figure 3 is Figure 2 the flow channel distribution diagram of the screen of the reverse osmosis membrane element in Figure 4 It is the schematic diagram of the geometric design parameters and calculation domain of the inlet screen unit of the present invention; Figure 5 It is the flow chart of the geometric parameter optimization method of the inlet screen unit of the present invention; Figure 6 It is the multi-scale hierarchical optimization design framework of the reverse osmosis membrane element and system for high-flux membranes of the present invention; Figure 7 It is the iterative graph of the objective function for the geometric parameter optimization of the inlet screen unit of the present invention; Figure 8 It is the schematic diagram of the result of the geometric parameter optimization of the inlet screen unit of the present invention; Among them, Figures 1-3 in, 1. Reverse osmosis unit; 11. Inlet screen; 12. Reverse osmosis membrane sheet; 2. Produced water central pipe; 21. Diversion port. Detailed Embodiments

[0015] The technical solution of the present invention will be further described below with reference to the drawings.

[0016] As Figure 2 shown, the embodiment of the present invention provides a multi-scale intelligent design method for a reverse osmosis membrane element and process for high-flux membranes, including the following steps: (1) Establish a three-dimensional multi-physics field coupling CFD model for the reverse osmosis membrane element; specifically as follows: The reverse osmosis membrane element includes a reverse osmosis unit 1 and a product water central pipe 2. A number of pure water diversion ports 21 are provided on the pipe wall of the product water central pipe, and are evenly distributed on the outer wall of the product water central pipe 2. The reverse osmosis unit 1 is wound around the circumference of the product water central pipe 2. The reverse osmosis unit includes an inlet spacer 11, a reverse osmosis membrane sheet 12, and a product water spacer stacked in sequence, and the structure of the inlet spacer 11 is composed of a number of neatly arranged and closely spaced sub-units of a V-like structure.

[0017] Among them, the V-like structure refers to a structure whose cross-section is composed of basic functions such as parabola, trigonometric function, absolute value function, etc. And the inlet spacer includes the following design parameters: the width W1 of the sub-unit, the thickness H1 of the sub-unit, half of the height H2 of the sub-unit, the distance Ws between adjacent sub-units in the x-axis direction, and the distance Ls between adjacent sub-units in the y-axis direction. In addition, the design parameters also include the parameters of the basic function f ( x ). The width W1 of the sub-unit is 0.1 - 1 mm; the thickness H1 of the sub-unit is 0.01 - 0.1 mm; half of the height H2 of the sub-unit is 0.3 - 1 mm; the distance Ws between adjacent sub-units in the x-axis direction is 0.2 - 1 mm; the distance Ls between adjacent sub-units in the y-axis direction is 0.4 - 2 mm.

[0018] The Navier-Stokes equation is used to describe the motion of viscous incompressible fluids and is expressed as: ; The convection-diffusion equation is used to describe the mass transfer process and is expressed as: ; In the formula, represents the fluid density, represents the fluid velocity, represents the fluid pressure, represents the unit matrix, represents the dynamic viscosity of the fluid, represents the gradient operator, represents the diffusion coefficient tensor, represents the fluid concentration.

[0019] (2) On the submillimeter scale, a hybrid intelligent optimization algorithm combining Bayesian optimization and pattern search is adopted to optimize the geometric parameters of the inlet grid unit. Specifically as follows: The hybrid intelligent optimization method includes global optimization and local optimization. Among them, the global optimization adopts the Bayesian optimization algorithm, and the algorithm includes the following steps: Obtain sampling points by randomly sampling the target model, and use the sampling points to train the Gaussian process model; Determine the next evaluation point by calculating the acquisition function; After evaluating the new point, update the posterior distribution of the Gaussian process according to the new observation results; Return to training the Gaussian process model with the sampling points and repeat this process until the convergence criterion is met. Among them, the convergence criterion is a criterion jointly composed of the maximum number of evaluations, the posterior uncertainty criterion, and the objective function convergence criterion. Among them, the posterior uncertainty criterion means that if the variance of the model near the optimal solution is small enough, it indicates that the model is already certain enough that the current solution is optimal, and at this time, it can be stopped; The objective function convergence criterion means that if the change in the optimal objective value is less than a preset threshold within a certain period of time or multiple iterations, it is considered that the algorithm has converged and can be stopped.

[0020] Among them, the acquisition function refers to the expected improvement function, and its expression is: ; Among them, is the optimization variable composed of design parameters , is the mean of the Gaussian process at the variable , is the currently known optimal objective function value, is a small positive constant, is the cumulative distribution function of the standard normal distribution, is the standard deviation of the Gaussian process at the variable , is the probability density function of the standard normal distribution. The variable with the largest final acquisition function value will be used as the next evaluation point.

[0021] The local optimization part adopts the pattern search algorithm, and the algorithm includes the following steps: Determine the initial solution, the initial grid size, and the positive definite expansion set; Evaluate the objective function value of the current point and perform a search operation. If a better value cannot be found, traverse each subset in the positive definite expansion set and perform a polling operation. Otherwise, set this solution as the current position; If a better value cannot be found in the polling operation, reduce the grid size and execute the search step again. Repeat this process until the convergence criterion is met. Otherwise, set this solution as the current position; Among them, the grid is a conceptual grid composed of solutions, and the algorithm will select a solution from it and calculate the objective function with this solution; The initial solution is set to the optimal solution after Bayesian optimization; The positive definite expansion set is set to A set of search directions, where n represents the dimension of the design variables, expressed as . In addition, the convergence criterion refers to the criterion jointly composed of the objective function convergence criterion, the minimum tolerance of the grid size, and the maximum number of iterations. Among them, the minimum tolerance of the grid size means that when the grid size is reduced to a very small threshold, it is considered that the algorithm has converged and can stop.

[0022] The optimization objective function for the geometric parameter optimization design of the inlet grid unit is: ; Among them, APLR represents the ratio of the axial pressure drop per meter between the designed membrane module and the commercial membrane module, ; and respectively represent the mass transfer coefficients of the designed membrane module and the commercial membrane module; is used to control the trade-off between fluid resistance and mass transfer ability, serves as a penalty factor to penalize the situation of excessive pressure drop.

[0023] (3) According to the optimization results, characterize the hydrodynamic and mass transfer characteristics through three-dimensional multi-physics field simulation; specifically as follows: within the set calculation domain, characterize the Darcy friction factor and Sherwood number for different Reynolds numbers. Due to the periodic structure of the inlet grid, the calculation domain is constructed as five sub-units continuous in the y-axis direction.

[0024] Specifically, the Darcy friction factor is calculated by the following formula: ; Among them, is the fluid density, is the average flow velocity at the inlet, is the length of the set calculation domain, is the hydraulic diameter, is the pressure drop of the calculation domain length.

[0025] The Sherwood number is calculated by the following formula: ; The average mass transfer coefficient of the membrane wall is calculated by the following formula: ; Among them, is the width of the set calculation domain, is the solute concentration in the bulk of the retentate, is the solute concentration at the membrane wall.

[0026] (4) Establish an industrial system model at the meter scale and optimize the process parameters and membrane parameters using a genetic algorithm. Specifically: The industrial system model at the meter scale consists of one-dimensional differential-algebraic equations: ; where k is the number of stages of the seawater reverse osmosis system, Q is the flow rate, is the transmembrane pressure, is the water flux, is the salinity in the bulk of the retentate, is the salinity in the bulk of the permeate, is the salinity at the membrane wall, is the dimensionless length, is the membrane area of the k-th stage, is the number of membrane elements in each pressure vessel in the k-th stage, and are the lengths of the membrane sheets perpendicular and parallel to the feed direction (x and y axis directions) in the optimization module, is the water permeability coefficient, is the reflection coefficient, is the osmotic pressure coefficient, is the intrinsic rejection rate of the membrane.

[0027] Specifically, the flow rate can be expressed as where, is the k number of pressure vessels in the k-th stage, is the number of inlet spacer sheets for each membrane element, and are the feed channel height and feed channel porosity, respectively.

[0028] The membrane area of the k-th stage can be expressed as , where, is the length of the membrane sheet parallel to the feed direction (y axis direction) in the commercial module, is the membrane area of the commercial membrane element.

[0029] The salinity in the bulk of the permeate and at the membrane wall can be expressed as , where, is the salt permeability coefficient, is the average unit mass transfer coefficient at the membrane wall.

[0030] The process and membrane parameters optimized at the meter scale include: the water permeability coefficient and salt permeability coefficient of the membrane, the number of pressure vessels at each stage, the number of membrane elements in each pressure vessel, the transmembrane pressure at the inlet of each stage, the number of spacer sheets 。And the optimization objective function for the optimization design of the process and membrane parameters is: ; including the annualized cost of the membrane and the power consumption cost per cubic meter of fresh water, where is the total required membrane area, is the cost per unit area of the membrane, is the annual amortization factor, is the annual operating time, is the permeate flow rate, is the energy cost per kilowatt-hour, is the specific energy consumption; Specific embodiments: The V-like structure is represented by using a cosine function, a parabola function, and an absolute value function, which are mathematically expressed as: ; Table 1 Geometric design parameters of the inlet spacer and the corresponding dimensions of the computational domain optimized by using the hybrid intelligent optimization algorithm under different trade-off factors ; Based on the conditions of an inlet salinity of 35000 ppm, a water recovery rate of 50%, a pump efficiency of 85%, an energy recovery efficiency of 95%, and a concentration polarization limit not exceeding 1.2, the process and membrane parameters of an industrial system model at the meter scale are optimized (taking a two-stage reverse osmosis system as an example) for different costs per unit area of the membrane .

[0031] The specific energy consumption of the two-stage reverse osmosis system is calculated by the following formula:

[0032] where and represent the flow rates at the inlet of the first stage, the outlet of the first stage, and the outlet of the second stage respectively, and represent the transmembrane pressures at the inlet of the first stage, the outlet of the first stage, and the outlet of the second stage respectively, represent the hydraulic pressures at the inlet and the outlet of the second stage respectively, represents the pump efficiency of the seawater reverse osmosis system, represents the energy recovery efficiency.

[0033] Since restricting the maximum concentration polarization factor can effectively alleviate membrane fouling and scaling, which is crucial for a highly permeable seawater reverse osmosis system. The water production recovery rate , the average permeation flux , the brine salinity at the outlet of each stage and the average inlet flow velocity , and the average salinity of the permeate and other parameters also need to be within the range of engineering parameters. At the same time, the occurrence of negative driving force (i.e., the osmotic pressure difference exceeds the operating pressure) in the membrane module needs to be avoided. For a multi-stage RO system (the number of stages is set in this embodiment ), when optimizing the design of the process and membrane parameters, the constraint conditions are summarized as follows: ; Among them, the concentration polarization factor is calculated by the following formula:

[0034] The transmembrane pressure at the inlet of the first stage is calculated by the following formula:

[0035] Table 2 Design parameters of the industrial system model and their corresponding value ranges ; The optimization vector composed of system design parameters is optimized by the genetic algorithm to obtain the optimization results in Table 3.

[0036] Table 3 Optimized optimal design parameters and numerical simulation results .

Claims

1. A multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-throughput membranes, characterized in that It includes the following steps: (1) Establish a three-dimensional multi-physical field coupled CFD model of the reverse osmosis membrane element; (2) On the sub-millimeter scale, use a hybrid intelligent optimization algorithm combining Bayesian optimization and pattern search to optimize the geometric parameters of the inlet spacer unit; (3) According to the optimization results, characterize the hydrodynamic and mass transfer characteristics through three-dimensional multi-physical field simulation; (4) Establish an industrial system model at the meter scale, and use the genetic algorithm to optimize the process parameters and membrane parameters.

2. The multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes according to claim 1, characterized in that, The reverse osmosis membrane element includes: a reverse osmosis unit (1) and a permeate central pipe (2); a plurality of pure water diversion ports (21) are provided on the pipe wall of the permeate central pipe (2), and the reverse osmosis unit (1) is wound around the circumferential direction of the permeate central pipe (2); the reverse osmosis unit (1) includes an inlet spacer (11), a reverse osmosis membrane sheet (12) and a permeate spacer stacked in sequence, wherein the inlet spacer (11) is composed of a plurality of closely arranged V-shaped sub-units.

3. The intelligent design method for multi-scale reverse osmosis membrane elements and processes for high-flux membranes according to claim 2, characterized in that The cross-sectional shape of the V-shaped sub-unit is defined by a parabola function, a cosine function or an absolute value function; the V-shaped sub-unit generates eddy currents in the inlet channel to improve the mass transfer coefficient and suppress concentration polarization.

4. A multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes according to claim 1, characterized in that, The three-dimensional multi-physical field coupled CFD model is constructed based on the Navier-Stokes equation and convection-diffusion.

5. A multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes according to claim 1, characterized in that The hybrid intelligent optimization algorithm is specifically as follows: use Bayesian optimization as the global search to generate an initial optimal solution; starting from the initial optimal solution, use the pattern search algorithm for local fine optimization.

6. The intelligent design method for multi-scale reverse osmosis membrane elements and processes for high-flux membranes according to claim 5, characterized in that Bayesian optimization uses the expected improvement function EI as the acquisition function; the pattern search algorithm is specifically as follows: use a positive definite extended set to define the search direction; update the solution by iteratively executing search and polling operations, and terminate when the grid size is less than the threshold.

7. A multi-scale intelligent design method for reverse osmosis membrane elements and processes for high-flux membranes according to claim 1, characterized in that, The optimization objective function in step (2) is: ； wherein, APLR represents the ratio of the axial pressure drop per meter between the designed membrane module and the commercial membrane module, ; and respectively represent the mass transfer coefficients of the designed membrane module and the commercial membrane module; is used to control the trade-off between fluid resistance and mass transfer ability, as a penalty factor to penalize the situation of excessive pressure drop.

8. The intelligent design method for reverse osmosis membrane elements and processes at multiple scales for high-flux membranes according to claim 1, characterized in that, The industrial system model at the meter scale is described by one-dimensional differential-algebraic equations.

9. The multi-scale intelligent design method of a reverse osmosis membrane element and process for a high-flux membrane according to claim 8, wherein, Step (4) is specifically as follows: The formula of the one-dimensional differential-algebraic equation is as follows: ； where k is the number of stages of the seawater reverse osmosis system, Q is the flow rate, is the transmembrane pressure, is the water flux, is the salinity in the bulk of the retentate, is the salinity in the bulk of the permeate, is the salinity at the membrane wall, is the dimensionless length, is the membrane area of the k-th stage, is the number of membrane elements in each pressure vessel in the k-th stage, and are the lengths of the membrane sheets in the optimization module perpendicular and parallel to the feed direction, i.e., in the x and y axis directions, is the water permeability coefficient, is the reflection coefficient, is the osmotic pressure coefficient, is the intrinsic rejection rate of the membrane; The flow rate is expressed as wherein is the number of pressure vessels at the k th level, is the number of inlet spacer sheets for each membrane element, and are the feed channel height and the feed channel porosity respectively; The membrane area of the k-th stage is expressed as , where is the length of the membrane sheet parallel to the feed y-axis direction in the commercial module, is the membrane area of the commercial membrane element; The salinity at the permeate main body and the membrane wall is expressed as , where is the salt permeability coefficient, is the average unit mass transfer coefficient at the membrane wall; The optimized process and membrane parameters at the meter scale include: the water permeability coefficient of the membrane and the salt permeability coefficient , the number of pressure vessels at each stage , the number of membrane elements in each pressure vessel , the transmembrane pressure at the inlet of each stage , the number of spacer sheets in each membrane element ; and the optimization objective function for the optimized design of the process and membrane parameters is: ； including the annualized cost of the membrane and the power consumption cost per cubic meter of fresh water, where is the total required membrane area, is the cost per unit area of the membrane, is the annual amortization factor, is the annual operating time, is the permeate flow rate, is the energy cost per kilowatt-hour, is the specific energy consumption.

10. An electronic device, characterized in that, It includes a memory and a processor, the memory stores a computer program, and when the processor executes the program, it implements the steps of the method described in any one of claims 1-9.

Images