Method for optimizing struvite production parameters based on response surface method and Minteq combined regulation

Abstract

The invention relates to a method for optimizing struvite production parameters based on response surface method and Minteq combined regulation and control, and belongs to the technical field of parameter optimization. The invention discloses a method for optimizing struvite production parameters based on response surface method and Minteq combined regulation and control. The method has the following advantages: (1) the pH can be regulated and controlled according to the concentration of heavy metals in wastewater, the concentration of bicarbonate maximizes the recovery rate of phosphorus, the occurrence of heavy metals in struvite products is minimized, and the yield of the struvite products is improved; optimal process parameters are provided for safe and healthy application of agricultural soil fertilization; (2) the adopted response surface method has the characteristics of simple operation, high feasibility and accurate prediction; (3) regulation and control can be performed according to the concentration and alkalinity of heavy metals contained in phosphorus-containing wastewater of pig manure wastewater, and high-quality phosphorus recovery of different types of wastewater is realized; and (4) parameter regulation and control can be carried out according to the measured wastewater with phosphorus recovery value, the dosage of sodium hydroxide is saved, and the cost is greatly saved.

Description

Technical Field

[0001] This invention belongs to the field of parameter optimization technology, and relates to a method for optimizing the production parameters of struvite based on the combined control of response surface methodology and Minteq. Background Technology

[0002] Currently, my country is the world's largest pig-producing country, and pig farms are gradually developing towards highly large-scale and intensive operations. However, the development of the pig industry and the increasing demand for livestock products have resulted in high concentrations of ammonia nitrogen (NH4-N, approximately 230-3200 mg / L) and orthophosphate (PO4-P, approximately 100-200 mg / L) in pig manure wastewater, which has a significant impact on the environment. Direct discharge without treatment will lead to severe eutrophication. Magnesium ammonium phosphate (MgNH4PO4•6H2O, MAP), commonly known as struvite, is widely considered a high-quality slow-release fertilizer that can simultaneously recover N and P. The struvite crystallization process has been widely used to treat pig manure wastewater to recover nitrogen and phosphorus nutrients.

[0003] However, due to the addition of heavy metals (especially Cu and Zn) to pig feed to prevent pig diseases and promote pig growth, the emissions of heavy metals such as Cu and Zn in feed additives far exceed the utilization rate, resulting in excessive levels of heavy metals (especially Cu and Zn) in the excreted livestock wastewater. Consequently, heavy metals are present on the struvite recovered from pig manure wastewater, which limits the safe agricultural use of struvite as fertilizer.

[0004] The occurrence of heavy metals on struvite is influenced by factors such as pH, heavy metal concentration, bicarbonate, and foreign ions. Current research on the effects of Cu and Zn on struvite occurrence is limited to single-factor studies, and the interactions between these factors cannot be effectively analyzed. This results in the applicability of some factors not being effectively adjustable when other factors are involved. Furthermore, it is crucial to adjust the pH and HCO3- concentration according to the desired heavy metal concentration range to minimize heavy metal occurrence on struvite crystals, while simultaneously improving phosphorus recovery, thereby enhancing the safe agricultural use of struvite as a slow-release fertilizer.

[0005] Response surface methodology (RSM) offers advantages in optimizing stochastic processes through statistical experiments. Its regression models closely integrate mathematical statistics with computer science, reducing computational complexity while saving on manual computation costs. By experimenting with multiple factors proposed in the model and substituting the results into the model, the quantitative laws governing the relationship between the experimental index (dependent variable) and each factor (independent variable) are identified. This reveals the interactions between factors on the dependent variable, ultimately leading to the optimal combination of independent variables required by the target dependent variable. The regression model established by RSM is a complex multidimensional surface space, which more closely approximates actual results compared to traditional experimental models. The Central Composite Design (CCD) model, a type of RSM, possesses rotational properties absent in Box-Behnken designs (BBD). It is widely used due to its advantages such as fewer experiments, better predictive performance, higher precision, and facilitating multi-factor interactions.

[0006] Minteq is a classic hydrogeochemical equilibrium calculation software jointly developed by the US Environmental Protection Agency (EPA) and the Royal Institute of Technology in Sweden. It incorporates a database of over a thousand minerals, complexes, redox reactions, and surface reactions. By solving simultaneous equations of mass balance, charge balance, and equilibrium constants, it can quickly provide the speciation of each component, saturation index (Sa), and potential precipitation / dissolution pathways under given water quality conditions. It has been widely used to predict the crystallization trends of magnesium ammonium phosphate (struvite) and other byproducts in wastewater treatment. Introducing Minteq into struvite process optimization allows for quantitative assessment of the occurrence speciation and co-precipitation risk of heavy metals such as Cu and Zn before experiments. Combined with response surface methodology, it enables a closed-loop optimization strategy of "simulation first, experiment second, and calibration third," significantly reducing trial-and-error costs and improving parameter reliability.

[0007] Therefore, it is necessary to conduct in-depth research on new methods for optimizing the parameters of struvite production. Summary of the Invention

[0008] In view of this, the purpose of this invention is to provide a method for optimizing the production parameters of struvite based on the combined control of response surface methodology and Minteq.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] 1. A method for optimizing struvite production parameters based on the combined use of response surface methodology and Minteq, the method comprising the following steps:

[0011] (1) pH, heavy metal concentration (X3, where the heavy metal concentration is the total concentration of Cu and Zn), heavy metal molar ratio (where the heavy metal ratio is the molar ratio of Cn and Zn), and bicarbonate concentration (X4) are used as independent variables, and Cu content, Zn content, and phosphorus recovery rate are used as dependent variables.

[0012] (2) Conduct single-factor experiments on the independent variables and analyze the range of variation of the extreme values ​​of the dependent variables to determine the experimental range required for the response surface model;

[0013] (3) Based on the interaction relationship between the four independent variables and the three dependent variables, the experimental scheme of the response surface method was designed using Design-Expert software and the central composite design (CCD) model was selected.

[0014] (4) A batch reactor was simulated using a beaker and a magnetic stirrer. The experiment was conducted according to the experimental plan, and three sets of parallel data were used for comparison. Each set of data included PO4 in the supernatant used to calculate the phosphorus recovery rate (γ3). 3- Concentration, residual Cu concentration in the supernatant used to calculate Cu content (γ1), and residual Zn concentration in the supernatant used to calculate Zn content (γ2);

[0015] (5) Substitute the data from each group in step (4) into the Design-Expert software for data processing, and use pH, heavy metal concentration (Cu+Zn), heavy metal molar ratio, and bicarbonate concentration as independent variables to calculate the response value of the dependent variable using the following quadratic multiple regression model:

[0016]

[0017] In the formula: Y i X is the response value of the dependent variable. i ß is the independent variable. i β is the coefficient of the linear term, β0 is the constant term, β ij β is the coefficient of the interaction term. ii The coefficient of the quadratic term;

[0018] (6) For the equation coefficients (β0, β) in step (5) i β ij β ii The regression model is subjected to analysis of variance (ANOVA), p-test, F-test and significance test. A model with good predictive performance can be obtained after the following conditions are met: the coefficient of determination (RSquared2) > 0.9, the adjusted coefficient of determination (Radj2) > 0.9, the model (Mode1) is significant and the lack of fit is not significant.

[0019] (7) Select any set of parameters within the range of independent variables, repeat the experiment to analyze the degree of agreement between the model prediction variable value and the experimental value, so as to further verify the accuracy of the model. The range of independent variables is determined by the single-factor experiment in step (2). The specific method is as follows: First, perform single-factor scanning on the four factors of pH, heavy metal concentration, heavy metal molar ratio and bicarbonate concentration respectively, find the high-low boundary of each factor that causes the response to have extreme values, and then combine these four boundaries into the range of independent variables for subsequent modeling and verification. The model prediction variable value is the prediction result given by the quadratic multiple regression model for the three dependent variables of Cu content Y1 (mg / L), Zn content Y2 (mg / L) and phosphorus recovery rate Y3 (%).

[0020] (8) At the same time, the minimum value of Cu content (Y1), the minimum value of Zn content (Y2) and the maximum value of phosphorus recovery rate (Y3) are used as target values. The priority is set according to the importance of heavy metals and phosphorus recovery rate. The Optimization module of Design-Expert is used to optimize and obtain the best parameter combination for producing struvite.

[0021] Preferably, the unit of the heavy metal concentration (X3) is µM, the unit of the bicarbonate concentration (X4) is mM, the unit of the Cu content (Y1) is mg / g, and the unit of the Zn content is mg / g.

[0022] Preferably, X1 is pH, X2 is the initial molar ratio of Cu to Zn, and X3 is the total concentration of Cu and Zn;

[0023] The synthetic wastewater includes ammonium chloride, disodium hydrogen phosphate dodecahydrate, magnesium chloride hexahydrate, sodium bicarbonate, zinc sulfate heptahydrate, copper sulfate pentahydrate, sodium hydroxide, hydrochloric acid, nitric acid, and sodium chloride.

[0024] Preferably, the pH range is 8-9, the total heavy metal concentration ranges from 40-80 µM, the heavy metal molar ratio ranges from 0.5-2, and the bicarbonate concentration ranges from 10-30 mM.

[0025] Preferably, in step (3), the surface method design scheme of the central composite design model is as follows:

[0026]

[0027] Preferably, in step (4), the batch reactor includes a magnetic stirrer, a rotor, a beaker, and a pH meter;

[0028] The reaction time in the batch reactor was 30 min. During the experiment, 2 mL of sample was taken every 1 min and the experiment was terminated with 2 µL of 10% hydrochloric acid.

[0029] In the batch reactor, the PO4 content of the supernatant was measured using an ultraviolet spectrophotometer. 3- The concentrations of Cu and Zn in the supernatant were measured using atomic absorption spectrometry.

[0030] Preferably, in step (5), the expression for the quadratic multiple regression model is:

[0031] Y1=19.27-4.9+3.79B+0.12C-0.15D-0.45AB-0.01AC+0.01AD+0.005BC+0.006BD-0.0002CD+0.32A 2 -0.08B 2 +0.00009C 2 +0.0009D 2 ,

[0032] Y2=-49.22+11.1A+0.06B+0.10C+0.35D-0.24AB-0.004AC-0.03AD-0.01BC+0.05BD-0.001CD-0.63A 2 +0.42B 2 -0.0001C 2 -0.002D 2 ,

[0033] Y3=-14.05+3.14A-0.11B-0.006C+0.09D+0.002AB+0.0004AC-0.01AD+0.002BC+0.0006BD+0.00002CD-0.16A 2 -0.02B 2 -3.79E-07C 2 -0.0002D 2 .

[0034] The beneficial effects of the present invention are as follows: The present invention discloses a method for optimizing the production parameters of struvite based on the combined control of response surface methodology and Minteq, which has the following advantages: (1) The method for optimizing the production parameters of struvite based on the combined control of response surface methodology and Minteq disclosed in the present invention can simultaneously adjust the pH according to the heavy metal concentration in the wastewater, maximize the phosphorus recovery rate by adjusting the bicarbonate concentration, minimize the occurrence of heavy metals in struvite products, and provide the best process parameters for safe and healthy application of agricultural soil fertilization; (2) The response surface methodology used in the present invention has the characteristics of simple operation, high feasibility, and accurate prediction; (3) The present invention can adjust the parameters according to the heavy metal concentration and alkalinity contained in the phosphorus-containing wastewater of pig manure wastewater, so as to achieve high-quality phosphorus recovery from different types of wastewater; (4) The present invention can adjust the parameters according to the measured wastewater with phosphorus recovery value, save the amount of sodium hydroxide added, and save a large amount of cost.

[0035] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0036] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0037] Figure 1 pH and bicarbonate (HCO3) - Three-dimensional response surface plot of phosphorus recovery rate under concentration coupling effect;

[0038] Figure 2 pH and bicarbonate (HCO3) - Three-dimensional response surface plot of Cu accumulation under concentration coupling effect;

[0039] Figure 3 pH and bicarbonate (HCO3) - Three-dimensional response surface plot of Zn stock under concentration coupling effect;

[0040] Figure 4 This is a graph verifying the good agreement between the predicted values ​​from the quadratic multiple regression model and the experimentally measured values.

[0041] Figure 5 The graphs show the fiber diameter diagnostic chart, the comparison between predicted and actual values ​​(a, b, c), the normal probability graph (d, e, f), the comparison between studentized residuals and predicted values ​​(g, h, i), and the comparison between studentized residuals and run numbers (j, k, l). Detailed Implementation

[0042] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0043] Example 1

[0044] (1) Prepare 400 mL of synthetic wastewater in a 500 mL beaker: NH4Cl 1.92 g, Na2HPO4·12H2O 0.716 g, MgCl2·6H2O 0.975 g, to obtain NH4Cl. + / PO4 3- =10, Mg / P=1.2; Adding 1.60 mg of CuSO4·5H2O and 2.88 mg of ZnSO4·7H2O, so that Cu... 2+ =30 µM, Zn 2+ =30 µM (total heavy metals 60 µM, Cu / Zn=1), NaHCO3 0.336 g makes HCO3 - =20 mM; adjust pH to 8.5±0.05 with 0.1 M NaOH. (2) Stir magnetically at 25 ℃ and 300 r / min for 30 min, take 2 mL every 1 min and filter through a 0.45 µm filter membrane, and immediately add 2 µL of 10% HCl to stop; determine PO4 by ammonium molybdate spectrophotometry. 3- , flame AAS to measure Cu and Zn; the phosphorus recovery rate Y3=94.2%, Cu content Y1=0.172 mg / L, and Zn content Y2=1.39 mg / L were calculated. (3) Substituting the above measured Y1, Y2, and Y3 into Design-Expert 13, the CCD model under secondary regression yielded: Y1=0.172 (experimental) vs 0.1723 (predicted), with a relative error of 0.2%; Y2=1.39 vs 1.39 (predicted), with a relative error of 0%; Y3=94.2% vs 92.9%, with a relative error of 1.4%; the model R²>0.9, and the lack of fit was not significant, so the verification was passed. (4) With the goal of minimizing Y1, minimizing Y2, and Y3>90%, the software gave the optimal solution: pH=8.9, Cu / Zn=0.5, total heavy metals 80 µM, HCO3 -=30 mM; the experiment was repeated under these conditions, and the measured values ​​were Y3=98.3%, Y1=0.045 mg / L, and Y2=0.28 mg / L, all of which met the 95% confidence level of the prediction interval and were determined as the final optimized parameters.

[0045] Example 2 (1) Take another 400 mL of synthetic wastewater, adjust NH4Cl 2.30 g, Na2HPO4·12H2O 0.859 g, MgCl2·6H2O 1.17 g, and maintain NH4Cl. + / PO4 3 =10, Mg / P=1.2; Add 3.19 mg CuSO4·5H2O and 1.15 mg ZnSO4·7H2O to make Cu 2+ =50 µM, Zn 2+ =20 µM (total heavy metals 70 µM, Cu / Zn=2.5), NaHCO3 0.504 g makes HCO3 - =30 mM; adjust pH to 9.0±0.05 with 0.1 M NaOH. (2) Stir for 30 min, take samples, terminate, and determine the phosphorus recovery rate Y3=97.1%, Cu content Y1=0.092 mg / L, and Zn content Y2=0.14 mg / L. (3) Substitute the data into the same CCD model, the predicted values ​​are: Y1=0.089 mg / L, Y2=0.15 mg / L, Y3=96.5%; the relative errors are 3.4%, 7.1%, and 0.6% respectively, and the model still satisfies R²>0.9 and the lack of fit is not significant. (4) Optimize again with Y1<0.1 mg / L, Y2<0.2 mg / L, and Y3>95% as the target, and the software output is: pH=9.2, Cu / Zn=2, total heavy metals 60 µM, HCO3 - =40 mM; According to the verification of this combination, the measured Y3=96.8%, Y1=0.068 mg / L, and Y2=0.11 mg / L all fell within the 95% prediction interval, and were determined to be another set of usable optimization parameters.

[0046] Example 3

[0047] The pig manure used in the experiment came from a pig farm in the suburbs of Wuhan. Regarding pH and HCO3... - The combined effects design is shown in Table 1.

[0048] Table 1 pH and HCO3 - Joint impact design

[0049]

[0050] Many previous studies have shown that the heavy metal content in struvite is mostly below 100 μM, with only a small number of studies showing that it can rise to 250 μM or higher, as shown in Table 2.

[0051] Table 2 Experimental design for the influence of heavy metals Cu and Zn and their molar ratio on the composition of guano stone.

[0052]

[0053] To further investigate the compositional changes of metal ions in struvite, this experiment used synthetic wastewater (a mixture of ammonium chloride, disodium hydrogen phosphate dodecahydrate, magnesium chloride hexahydrate, sodium bicarbonate, zinc sulfate heptahydrate, copper sulfate pentahydrate, sodium hydroxide, hydrochloric acid, and nitric acid). The N / P molar ratio in this synthetic wastewater was 10, consistent with real pig farm wastewater. By adding magnesium chloride hexahydrate as a magnesium source, the initial Mg / P molar ratio in the solution was 1.2, ultimately forming struvite. To better simulate HCO3-... - Cu 2+ and Zn 2+ In a concentrated pig manure wastewater system, NaHCO3, ZnSO4⋅7H2O, and CuSO4⋅5H2O were added to the initial wastewater solution. All chemicals used in the batch experiments were of analytical grade. 400 mL of the synthetic wastewater was stirred in a beaker at 300 rpm for 1 min using a magnetic stirrer to enhance the reaction of the three different ions. Three replicate experiments were conducted to confirm the reproducibility of the experiment. Through multiple experiments, the properties of struvite and the incorporation of heavy metals into struvite were studied in relation to the initial pH (8.0–9.0) and HCO3- of the solution. - (0–30 mM), Cu / Zn molar ratio (0–3), Cu and Zn content (0–500 μM). Solutions with different initial pH values ​​were obtained using 1 M HCl and 5 M NaOH solutions.

[0054] The first round of experiments was designed to investigate the single-factor effects of different initial concentrations (Cu and Zn) of heavy metals (2–500 μM) in struvite. The second round of experiments focused on the adsorption of metals as the initial Cu / Zn molar ratio in the solution changed under a total Cu and Zn concentration of 80 μM. Both rounds of experiments were conducted at pH 9.0, and the experimental design is shown in Table 1. In short, 2 mL of liquid sample was added using a syringe at 0 min, 1 min, 3 min, 5 min, 7 min, 15 min, and 30 min, and then immediately filtered through a 0.45 μm microporous filter. Subsequently, 0.2 mL of 5% HCl solution was added to terminate the precipitation, and the mixture was stored at 4 °C for subsequent testing and analysis. The collected white precipitate was continuously dried under vacuum at 38 °C for 48 h, and the crystals were characterized using various instruments. To investigate the differences between the synthesized struvite and pig manure water, the second round of experiments using real pig manure water was repeated three times. The guano from synthetic wastewater and the liquid obtained from filtering real pig wastewater were dissolved in diluted HNO3 solution (1%), and then the heavy metal content in the solution was analyzed and compared.

[0055] In this experiment, the central coincidence design (CCD) was used to determine the residual amount of heavy metals in four factors during the struvite crystallization process, such as pH (A), Cu. 2+ / Zn 2+ molar ratio (B), Cu 2+ and Zn 2+ The total concentration and HCO3- concentration were measured in 30 replicate experiments at 5 levels, as shown in Table 3. To collect the heavy metal content in struvite, Y1 and Y2 represent the absorption of Cu (mg / L) and Zn (mg / L), respectively, and Y3 represents the P recovery rate (%). The experimental design of the RSM model is described in more detail in the references.

[0056] Table 3. Actual and Indirect Independent Variables

[0057]

[0058] To predict the linear relationship between input parameters (X1, X2, X3, and X4) and response variables (Y1, Y2, and Y3), a PSO-BP (Particle Swarm Optimization Backpropagation) artificial neural network was used in the ANN toolbox of MATLAB R2016a (Mathworks, Inc. MA, the USA) to accommodate more samples. The network structure was divided into an input layer, hidden layers, and an output layer (Fig. S1). Based on the type and number of independent variables, the input layer contains four neurons, namely pH, Cu... 2+ Zn2+ and HCO3 - Content. Based on the type and number of dependent variables, the output layer contains three neurons: Cu adsorption, Zn adsorption, and P recovery. All experimental data, totaling 67 samples (as shown in Table 4), were used for training in this experiment. The independent variable values ​​are as follows: pH (7.5–9.5), HCO3- - The content of heavy metals (0–40 mM), Cu (0–500 μM), and Zn (0–500 μM) is also shown. The PSO-BP neural network structure is described in more detail in Table 5. Furthermore, based on the ANN, the heavy metal absorption rate, dingyi, is represented by the comprehensive index F, which is also described in more detail in Table 5.

[0059] Table 4 Experimental data (n=3)

[0060]

[0061]

[0062] Table 5 Heavy metal content of pig manure water reported in previous studies

[0063]

[0064] 1. Phosphorus recovery and heavy metal contamination from struvite

[0065] 1.1 Effect of initial metal content and Cu / Zn molar ratio

[0066] Due to the rapid crystallization process, PO4 3- The content decreased to a minimum within 2 minutes and remained at a stable level. With increasing initial Cu or Zn concentration (0–200 μM) (slightly increasing when heavy metal concentration reached 500 μM), the PO4 content in the solution... 3- The residual concentration of heavy metals in the wastewater consistently remained at a low level. Consequently, the low to medium concentrations of heavy metals in the wastewater had almost no impact on phosphorus recovery, likely due to the lower molar ratio of the metals relative to phosphorus. This, in turn, may be a result of low concentrations of copper, zinc, and chromium, which are lower than the added Mg²⁺, thus reducing side reactions. Similarly, in previous studies of batch synthesis wastewater, the Ca / Mg ratio had a negligible and acceptable effect on phosphorus recovery when the proportion was below 1 / 3. Furthermore, although at negligible levels, changes in the Cu / Zn ratio led to changes in PO₄²⁻. 3- The concentration difference. However, no obvious trend was observed; the coexistence of Cu and Zn resulted in excess PO4. 3- At higher concentrations, phosphorus recovery is significantly inhibited.

[0067] However, the increase in initial metal concentration directly led to differences in metal removal rates. At an initial concentration of 2 µM, the residual Cu concentration decreased from 0.127 mg / L to 0.054 mg / L, a reduction of 57.8%; at an initial concentration of 500 µM, it decreased from 37.77 mg / L to 29.8 mg / L, a reduction of 6.7%. It can be seen that although the removal rate of Cu (Y1) increased, the Cu removal rate (F1) decreased while the initial Cu content increased. In contrast, the zinc content in the wastewater underwent significant changes and increased with increasing zinc content. More specifically, at an initial concentration of 20 µM, it decreased from 0.13 mg / L to 0.073 mg / L, a reduction of 43.6%; at an initial concentration of 500 µM, it decreased from 32.5 mg / L to 3.18 mg / L, a reduction of 90.2%. Therefore, the adsorption rate of zinc (F2) by struvite increases with increasing initial concentration, the opposite of the adsorption pattern for copper. This difference can be attributed to their different adsorption mechanisms. Additionally, Cu... 2+ (Y1) and Zn 2+ The removal rate of (Y2) varies with the Cu / Zn molar ratio. When the solution contains only Zn (80 µM) at a Cu / Zn ratio of 0, the maximum Zn removal rate is 74.5%. However, as the Cu / Zn ratio increases, the Zn removal rate decreases to 51% at a Cu / Zn ratio of 3. Conversely, the Cu removal rate at different Cu / Zn ratios does not show a clear pattern, but is generally below 50%, lower than the Zn content at the same Cu / Zn ratio. In summary, in the competitive adsorption process, zinc adsorption is always preferential to copper.

[0068] 1.2 pH and HCO3 - The combined effects

[0069] Based on the above results, the initial pH and HCO3 - The increased content led to lower PO4 levels in the wastewater. 3- It increases the content and accelerates phosphorus recovery and struvite formation. At pH 8.0, higher HCO3 content... - The content is conducive to the formation of struvite, leading to PO4. 3- The content was significantly reduced. Higher PO4 levels. 3- The concentration level provides a strong buffering effect, maintaining a stable solution pH and ensuring continuous crystallization of struvite. However, at pH 8.0 and 9.0, HCO3-... - The effect of the concentration decreased significantly, resulting in a smaller decrease in the solution pH. This can be explained by the rapid crystallization of struvite at higher pH values, with the addition of HCO3. - It did not produce a similar effect to pH=8.0.

[0070] The following describes pH and HCO3.- Removal of Cu and Zn under combined concentration effects. Higher pH and HCO3 concentrations... - Both inhibited the removal of copper from wastewater, potentially reducing the amount of copper residue in the crystals. However, at different pH and HCO3 levels... - The combined effects of these factors lead to more complex changes in zinc concentration in zinc-containing wastewater. For example, at pH 8.0, zinc removal is promoted, while HCO3-... - The removal of zinc from wastewater was inhibited at pH 9.0. When HCO3... - At a concentration of 30 mM, increasing pH significantly inhibited Zn removal. Furthermore, the concentrations of copper and zinc fluctuated within 7 minutes, indicating that release and re-adsorption may occur during struvite crystallization. Previous studies have also found that higher wastewater pH in struvite inhibits the incorporation of heavy metals (Cu, Zn, Cr, Pb, etc.), confirming the above conclusions. Specifically, Huang et al. reported that as the pH in solution increased from 8.0 to 10.5, the Cu, Zn, and Cr contents in struvite decreased from 4.8 mg / L, 6.8 mg / L, and 9.5 mg / L to 0.4 mg / L, 2.1 mg / L, and 7.4 mg / L, respectively, representing decreases of 91.7%, 69.1%, and 22%. To date, only As has shown the opposite effect at higher solution pH.

[0071] 1.3 Mutual Influence Based on Response Surface Method

[0072] The regression model coefficients and model variance analyses for Y1, Y2, and Y3 are shown in Tables 6, 7, and 8, respectively. The ANOVA results demonstrate the accuracy of the RSM model. The RSM results further confirm that the P recovery rate is mainly affected by pH and HCO3-. - The effect of pH on HCO3 (P<0.0001) and pH value on HCO3 - There was a significant interaction between them (p < 0.0001). At pH 8, the recovery rate of P increased with increasing HCO3- - The concentration of phosphorus increased significantly with increasing pH; however, when the pH increased to 9, the phosphorus recovery rate remained almost stable. Furthermore, heavy metal concentrations (0–100 µM) had no significant effect on phosphorus recovery (P > 0.05). These results further validate the accuracy of previous experimental results. Although the p-values ​​of Cu / Zn(B) and its interaction term with Cu+Zn(BC) are both less than 0.05, this indicates the existence of a significant model term. The relationship between pH and HCO3- - In comparison, these two model terms show the least change on the response surface. Therefore, except for extremely high metal concentrations, it can be assumed that they do not affect the recovery rate of P.

[0073] According to ANOVA results, the absorption of heavy metals (Cu and Zn) depends on pH, the molar ratio of Cu to Zn, the total content of Cu and Zn, and HCO3-. - (p < 0.005) These four factors. The interaction effects on copper adsorption capacity are, in descending order: AC (p < 0.0005), AB (p < 0.005), AD (p < 0.05). Copper adsorption capacity is positively correlated with total heavy metal concentration and copper-zinc ratio; pH and HCO3... - There is a negative correlation. As pH increases, HCO3... - The inhibitory effect on copper adsorption capacity was significantly reduced. Furthermore, the copper adsorption capacity predicted by RSM was consistent with the experimental results. All four factors significantly affected zinc adsorption (p<0.0001) (as shown in Table 7). Considering the presence of Zn, the order of interaction was BD (p<0.0001), CD (p<0.00001), AD (p<0.0005), and BC (p<0.001). Zinc adsorption was positively correlated with the total heavy metal concentration.

[0074] Table 6. Analysis of variance values ​​of copper adsorption in the RSM-CCD model.

[0075]

[0076] Table 7. Analysis of variance values ​​of zinc adsorption in the RSM-CCD model.

[0077]

[0078] Table 8. Analysis of variance values ​​of zinc adsorption in the RSM-CCD model.

[0079]

[0080] Figure 1 pH and HCO3 - Three-dimensional response surface plot of phosphorus recovery under concentration coupling effect. From Figure 1 It can be seen that when the pH increases from 8.0 to 9.0, HCO3... - As the concentration increased from 0 to 30 mM, the response surface exhibited a monotonically increasing trend, with the phosphorus recovery rate rapidly increasing from approximately 55% to >98%. The steepness of the surface was greatest near pH=8.5, indicating that HCO3- - The buffering effect and pH synergistically promoted struvite crystallization, consistent with the highly significant (p<0.0001) results in the AD section of Table 7.

[0081] Figure 2 pH and HCO3 - Three-dimensional response surface plot of Cu occurrence under concentration coupling effect. From Figure 2It can be seen that the Cu content continuously decreases with increasing pH and HCO3⁻, and decreases further at pH=9.0 and HCO3⁻. - The concentration drops to a minimum of 0.05 mg / L at the peak of 30 mM, the surface is concave, and the contour lines are more densely distributed along the pH axis, indicating that pH has a stronger inhibitory effect on Cu adsorption than HCO3. - This aligns with the main effect dominance of item A in Table 5, where the F value is 146.47 (p<0.0001).

[0082] Figure 3 This is a three-dimensional response surface plot of Zn occurrence under the coupled effect of pH and HCO3⁻ concentration. From... Figure 3 It can be seen that the Zn content is most sensitive to changes in HCO3⁻ in the pH range of 8.0–8.5. When HCO3⁻ increases from 0 to 30 mM, the Zn content drops sharply from 2.8 mg / L to 0.3 mg / L. When the pH is further increased to 9.0, the curve becomes flatter, and the decrease is less than 0.1 mg / L, indicating that high HCO3⁻ has fully converted Zn into soluble complexes such as ZnCO3(aq), inhibiting Zn from entering struvite. This is consistent with the calculated result that the φ(ZnCO3) ratio increases from 20% to 85%.

[0083] Figure 4 This is a graph verifying the agreement between the predicted values ​​of the quadratic multiple regression model and the experimentally measured values. Figure 4 It can be seen that the 30 validation points of Y1, Y2 and Y3 are evenly distributed on both sides of the 45° diagonal, with R² values ​​of 0.953, 0.981 and 0.989 respectively, and the maximum relative error is <7%. Moreover, the learned residuals are all within ±3, indicating that the model has high prediction accuracy and no systematic bias, and can be used for subsequent optimization.

[0084] Figure 5 Fiber diameter diagnostic plots: predicted vs. actual values ​​(a, b, and c); normal probability plots (d, e, and f); studentized residuals vs. predicted values ​​(g, h, and i); studentized residuals vs. number of runs (j, k, and l). From Figure 5 As can be seen, no funnel shape or obvious trend band appeared in any of the subplots, the residuals were normally distributed (Anderson-Darling p>0.05), and the order of operation was not correlated with the residuals, which proved that the model had good homogeneity of variance and independence, further confirming the reliability of the regression model.

[0085] In summary, this invention discloses a method for optimizing the production parameters of struvite based on the combined regulation of response surface methodology and Minteq, which has the following advantages: (1) The method for optimizing the production parameters of struvite based on the combined regulation of response surface methodology and Minteq disclosed in this invention can simultaneously regulate the pH according to the heavy metal concentration in the wastewater, maximize the phosphorus recovery rate by adjusting the bicarbonate concentration, minimize the occurrence of heavy metals in struvite products, and provide the best process parameters for safe and healthy application of agricultural soil fertilization; (2) The response surface methodology used in this invention has the characteristics of simple operation, high feasibility, and accurate prediction; (3) This invention can regulate the heavy metal concentration and alkalinity in phosphorus-containing wastewater of pig manure wastewater according to the concentration, so as to achieve high-quality phosphorus recovery from different types of wastewater; (4) This invention can regulate the parameters according to the measured wastewater with phosphorus recovery value, save the amount of sodium hydroxide added, and save a lot of costs.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing struvite production parameters based on the combined use of response surface methodology and Minteq, characterized in that, The method includes the following steps: (1) pH, heavy metal concentration, heavy metal molar ratio and bicarbonate concentration were used as independent variables, and Cu content, Zn content and phosphorus recovery rate were used as dependent variables. (2) Conduct single-factor experiments on the independent variables and analyze the range of variation of the extreme values ​​of the dependent variables to determine the experimental range required for the response surface model; (3) Based on the interaction relationship between the four independent variables and the three dependent variables, the experimental scheme design of the response surface method was implemented by using Design-Expert software and selecting the central composite design model. (4) A batch reactor was simulated using a beaker and a magnetic stirrer. The experiment was conducted according to the experimental plan, and three sets of parallel data were used for comparison. Each set of data included PO4 in the supernatant used to calculate the phosphorus recovery rate. 3- Concentration, residual Cu concentration in the supernatant used to calculate Cu content, and residual Zn concentration in the supernatant used to calculate Zn content; (5) Substitute the data from each group in step (4) into the Design-Expert software for data processing, and use pH, heavy metal concentration, heavy metal molar ratio, and bicarbonate concentration as independent variables to calculate the response value of the dependent variable using the following quadratic multiple regression model: ， In the formula: Y i X is the response value of the dependent variable. i ß is the independent variable. i β is the coefficient of the linear term, β0 is the constant term, β ij β is the coefficient of the interaction term. ii The coefficient of the quadratic term; (6) Perform variance analysis, P test, F test and significance test on the equation coefficients and regression model in step (5). A model with good predictive effect can be obtained after satisfying the following conditions: coefficient of determination > 0.9, corrected coefficient of determination > 0.9, model significance, and lack of significant value. (7) Select any set of parameters within the range of independent variables, repeat the experiment to analyze the degree of agreement between the model prediction variable value and the experimental value, so as to further verify the accuracy of the model. The range of independent variables is determined by the single-factor experiment in step (2). The specific method is as follows: First, perform single-factor scanning on the four factors of pH, heavy metal concentration, heavy metal molar ratio and bicarbonate concentration respectively, find the high-low boundary of each factor that causes the response to have extreme values, and then combine these four boundaries into the range of independent variables for subsequent modeling and verification. The model prediction variable value is the prediction result given by the quadratic multiple regression model for the three dependent variables of Cu content Y1, Zn content Y2 and phosphorus recovery rate Y3. (8) At the same time, the minimum value of Cu content, the minimum value of Zn content and the maximum value of phosphorus recovery rate are used as target values. The priority is set according to the importance of heavy metals and phosphorus recovery rate. The Optimization module of Design-Expert is used to optimize and obtain the best parameter combination for producing struvite.

2. The method according to claim 1, characterized in that, The unit for the heavy metal concentration is µM, the unit for the bicarbonate concentration is mM, the unit for the Cu content is mg / g, and the unit for the Zn content is mg / g.

3. The method according to claim 1, characterized in that, X1 is pH, X2 is the initial molar ratio of Cu to Zn, and X3 is the total concentration of Cu and Zn. The synthetic wastewater includes ammonium chloride, disodium hydrogen phosphate dodecahydrate, magnesium chloride hexahydrate, sodium bicarbonate, zinc sulfate heptahydrate, copper sulfate pentahydrate, sodium hydroxide, hydrochloric acid, nitric acid, and sodium chloride.

4. The method according to claim 1, characterized in that, The pH range is 8-9, the total heavy metal concentration ranges 40-80 µM, the heavy metal molar ratio ranges 0.5-2, and the bicarbonate concentration ranges 10-30 mM.

5. The method according to claim 1, characterized in that, In step (3), the surface method design scheme of the central composite design model is as follows: 。 6. The method according to claim 1, characterized in that, In step (4), the batch reactor includes a magnetic stirrer, a rotor, a beaker, and a pH meter; The reaction time in the batch reactor was 30 min. During the experiment, 2 mL of sample was taken every 1 min and the experiment was terminated with 2 µL of 10% hydrochloric acid. In the batch reactor, the PO4 content of the supernatant was measured using an ultraviolet spectrophotometer. 3- The concentrations of Cu and Zn in the supernatant were measured using atomic absorption spectrometry.

7. The method according to claim 1, characterized in that, In step (5), the expression for the quadratic multiple regression model is: Y1=19.27-4.9+3.79B+0.12C-0.15D-0.45AB-0.01AC+0.01AD+0.005BC+0.006BD-0.0002CD+0.32A 2 -0.08B 2 +0.00009C 2 +0.0009D 2 , Y2=-49.22+11.1A+0.06B+0.10C+0.35D-0.24AB-0.004AC-0.03AD-0.01BC+0.05BD-0.001CD-0.63A 2 +0.42B 2 -0.0001C 2 -0.002D 2 , Y3 = -14.05 + 3.14A - 0.11B - 0.006C + 0.09D + 0.002AB + 0.0004AC - 0.01AD + 0.002BC + 0.0006BD + 0.00002CD - 0.16A 2 -0.02B 2 -3.79E-07C 2 -0.0002D 2 。

Images