| The traditional drinking water treatment processes worldwide include following stages: coagulation(mixing and flocculation), sedimentation, filtration and disinfection. Among them, flocculation is an indispensable part of the process, which has nonlinear and hysteresis effect characteristics. The efficiency of flocculation has important implication on sedimentation and filtration process. Several parameters, such as the effluent turbidity, permanganate index, ammonia nitrogen, et.al of settled water, are used to characterize efficiency and control the coagulant dosage in most water plants. However, this practice has both time and feedback hysteresis effect, which means it fails both in quickly reflecting the efficiency of flocculation and feedback and controlling the dosage. Firstly, coagulant and the optimum dosage were selected by coagulation and sedimentation beaker test. Then detected the floc by using the industry digital camera, and effectively identified, treated and obtained the floc performance parameters(floc number, equivalent size and fractal dimension) at the transitional section based on machine vision system by applying micro-vortex flocculation technology with independent intellectual property of our school. Investigated how the change of flocculation time or coagulant dosage influences floc number, equivalent size and fractal dimension; studied on the correlation between floc number, equivalent size, fractal dimension and the quality of settled water(turbidity and ξ potential), and analyzed the relationship between the fractal dimension and equivalent diameter. The test results could provide reference data for resolving problems, such as hysteresis effect and dosage accuracy, et.al, in coagulant dosing control process.The raw water used in the test was taken from Kongmu Lake of East China Jiaotong University, which ammonia nitrogen and total phosphorus index is relatively high because of school’s domestic sewage. The values of parameters reflecting water quality are listed below: water temperature is 8~26 ℃, p H value is 6.37~7.15, ξ potential is-32.08~-23.12 m V, turbidity is 8.812~28.165 NTU, potassium permanganate index(CODMn) is 5.220~23.560 mg/L, ammonia nitrogen(NH3-N) is 4.560~12.050 mg/L and total phosphorus(TP) is 0.250~2.760 mg/L. The main test results and conclusions are as follows:1. Coagulation and sedimentation beaker testThrough series of coagulation and sedimentation beaker tests, the removal rate of turbidity, CODMn, NH3-N and TP in the flocculation process with three coagulant(polyaluminium chloride(PAC), ferric chloride(Fe Cl3) and aluminum sulfate(Al2(SO4)3)) was studied. It turns out that the optimal coagulant is PAC and the optimal dosage of 18 mg/L.2. The study of floc detection in flocculation process based on machine visionThe test detected the floc by using the industry digital camera at the transitional section, and obtained floc performance parameters, including the number, equivalent size and fractal dimension, and object detection identification process by using computer and software of Floc Processor-Microsoft Visual Studio, and each parameter values are within normal range, indicating that the use of machine vision systems for image preprocessing, detection and recognition, analysis of mathematical morphology and floc feature extraction has good feasibility and practical operation, which could provide technical support for the follow-up study of floc performance parameters in the flocculation process.3. Effect of flocculation control condition on floc performance parametersThis stage of the test is to investigate the impact of flocculation time and coagulant dosages on floc performance parameters:(1) Effect of different flocculation time on floc number, equivalent diameter and fractal dimension when the PAC dosage of 18 mg/L is studied, and the results are as follow: with the reduction of flocculation time(water flow rate increased), the number of floc increases at first, reduces then, and finally increases again; while the equivalent size and fractal dimension increase with an more substantial speed at first and then decrease.(2) Effect of different coagulant PAC dosages on floc number, equivalent diameter and fractal dimension was studied when the flocculation time of 17.1 min, and the results could be concluded as follow: With the increment of dosage, the number of floc shows a growing trend, while the floc equivalent size and floc fractal dimension tend to increase at first and then decrease gradually.4. The relationship between floc performance parameters and the quality of settled waterThis stage of the test is aiming to study the relationship between performance parameters, including floc number, equivalent diameter, fractal dimension, and the quality of settled water(turbidity and ξ potential) by changing PAC dosage while the flocculation time was remain 17.1 min.(1) The turbidity was decreasing with the increment of floc number during the process of increasing the dosage to the optimum dosage, which showed negative correlation between them. After that, the turbidity was increasing with floc number increasing, showing positive correlation when the dosage was increased further, the relation of 2 xxy(10)-(28)42522.5195076.000444.0 and correlation coefficient R2 of 0.982. With the increase of PAC dosage, the variation trends for both floc equivalent diameter and fractal dimension and turbidity of settled water were negative correlated, correlation coefficient R2 of 0.851 and 0.875, the relation between equivalent diameter, and turbidity of 2 xxy(10)-(28)11547.784602.1042369.4, the relation between fractal dimension and turbidity of 2 xxy(10)-(28)43912.3801026.4473575.12. The turbidity of settled water decreased when floc equivalent diameter and fractal dimension were increasing.(2) With the increment of PAC dosage, the trends of floc number and ξ potential value of settled water are positively correlated, the correlation coefficient R2 of 0.868, the relation of 2 xxy(10)-(28)88725.29208.100858.0.The ξ potential value gradually changed from negative to positive when the floc number increased. The relationship between floc equivalent diameter variation trend and that of ξ potential value of settled water was positively correlated, the correlation coefficient R2 of 0.879 and the relation of 2 xxy-(10)(28)75676.1409273.681045.3. The ξ potential value changed from negative to positive as the floc equivalent diameter increased. The relationship between floc fractal dimension and ξ potential presented a linear positive correlation when the value of ξ potential varies from negative to positive, the correlation coefficient R2 of 0.942, the relation of xy-(28)68.33598.17, and fractal dimension increased in this process. However, the relationship between floc fractal dimension and ξ potential presents a negative linear correlation when the positive value of ξ potential increased, the correlation coefficient R2 of 0.954, the relation of xy(10)-(28)432.146.7, the value of ξ potential became large and fractal dimension value decreased in this process.5. The relevance between floc fractal dimension and equivalent diameterThis stage of the test is to investigate the relevance of floc fractal dimension and equivalent diameter, while the flocculation time was remain 17.1 min(influent flow rate of 5 m3/h), PAC dosage of 20 mg/L. The vary range of floc equivalent diameter is 0.417~2.989 mm, and the average value is 1.345 mm. The vary range of floc fractal dimension value is 1.05~1.96, and the average value is 1.66 during the test. The test result shows that there is a good relevance between the floc fractal dimension and equivalent diameter, the relation of 0224.1 d D13147.0f×(28), which is in line with an exponential relationship, and the correlation coefficient R2 is 0.826. When floc equivalent diameter less than about 2 mm, fractal dimension value increased with the increament of the equivalent diameter. However, when floc equivalent diameter larger than 2 mm, the value of fractal dimension decreased gradually deviated from the curve when the equivalent diameter increased. Key Words: micro-vortex flocculation, machine vision, floc, equivalent diameter, fractal... |
PDF Full Text Request