| High energy consumption and membrane fouling are the main obstacles on the membrane technology extending on waste water treatment. For the advantage on anti-fouling and backwashing, dynamic membrane (DM) technique is drawn more attraction recently. Through the theoretical analysis and experiments, the pre-coated dynamic membrane is utilized to optimize the process of cross flow filtration. The objectives of this paper are to reduce membrane fouling, increase permeate flux and simplify membrane washing with DM. The performance of DM was investigated in the long-term run of crossflow membrane bioreactor (CMBR) through the quality of permeate and flux decrement with time, in order to validate the feasibility of precoated DM in CMBR. The combination of CMBR and A/O (anaerobic/aeration) process was adopted to treat caplactam wastewater in this investigation. The denitrificability and organism activity was examined. In the late period of investigation, addition of power active carbon (PAC) into the reactor was expected to change the structure and characteristics of sludge flocs and improve the performance of the system.Firstly, the model of membraning process based on the particle force balance was founded through the analysis of the forces on the particles in the cross flow could calculate the permeate flux and cake thickness in steady-state and unsteady-state respectively. In addition, pores blockage and variety of curvature of ceramic tube were taken into account in this model. These parameters, such as particle size, crossflow velocity and transmembrane pressure, influenced the cake thickness and permeate flux in steady-state dramatically, while those including the transmembrane pressure, particle size and particle concentration of bulk related to the time needed to reach the steady-state. It was not shown a distinct linear relationship between the the pore clogging rate and particle concentration in the results. But pore clogging rate increased with the particles concentration and pressure clearly. The experiments, in which 1250 mesh and 600mesh kaolin were used as DM material were carried out under different conditions, i.e. diverse flow rate, pressure and bulk concentration. The results showed that the values from model simulation were up to the experimental data when thesystem coefficient of A;was taken in account as 2 and 10, corresponding to 1250 mesh and 6000mesh kaolin respectively. The continual decease of flux was found after the system was reach the steady-state.In order to determine the optimal membraning condition, membraning experiments were carried out under different pressure (0.1 > 0.2MP), cross flow rate (0.5 n I.On 1.5 ms"1) and particle concentration (0.5 ^ 0.75 ^ l.Og-L"1) with kaolin of 1250 mesh, 4000mesh, 6000mesh respectively. The property of DM was estimated by the quality of permeate and flux. The DM made under 0.1MP pressure, 1.5 m-s"1 initial flow velocity with 6000 mesh kaolin of membraning for 30min are considered as the best one. The DM filtration experiments illustrate that DM filtration was prior to the direct filtration obviously. But there is no evident advantage for DM in secondary effluent treatment as far flux as was concerned. Flow velocity over membrane surface was the most sensitive factor for flux among all these operational parameters. Either the higher flow velocity or the lower mixed liquid suspended solid (MLSS) concentration can lead to higher flux in pseudo steady-state while increasing of pressure all along cannot get the increasing of flux linearly. Beyond the critical pressure, there was no evident promotion in flux by increase TMP. The relation between flux and MLSS concentration can be expressed as J = aexp(bMLSS>). Compared to the national criterion on domestic and industrial waterreclaim, permeate of DM in secondary effluent and activated sludge can be reused.Analysis on the membrane resistances distribution showed that the external membrane fouling was the primary portion. In the secondary effluent treatment, cake resistance, Ref, was the largest part among the total resistance, while in activated sludge treatment the resistance of concentration polarization, Rp, was the largest. The flux was most sensitive to the crossflow velocity in both secondary effluent and activated sludge filtration. Particularly in sludge filtration, the portion of Rp to the total resistance was decreasing with the decrement of CFV, which related to the thickness of concentration polarization layer. Cross-flow velocity and MLSS concentration influence the resistance distribution obviously. According to membrane back-wash investigation, DM can prevent the membrane support from internal fouling comparing with microfiltrationinorganic membrane. According to the chemical washing and inorganic foulant analysis, it was found that the main inorganic foulant were calcium and magnesium compound. Comparison on either solo chemical lixiviation showed that alkali lixiviation was better than acid lixiviation. And lixiviation sequence of lixiviation in alkali (O^mgL'1 NaOH) for lh + lixiviation in acid (0.2 mg-L"1 HC1) was prior to the conversed order. After the reduplicative backwashing, the regenerated flux of DM was up to 90%, while that of direct filtration was only 60%. By the protection of DM, the flux of membrane support can regenerate back to 45-55 %of the virgin flux after hydraulic back-wash. And the fouling accumulation almost cannot be detected during the DM filtration.The combination of DM and CMBR was brought out fist time. DM was tested in CMBR for long run period ranging over 230d, which was designed as anoxic/aerobic process to treating synthetic caprolactam wastewater. As system launched, it was found that MLSS or MLVSS concentration collapsed and oxygen use rate (OUR) was decreasing from 6.84mgO2-L"1h"1 to 3.96mgO2-L~1h"1 first and increasing 8.44 mgC^-L^-h'1 at 12th d then. The characteristics of permeate and supernatant were measured and the results illustrated that the concentration of CODcr in supernatant maintained below lOOmg-L"1, while that in permeate below 20 mg-L'1. There was no evidence shown that the contribution of membrane rejection for nitrogen removal. Soluble microbial product (SMP) was accumulated in the reactor for the rejection of membrane and degraded after being rejected in reactor for long enough. The observed sludge productivity, FObs> of CMBR was 0.24gMLSSgCOD"', which was much smaller than the conventional processes. OUR was measured after inhibitor addition. It was found that activities of the heterotrophic organisms, Nitrosomonas group, Nitrobacter group were decreasing because of low F7M ratio and accumulation of SMP. But it did not effect on the nitrogen removal efficiency. The concentration of TOC was up and down within the run period and reached the peak at 60 mg-L"1 at 35th d and then leveled off, and it kept constant along in the permeation. In the last 50d, PAC was added into the reactor. SOUR still remained about 1.2 mgO2gMLSS"1h'1 while the mean particle size of sludge was up to 35urn from 13.5um caused by the aggregation of PAC. Thepossible decrement of oxygen convert efficiency induced by the augmentation of floes, the TN removal declined faintly. But the system was not so sensitive to the addition of PAC in summary.The data of flux showed that flux decreased along even after the pseudo steady-state and leveled off then. The operational parameters affect the flux decrement evidently. It took 89h for flux decreasing to 20L-m"2-h"1 under 0.2MP pressure, while it did 278h under 0.05MP, which indicated that the higher flow velocity caused slower flux decrement rate. The span of flux decreasing to 20 Lm"2-h"' was 85h, 136h, 232h under flow velocity 0.5, 1.0, 1.5 m-s"1 respectively. If the concentration of MLSS was higher, the flux decrement rate was bigger and the proportion of resistance of concentration polarization, Ref, in total resistance was higher. Under the same filtration condition, DM filtration was better than direct filtration as standing-time and flux regarded. The flux of DM benefited from the addition of PAC, either as flux or on standing time were concerned, system operation got more satisfactory, relating to the specific resistance of cake optimized by PAC. The technology of DM combining with CMBR was proved excellent in membrane fouling reduction and membrane washing simplification and would be promising well in practice.
|
PDF Full Text Request