• Membrane Journal
• /
• v.14 no.2
• /
• pp.149-156
• /
• 2004

This work focused on the degradation of natural organic matter (NOM) present in lake water using a combined pkotocatalysisimicrofiltration (MF) process. The system performances were investigated in terms of organic removal efficiency and membrane permeability. The addition of iron oxide particles (IOP) into the photocatalytic membrane reactor improved initial NOM removal by sorption, but during photocatalysis the removal efficiency was reversed, probably due to the scattering of UV light by IOP. The modification of TiO$_2$ surfaces by IOP deposition was conducted to enhance the photocatalytic NOM removal efficiency. A minimal amount of Impregnation of IOP on TiO$_2$ surfaces was required to prevent the light scattering effect as well. The coating of MF membranes with IOP helped to improve the NOM removal efficiency while sorbing NOM by IOP. Regardless of tile operating conditions and particles addition examined, no significant fouling was occurring at a flux of 15 L/$m^2$-h during entire MF operation.

• Membrane Journal
• /
• v.18 no.1
• /
• pp.53-64
• /
• 2008

Fouling behavior of polyethylene capillary membranes was examined by measuring the flux of bentonite colloidal suspensions through the cross flow micro filtration. The membrane fouling was due to the three mechanisms: the cake formation on the membrane surface, the standard pore blocking and the complete pore blocking by particles. These mechanisms were simultaneously responsible for the membrane fouling, being significantly governed by the cake filtration. In the total fouling at $1.0kg/cm^2$ TMP condition, the complete blocking was 3.36%, the standard blocking 3.18% and the cake filtration 96.05%. For 1000 ppm feed solution, the complete blocking was 1.71% compared with the standard blocking of 1.90% and the cake filtration of 96.39%. And 96.14% of the total fouling was generated at the initial period of filtration. The cake filtration effect was larger on $0.34{\mu}m$ pore membrane than on $0.24{\mu}m$ pore membrane. With the increase in cross flow velocity, the component fouling decreased by 10.20%, and the ratio of pore blocking to total fouling increased.

• Journal of Korean Society of Water and Wastewater
• /
• v.29 no.2
• /
• pp.233-241
• /
• 2015

Polytetrafluoroethylene (PTFE) membrane has high resistance to chlorine, which is a great advantages in chemical cleaning to recover water flux during membrane processes in drinking water systems. A humic kaolin water with approximately 4 mg/L of DOC and 10 NTU of turbidity was prepared as a feed water. Coagulation pretreatment with or without settling was applied. The coagulation with settling showed the greatest water production. The reduced flux was effectively recovered by NaOCl cleaning, i.e., 21% recovery by 50 mg/L of NaOCl cleaning and 49% recovery by 500 mg/L NaOCl cleaning. The images of SEM and AFM analyses were corresponded to the water flux variation. However, when the floc was accumulated on the membrane surfaces, the efficiency of NaOCl cleaning was substantially limited. In addition, dynamic contact angle became greater after cleaning, which indicates changes in characteristics of fouling layer such as surface hydrophobicity. Proper cleaning technologies during enhanced backwash using NaOCl would expand application of PTFE membranes in drinking water systems.

• Journal of Korean Society of Water and Wastewater
• /
• v.18 no.6
• /
• pp.770-777
• /
• 2004

Three pilot-scale membrane systems were operated using lake water as influent in this study. Microfiltration (MF) membrane with pore size of 0.01 m was used in Systen I of which filtration mode was set at constant pressure of $1kgf/cm^2$. Ultrafiltration (UF) membranes with molecular cutoff (MWCO) of 80,000 and 13,000 were used in System II-1 and II-2, respectively. Constant flow mode was applied at the range between 0.7 and $1.5m^3/m^2{\cdot}d$ (average of $1.1m^3/m^2${\cdot}d) for System II-1 and between 0.37 and $1.65m^3/m^2{\cdot}d$ (average of $1.18m^3/m^2{\cdot}d$) for System II-2. In System I, the flux changed from $1m^3/m^2{\cdot}d$ to $0.2m^3/m^2{\cdot}d$ during the operation time of 5 months. System II showed recovery of 94% under the allowable maximum pressure of $3kgf/cm^2$ during the same operation period. From these results, the efficient operation was observed in constant flow mode with respect to filtration time and recovery. Average filtrate turbidity showed 0.0071 NTU in System I and 0.0054 NTU in System II, which implied that high turbidity removal was obtained in both MF and UF systems with no significant difference between MF and UF. From the fact that membrane flux depends largely on membrane type and operation mode, a guideline of optimum design and operation should be suggested for application of membrane systems to full scale water treatment.

• Applied Chemistry for Engineering
• /
• v.17 no.3
• /
• pp.310-316
• /
• 2006

In this study, activated carbon in a powder form was used to remove dissolved ammonia which causes a fouling smell in water. Since the adsorption capacity of common powder activated carbon is not high enough, we prepared powder activated carbon deposited on an acid solution to enhance the adsorption capacity. The acid-impregnated activated carbon was applied on the surface of porous fibril support ($10{\sim}50{\mu}m$) by which adsorption and separation processes take place simultaneously by varying effective pressure. As the result, the ammonia removal efficiency is above 60% in the mixing process which is 10~15% higher than general powder activated carbon. From the result of an experiment on the pure permeable test of a dynamic membrane, its transmittance is 400~700 LMH (liter per hour), indicating that the prepared membrane works as a microfiltration membrane. Therefore, it is expected that the membrane prepared in this way would improve the efficiency of water treatment than conventional membranes.

• Membrane Journal
• /
• v.25 no.2
• /
• pp.180-190
• /
• 2015

PVDF (polyvinylidene fluoride) nanofiber has the advantages such as excellent strength, chemical resistance, nontoxic, non-combustibility. Flat membranes with 0.3 and $0.4{\mu}m$ pore size respectively, were manufactured by PVDF nanofiber, and then each spiral wound module was prepared with them. A woven paper was not included in preparing the module with $0.3{\mu}m$ pore size; however, it was included the module with $0.4{\mu}m$ pore size. The permeate fluxes and rejection rates of the two modules were compared using pure water and simulation solution including kaolin and humic acid. The recovery rate and filtration resistance were calculated after water back-washing. In addition, the effect of flow rate and trans-membrane pressure on treatment efficiency and filtration resistance were investigated for the spiral wound module with $0.4{\mu}m$ pore size.

• Korean Membrane Journal
• /
• v.10 no.1
• /
• pp.26-32
• /
• 2008

The Gongji stream water of Chuncheon city was filtrated by 2 kinds of tubular alumina ceramic MF membranes with periodic $N_2$-back-flushing. $N_2$-back-flushing time (BT) was changed in $0{\sim}50$ sec at fixed filtration time (FT), or back-flushing period, of 4 min for NCMT-5231 membrane ($0.05\;{\mu}m$). Then, FT was changed in $0{\sim}32$ min at fixed BT of 40 sec for NCMT-7231 $0.1{\mu}m)$). In the viewpoints of total permeate volume ($V_T$), dimensionless permeate flux ($J/J_0$) and resistance of membrane fouling ($R_f$), the optimal $N_2$-BT was 50 sec, which was the longest BT, at 4 min FT for NCMT-5231. It means the longest BT was the most effective to minimize the membrane fouling, and we could acquire the most $V_T$. But the optimal FT for NCMT-7231 was 16 min in the viewpoint of $V_T$, and was 8 min in the viewpoints of $J/J_0$ and $R_f$ at fixed BT of 40 sec. The rejection rates were excellent as $80.6{\sim}96.6\;%$ for turbidity, $35.2{\sim}58.4%$ for $NH_3$-N, $16.3{\sim}45.2%$ for T-P and $16.3{\sim}45.2%$ for $COD_{Mn}$. However, the rejection rate of T-N was very low as $2.7{\sim}13.4%$ and it of TDS below 6.1%.

• Membrane Journal
• /
• v.19 no.1
• /
• pp.34-46
• /
• 2009

Effect of backflushing on the membrane fouling for polyethylene capillary membranes was examined by measuring the flux of $Al_2O_3$ colloidal suspensions through the cross flow microfiltration. In the comparison of with and without the application of backflushing, the hydraulic resistance to permeate flow of the suspension was less with backfluslng, but the Increasing rate in permeate resistance was higher. Regardless of backflushing, the cake filtration was dominant at the initial period of filtration with backflushing, being followed by the pore blocking. And at steady state, the fouling mechanism is almost governed by the cake filtration model. On the contrary, the pore blocking preceded the cake filtration in the initial stage of the original membrane before backflushing. And irrespective of backflushing, the ratio of cake filtration to total fouling increased, compared with that fur before backflushing. For the membrane with $0.24{\mu}m$ pore size, the permeate resistance was higher than that of $0.34{\mu}m$ pore size membrane. but the ratio of cake filtration was smaller than that of large pore membrane. In comparing the ratio of each fouling component to the total fouling for the case with backflushing pore blocking was 7.8% and cake filtration was 92.2%. for the case without backflushing, total fouling was composed of 9.6% pore blocking and 90.4% cake filtration.

• Membrane Journal
• /
• v.8 no.4
• /
• pp.203-209
• /
• 1998

The cutting oil emulsion microfiltration was carried out on dead-end call and crossflow systems equipped with 0.22 $\mu$m GVHP Millipore and 0.2 m stainless steel Mott microfiltration membranes, respectivdy. The oil drop size in the emulsion was varied from 0.07 to 0.22 $\mu$m. Cake filtration(CFM) and standard pore blocking models(SPBM) were applied to predict the permeation flux. The permeation fluxes of 0.01 vol% oil emulsion followed CFM for dead-end system very well under the condition of 400 rpm and below 100 kPa. The SPBM was, however, suitable for the permeation flux at 400 rpm and above 150 kPa. The oil layer on the membrane surface was destroyed and reproduced repeatedly as operating pressure was suddenly changed from 60 to 200 kPa, and then returned to 60 kPa. Also, we estimated the critical entry pressure(CEP) which is changed from CFM to SPBM, and CEP for dead-end system was around 100 kPa. The CEP increased from around 100 to 150 kPa for the crossflow system as the oil concentration increased from 0.01 to 0.03 vol% when Reynolds number was 7080.

• Membrane Journal
• /
• v.15 no.1
• /
• pp.52-61
• /
• 2005

Permeate flux decline in a microfiltration was analyzed by measuring the permeability of bentonite colloidal solution through polyethylene capillary membranes. The flux decline with time was due to the growth of cake layer on the membrane surface and to the pore blocking by particles. As the time approaches to steady state, the permeate flux is almost controlled by the cake filtration model. Faster flux decline at high trans-membrane pressure was attributed to the formation of denser packed cake layer and pore blocking. The ratio of permeate flux to the initial permeate flux, J/J₁, decreased with increasing the trans-membrane pressure, from 45% for 0.5 kg/sub f//㎠ to 38% for 2.0 kg/sub f//㎠. In comparing the ratio of each fouling component to the total fouling for the 0.5 kg/sub f//㎠ TMP condition, complete blocking was 23.4%, standard blocking was about 14.6% and cake filtration was 62.0%, respectively. Permeate flux through the membrane increases with cross flow velocity, and the effect of the variation of velocity is more significant at 1.0 kg/sub f//㎠ rather than at 2.0 kg/sub f//㎠ of the operation pressure. Permeate flux for the membrane having the average pore diameter of 0.34 ㎛ was higher than that for the membrane of 0.24 ㎛ pore size, with the higher flux with the low concentration of feed. On the operation using the membrane of 0.34 ㎛ pore, the pore blocking in the low concentration of 200 ppm is negligible relative to the pore blocking in the 1000 ppm feed.