Search

The yield of a distribution network represents one of the indicators most commonly used by water distributors to account for the state of the volumes of losses over a given period or to measure the evolution of these losses from one year to the next, generally speaking. However, contrary to what might be thought, complex situations and very differing interpretations are to be found on the back of the result of the calculation of the volumes of water lost in the distribution network. A study of drinking water network yield was carried out in Paris in 2001 so as to get a better grasp of the developments recorded since 1987. The consequences of the changes that have taken place during this period have each been measured: alterations to the organisation of the water system, developments affecting the structure of the supply system, technical improvements in the wholesale water metering systems and overall reduction in volumes consumed. [ABSTRACT FROM AUTHOR]

In conventional drinking water treatment processes, the amount of the filter backwash water covers nearly 3% of the total production. To reduce the water loss and waste discharge in the conventional drinking water treatment process, the Macao Water Supply Co. Ltd (SAAM) plans to recover the backwash water by Microfiltration (MF) membrane process as water resources are scarce and new environmental regulations are mandated in Macao. Generally, the filter backwash water from the conventional water treatment plant with sedimentation process is recycled to the source water to be treated again under certain conditions, and the sedimentation tank discharges most of the sludge. However, it is possible to recycle the backwash water directly to the inlet for direct filtration process due to the limitation of inlet turbidity. This paper describes how to apply MF technology to treat the backwash water of the direct filtration plant and to optimize MF operation. Without pre-treatment of the settling basin for backwash water, the operation of the MF pilot plant is proved to be stable and the permeate quality can meet EU drinking water standards. The pilot study shows that it is both economically and technically feasible to adopt MF technology in backwash water treatment. The main parameters to test MF process include flux, chemical cleaning duration and transmembrane pressure (TMP). They are 150-200 L/m2.h, 20 days and <1 bar respectively. The estimated cost including O&M and investment for a 1320-1760 m3/d backwash water treatment plant is USD 0.126-0.168/m3. [ABSTRACT FROM AUTHOR]

The authors studied the performance of a hybrid MF membrane system combining activated carbon adsorption and biological oxidation for the removal of turbidity, natural organic matter, ammonia and manganese from river water in pilot-plant experiments. In this system, part of the soluble organic matter was adsorbed by the activated carbon. Ammonia and soluble manganese were oxidized by microorganisms that were concentrated in a submerged tank. In preliminary experiments, it became clear that powdered activated carbon (PAC) addition improved membrane permeability, because PAC has a large volume of macro-pores, which can adsorb larger molecular-weight organic matter. In the pilot-plant experiment, however, it was difficult to slow the rate of membrane permeability loss by PAC addition only. This may be due to biological fouling, caused by operating the system at a high recovery rate in order to maintain a high concentration of microorganisms in the submerged tank. Therefore, backwashing using hypochlorite solution (50-200 mg/l of Cl2) was carried out every few hours. This simple chemical washing was effective in preventing bio-fouling of the membrane. [ABSTRACT FROM AUTHOR]

Membrane filtration processes have become widely used for low turbidity water sources that require particle removal and disinfection. In cases where the feed water contains high turbidity levels or high fouling tendencies, pre-treatment has been required to allow the membranes to operate efficiently. Submersion membranes have allowed direct treatment on many of these water supplies but operate under vacuum. This presents limitations on operating flux and design conditions for incorporating into existing facilities. The treatment unit designs are also very expensive to implement for small water treatment applications. A recently developed pressure driven ultrafiltration (UF) membrane is being tested on several different water sources and has demonstrated exceptional operation on high turbidity feed water sources. The module design includes a single potted end with the membrane fibers looped on the bottom to allow draining of solids from the module. The UF membrane filters outside to inside and has been operated on feed water turbidity levels up to 200 ntu. The availability of a pressure driven hollow fiber membrane that can efficiently filter high suspended solids water sources can provide a cost effective solution for many small to medium sized water supplies. Conditioning the feed water with a coagulant or direct filtration of high turbidity feed water can be used depending on the organic material present without settling or other clarification process. This eliminates the need for additional structural pretreatment and reduces overall system cost and size. Current limitations include the module size that makes this cost prohibitive for very large systems. Larger membrane systems where high-suspended solids are present in the feed water have been utilizing immersion membrane technologies more and more over the past few years. These systems typically operate under a vacuum and can be used for large water treatment plants as their design allows large filtration modules such as Memcor''s CMFS filter block at 880 m3/hr. This allows significant capital cost reductions and allows membrane filtration to be competitive to conventional filtration technologies for virtually all filtration applications. This paper will review the economics and performance of both pressure driven and immersion membrane systems for small to large water supply systems on feed water containing high-suspended solid levels. Actual testing data for the new pressure driven hollow fiber UF membrane and immersion membrane systems will be provided on different water supplies including unconditioned surface water, filter backwash water and coagulated surface water. [ABSTRACT FROM AUTHOR]

Recent research indicated that hydrophobic microporous hollow fiber membrane Contactor could be used to separate volatile contaminants from wastewater efficiently. The factors that influence the effect of separation of NH3/H2O are investigated carefully. It is found that the average diameter of micropores, porosity and gas penetration rate of polypropylene (PP) hollow fiber membrane are maximized synchronously at draw elongation 260%. The modules assembled with this PP hollow fiber membrane have good performance for separation of NH3/H2O. It is found that the overall mass transfer coefficient (K) only is dependent on the temperature and flux of ammonia solution and independent of the pressure and concentration of ammonia solution. Concentration of absorbing acid influences slightly the overall transfer coefficients K while the absorbent supply is sufficient. Under this condition, the relationship between transfer coefficient and raw ammonia water flow rate can be described using the Sieder and Tate equation. K is directly proportional to v1/3 (v stands for the flow rate of ammonia solution). The removal rate of NH3 was about 96-98% in the lab. Further industrial tests have been carried out individually in catalyst and chemical fertilizer factories to treat wastewater containing ammonium, whose concentration ranged from 1000-10,000g/m3. Wastewater treatment amount was 9.6 m3/day and 24 m3/day in catalyst and chemical fertilizer factories respectively. Average removal rate of NH3 was 93.5% during 5 months of continuous operation in the catalyst factory. The fouling is not heavy because the wastewater containing ammonia did not pass through the membrane wall. It is efficient to get rid of membrane fouling by washing the membrane with dilute acid and the cleanout period was 3 days. PP hollow fiber membrane can be produced by the melt-spinning and drawing method. This paper also investigates the relationship between processing condition of PP hollow fiber membrane and its structure and performance. [ABSTRACT FROM AUTHOR]

The preliminary study investigated the effect of pre-ozonation of synthetic raw water (bentonite and humic acid) on membrane fouling. The results showed that humic acid, not bentonite, was the dominant factor in membrane fouling. When the synthetic raw water was pre-treated using ozone and activated carbon, stable operation was possible at twice the flux of membrane filtration without pre-treatment. Thus, ozone and activated carbon reduced the concentration of organic matter (humic acid), leading to an increase in flux. Secondly, an MF membrane filtration device with ozone tolerance was attached downstream of the ozone reactor so that residual ozone could reach the membrane surface. When water was treated with MF membrane filtration in the presence of ozone, the flux was stable at 4 m/d. The membrane filtration resistance increased when ozone was not injected, and decreased when it was injected. This phenomenon was repeatable. In the presence of ozone, organic matter that could accelerate fouling as decomposed and converted to less fouled materials, resulting in a suppression of fouling as well as facilitating removal of the fouling layer during back washing. [ABSTRACT FROM AUTHOR]

Natural organic matter (NOM) present in raw water can not only impart color to water, but can also cause health risks associated with disinfection by-products (DBP). The most common DBPs found in drinking water are trihalomethanes (THMs) and haloacetic acids (HAAs), which are formed when NOM reacts with chlorine or chlorine based disinfectants. In order to prevent the formation of DBP''s, the US EPA has introduced a two stage Disinfectants-Disinfection Byproduct Rule (D/DBPR). Stage 1, finalized in November of 1998, established the maximum contaminant level (MCL) at 0.080 mg/L for total trihalomethanes (TTHMs) and 0.060 mg/L for five haloacetic acids (HAA5). This will be followed by Stage 2 (Final Rule due in 2002) of the D/DBPR, which may set lower contaminant levels at 0.040 mg/L and 0.030 mg/L for TTHM''s and HAA5, respectively. Enhanced coagulation has been identified as the best available technology for meeting the requirements of the D/DBP Stage 1 Rule for Total Organic Carbon (TOC) reduction and the removal of disinfection byproduct precursors. With enhanced coagulation, NOM, color and TOC reduction is achieved using a higher coagulant dosage than would be utilized for turbidity removal. The pH of the raw water is also commonly optimized to maximize process efficiency. The application of immersed ultrafiltration membranes using enhanced coagulation has been successfully applied for disinfection byproduct precursor (DBP), color and TOC removal for drinking water applications. With this process, a single tank coagulation-ultrafiltration process replaces the coagulation-flocculation-sedimentation-filtration stages of a conventional treatment plant. Enhanced coagulation ultrafiltration has three stages: 1) rapid mix, 2) coagulation/adsorption of NOM, and 3) ultrafiltration. A high solids concentration is maintained in the membrane tank to promote the adsorption of organics onto the small settling flocs. Subsequently, the barrier characteristics of the ultrafilter are used to separate flocculated particles, precipitates, and colloidal particles. Powdered activated carbon (PAC) can also be used in combination with enhanced coagulation, or alone to reduce disinfectant byproduct precursors by addition to the rapid mix stage upstream of the ultrafiltration membranes. Compared to conventional treatment, this novel method of water treatment results in higher color and TOC removal and requires less coagulant and PAC. The use of lower chemical dosage results in significantly less treatment residuals and reduced disposal costs. The system also has a small footprint since it is designed with a shorter hydraulic retention time as it is only necessary to form a floc that exceeds the membrane pore size. This paper presents the application of immersed ultrafiltration membranes using enhanced coagulation and PAC processes for color and TOC removal as well as the reduction of DBP precursors. It also presents pilot data on a number of applications where the process is being effectively used and compares it with performance data reported for conventional treatment facilities using enhanced coagulation. [ABSTRACT FROM AUTHOR]

The City of McAllen, Texas, with the assistance of CH2M HILL, has pilot tested an integrated membrane bioreactor (MBR)/reverse osmosis (RO) treatment train to reclaim municipal wastewater to a quality suitable for use as a new drinking water supply in the process called indirect potable reuse. Previous testing by the City (Phase 1) demonstrated the applicability and cost of microfiltration (Memcor and ZeeWeed systems) to enhance the quality of secondary effluent for subsequent treatment by RO and the feasibility of a membrane bioreactor system (ZenoGem) to produce RO feedwater directly from minimally processed sewage. Phase 2 testing, reported on in this paper, is designed to demonstrate reliable operation of MBR/RO treatment for processing screened, degritted sewage and that the effluent from such a train can meet all federal primary and State secondary drinking water regulations and comply with anticipated State requirements for indirect potable reuse. Results show the ZenoGem process to be reliable, require minimal operator attention and maintenance, produce an effluent that can be processed by RO with little fouling and that easily exceeds the City's current effluent discharge requirements relative to BOD, TSS and ammonia. The ZenoGem permeate quality exceeds RO feedwater criteria for turbidity and silt density index and RO system performance confirms minimal membrane fouling by particles. However, the high calcium hardness and phosphate levels in the City's wastewater (and ZenoGem permeate) caused mineral precipitation within the RO system when operated at higher recoveries. Precipitation can be controlled, however, by increased acidification of the RO feedwater. [ABSTRACT FROM AUTHOR]

The in-line coagulation/ultrafiltration (UF) combined process has been studied in order to improve membrane performance for the treatment of surface water with high organic content. Ultrafiltration experiments were performed using cellulose derivative hollow-fibre membranes with ferric chloride as coagulant. The efficacy of adding coagulant just before the membrane module depends on the filtration mode, the process configuration and the module design. In dead-end mode, permeation flux of 70 l/hm2 can be maintained, but high backwash frequency (low recovery) is necessary to prevent plugging. Cross-flow filtration was effective in reducing fouling and allows operation at higher permeation flux. In cross-flow filtration, the feed-and-bleed configuration (single stage continuous process) appears promising to reduce backwash frequencies: for a recovery of 95% and at a permeation flux of 120 l/h.m2, quasi-steady transmembrane pressure can be maintained with a low fouling rate (10 kPa in 400 min.). In cross-flow filtration, the use of curved modules with Dean vortices reduces the fouling rate by a factor of 2 compared with a conventional straight module. [ABSTRACT FROM AUTHOR]

The microfiltration and ultrafiltration processes are considered as matured membrane processes that are well established in industrial practice. Nevertheless, the main obstacles of their further development in the new competitive implementations are the economical problems. The key economic factors are permeate flux and energy consumption. However, although the cross-flow systems enable us to attain higher flux, it is usually very expensive. The high energy is consumed to maintain circulation velocity of the retentate that is sufficient for sweeping out the retained component from the membrane surface. Moreover in the case of cells separation the high intensity of the fouling and low cake permeability makes it necessary to apply additional efforts, such as backflushing, backpulsing, promoters of turbulence, vibrations, ultrasounds and many other. Therefore, dead-end systems are still quite competitive with cross-flow, especially for diluted (less than 0.5% of dry mass) suspensions or solutions. Cell separation with membranes is one of the most vivid problems for modern biotechnology, wastewater and water treatment. Membranes offer mild process conditions and high selectivity of separation. This enables us to solve a variety of problems such as cell culturing, fractionation, concentration, purification and sterilisation. The selected cells may be precisely separated from other components of broth and subsequently directed into the reaction space again in good conditions to ensure a quasi-continuous mode of operation. Moreover, membranes enable us to attain high efficiency of the bioconversion by removal of all product and inhibitors directly from the bioreactor. This is the reason for the huge interest in cell separation with membranes. The idea of the paper was to present the new concept of flux enhancement for cell separation on membranes. This concept lies in taking advantage of the specific rheological nature of biopolymers, which are the main foulants. The biopolymers retained on the membrane surface (i.e. on the top layer) can be applied as a lubricant for the cells that can settle on such a ‘movable layer’. As is shown, further in the paper, the thickness of the moving layer is lower and the flux is greater. The common movement of the cells and gel layer is very convenient from the cells integrity point of view. However the hydrodynamic conditions always play an important role in cross-flow systems; the resistance of ultrafiltration membranes may be reduced much more when compared with more open microfiltration membranes. [ABSTRACT FROM AUTHOR]

The vacuum membrane stripping process (VMS) for removing a VOC (trichloroethylene) from water was studied in a polyvinylidene fluoride (PVDF) hollow fiber membrane module containing 187 fibers with a length of 0.59 m. The porous PVDF asymmetric hollow fiber used was prepared by the wet phase inversion technique. The feed solution containing trichloroethylene (TCE) was passed through the lumen of the PVDF hollow fiber. Vapors of TCE and water were transported through the pores of the membrane into the shell side maintained at a sub-atmospheric pressure, and condensed in cold traps. The effects of down-stream pressure, solution temperature, feed concentration, and feed flow rate on TCE removal, TCE permeation flux, water permeation flux and TCE concentration of the permeated solution, as well as the mass transfer coefficient were investigated. The down stream pressure and solution temperature were identified as the major factors to control VOC removal and TCE concentration in the permeate stream. A small amount of the permeated solution with higher TCE concentration was obtained at the temperature of 50°C and pressure of 80 mmHg when about 90% TCE was removed from the water. [ABSTRACT FROM AUTHOR]

The River Trent runs through both Severn Trent and Anglian Water regions, and has been identified as a potential new water resource for both companies. It contains treated effluent from 4 million people and industry in the Midlands. As part of an ongoing project, comparing conventional water treatment processes with membrane technologies for treatment of the River Trent to potable standards, pre-treatment processes for ultrafiltration are being compared. Depth filtration gave the best particulate removal, whereas the package filters performed as per the manufacturers' specifications. Package filters tended to increase the small particulate fraction between 2-5 μm, so although these may remove problems caused by large particles they could aggravate pore blocking and membrane penetration related fouling. There is a direct trade off between the capital cost of pre-treatments vs. increased running costs (higher TMPs, more frequent cleaning, extensive cleaning requirements). [ABSTRACT FROM AUTHOR]

A chromium smelting process can result, under oxidising conditions, in the production of a dust which when slurried and pumped to a waste dump heap is likely to pollute the environment. South Africa is a large producer of chromium metal and the problem of the release of hexavalent chromium (Cr6+) to the environment is a serious concern. The department of Water Affairs and Forestry in South Africa monitors all environmental discharges and regulate the levels of toxic chemicals and pollutants. The ‘Tubular Filter Press’ a South African developed sludge dewatering system was proposed for the recovery of dust in this effluent. The filter press was operated for the dewatering of waterworks sludges and promises to be successful for this application. A single tube pilot study shows that chromium discharge to the environment can be eliminated. Feed solids concentrations of between 40 and 50 g/l were fed to the tubular filter resulting in cake solids concentrations in excess of 50%. The operating pressure of the system was between 200 and 300 kPa and the flux reduced to 200 L/m2.h during the filtration cycle. The filtration was modelled using a variable pressure, internal cylindrical compressible cake filtration model and operating parameters for a full scale plant were proposed. [ABSTRACT FROM AUTHOR]

At the Heemskerk water treatment plant of N.V. PWN Water Supply Company North Holland the first full scale integrated membrane system in the Netherlands is realized. The 20 Mm3/year (15 MGD) ultrafiltration (UF)/reverse osmosis (RO) plant servicing IJssel Lake water has been in operation since the second half of 1999. Disinfection is a major objective of the direct surface water treatment plant without chemical disinfection. After UF indicator organisms such as E. coli and Clostridia spores were absent. Integer UF membranes removed MS-2 phage spikes completely (<5.4 log), severely compromised membranes removed 2.7 log only. RO showed a high although incomplete removal of MS-2 phages (up to 4.8 log). Therefore integrity monitoring is a critical issue. Before installation, vacuum testing was applied with a criterion of 10 kPa/min. Many standard modules failed to pass the test. During operation UF is monitored by 1 μm particle counting with 0.05 μm particle counting as a quality control. A 3 log removal can be established. At this moment for RO conductivity measurement is applied enabling a 2 log removal monitoring. Short term monitoring of conductivity will be replaced by sulfate, increasing the monitoring range by 1 log unit. The total monitoring range of 5-6 log units easily satisfies the maximum disinfection requirement for the membrane steps for viruses of 3.6 log. To determine a wider range of disinfection requirements PWN together with Norit Membrane Technology developed the Spiked Integrity Monitoring (SIM®) System. Spiking with 14000 PAC particles/ml combined with 1 μm particle counting increased the potential for log removal monitoring of UF to 5.8 log units. Additionally PAC particle spiking increased the potential to detect UF fiber compromising. During the first 200 days of full scale operation, system performance is excellent. Membrane integrity control is very satisfactory and SIM®-System integrity control has proved to be a valid technique to establish a broader range of required inactivation capacities. [ABSTRACT FROM AUTHOR]

A cleaning protocol that effectively removes fouling from hollow fiber UF systems without excessive use of chemicals, product water or (long) down time is needed. Cross flushing with UF feed water has been reported to increase the net flux of hollow fiber systems by reducing the frequency of backwashing, the consumption of permeate and the system down time. In this study, the flux restoration achieved in a vertical and horizontal UF system employing an intermittent water and water/air cross flush were compared. The flux restoration in the vertical UF system was not improved by the addition of air to the water flush and a maximum flux restoration of 82% was achieved, irrespective of the presence of air. Similarly, in a horizontal ultrafiltration system, a maximum flux restoration of 82% was also achieved with a water flush (v = 1.63 m/s). However, the addition of air to the water flush decreased the flux restoration to 40% at the highest water/air ratio (33% air). Low flux restoration in the horizontal system was attributed to residual air in the module after cross flushing. Flushing with water alone (v = 1.63 m/s) yielded a wall shear stress of 16 Pa compared with 130 Pa and 279 Pa in the liquid film surrounding the air slugs in the horizontal and vertical UF system, respectively, with a water/air ratio of 2:1. Despite the high shear force on the cake layer accumulated when air was added to the system, the maximum flux restoration was 82% both with and without air. This was attributed to the fact that it was the filtration mechanism and not the shear force on the cake layer that limited flux restoration during cross flushing. To improve the flux restoration that can be achieved by the cross flushing process, the filtration mechanism must be manipulated to minimize blocking filtration and induce cake filtration from the beginning of each filtration cycle. [ABSTRACT FROM AUTHOR]

The glass ball inserted membrane module presented in this study is designed to enhance filtration in a flat sheet membrane. Three different modes of filtration experiments were conducted and compared to demonstrate flux enhancement due to the presence of glass balls: a normal dead-end filtration, a vortex flow filtration, and an enhanced vortex flow filtration using glass balls. In the case of enhanced vortex flow filtration the permeate flux was found to be three times as large as that of the dead-end filtration, and two times larger compared with the vortex flow filtration. In addition, the flux decline was observed to be relatively low. The effect of the amount of glass balls on the permeate flux was also investigated by changing the values of glass ball volume fraction from 0.059 to 0.356. It has been observed that the permeate flux shows a maximum value of the volume fraction of 0.119. For the glass ball inserted membrane module, the permeate flux tends to increase with the feed flow rate. [ABSTRACT FROM AUTHOR]

The use of powdered activated carbon in combination with ultrafiltration membranes is attracting increasing interest for the removal of organic compounds in drinking water treatment. The overall adsorption efficiency of this hybrid membrane process strongly depends on the reactor configuration and its operating conditions. Identification of the operating conditions yielding optimum carbon performance can be facilitated by the use of mathematical models describing the adsorption process. In this study, the effect of various designs and operating parameters on the efficiency of the adsorption process is discussed using an adsorption model previously developed and verified by the authors. This discussion includes the effect of filtration time, membrane reactor volume, carbon dosing procedure, carbon dose and carbon particle size on the adsorption removal of two selected micropollutants and dissolved organic matter. [ABSTRACT FROM AUTHOR]

Two pilot-scale wastewater treatment systems - direct membrane separation (DMS) and membrane bioreactor (MBR) systems - were designed and constructed in order to investigate the feasibility of membrane filtration technology for domestic wastewater treatment and water reuse. A submerged-type hollow-fiber microfiltration (HF-MF) membrane module with pore size of 0.1μm was employed to build each pilot system. The systems were tested using low- and high-strength domestic wastewaters and the system performance was continuously monitored for a long period to compare filtration characteristics and effluent quality in each system. The MBR system showed much better performance than the DMS system in terms of filtration characteristics and effluent quality. Even though the mixed-liquor suspended solid (MLSS) content in the MBR system was much higher than that in the DMS system, the MBR filtration resistance was much lower than the DMS filtration resistance. The DMS system was not able to remove dissolved organic matter, which seemed to be a major component of membrane fouling. The MBR effluent quality such as COD, BOD, TOC and T-N was more stable and better than the DMS. In the MBR process, the organic removal efficiency remained more than 95% regardless of fluctuation in influent qualities. The effluent quality of both systems was satisfying the legal standards for water reuse in Korea. Rejection of pathogenic microorganisms by membrane filtration was also investigated. [ABSTRACT FROM AUTHOR]

Pilot studies were conducted at Liu-Du water treatment located in Keelung, Taiwan. Raw water was first treated by conventional coagulation- sedimentation-filtration (CSF) processes, and then pumped into the pilot plants. For the pilot plants, both ozone/GAC and ozone/UF processes were operated simultaneously for NOM removal comparisons. Both systems were controlled at the same operating conditions with ozone dosing of 4.5, 6.0 and 9.0 mg/L. The results demonstrate that both ozone/UF and ozone/GAC processes can enhance NOM removal after the CSF process. More NOM was removed by the Ozone/GAC process because ozone broke the large molecules into smaller molecules, which increases the GAC adsorption potential. However, the ozone/UF process provides higher removal capacity in turbidity. In both processes, ozone dosing was not proportional to NOM removal of each process and 6.0 mg/L is found to be the optimal dosing. Owing to increasing adsorbability by ozonation, the ozone/GAC process is a very competitive process in NOM removal when associated with the CSF process. [ABSTRACT FROM AUTHOR]

Pilot testing of low-pressure membrane filtration (microfiltration and ultrafiltration) was performed on surface waters at two different pilot locations: the Alameda County Water District (ACWD) in Fremont, CA and the Medina River outside San Antonio, TX. Clarification pretreatment and adsorption pretreatment were evaluated at each site for impacts on operational membrane performance and on removal of dissolved organic compounds including total organic carbon (TOC), UV254, color and precursors to disinfection by-products. At the ACWD test site, studies were also conducted with spiking of taste and odor-causing compounds. In both pilot studies, clarification pretreatment was demonstrated to achieve some apparent improvement in operating efficiency. For example, an apparent increase in sustainable flux between 25% and 50% was demonstrated for clarified feedwater operation as compared to raw feedwater operation at the ACWD pilot site. Removal of dissolved organic compounds was achieved primarily through the pretreatment with clarification and/or addition of powdered activated carbon (PAC); however, some reduction of TOC and color was achieved via membrane filtration presumably due to removal of particulate/colloidal constituents. [ABSTRACT FROM AUTHOR]

Selective removal of heavy metal ions from industrial wet-process phosphoric acid (WPA) was studied, using the two liquid membrane systems: Hybrid Liquid Membrane (HLM) and Aqueous Hybrid Liquid Membrane (AHLM). Monothiophosphinic acid (Cyanex 302) in kerosene (for the HLM) and Polyvinylsulfonic acid (PVSA) aqueous solution (for the AHLM) were used as a liquid membrane (LM), separated from the feed and strip streams by neutral or cation-exchange membranes. On application of the HLM separation system, the permeation from the feed to the strip phase followed the selectivity sequence: Zn >> Cd > Cu with all types of membranes used. Results show that the selective separation of Cd and Zn from the WPA may be achieved by the HLM with cation-exchange membranes, using a single set-up installation at continuous processing. Copper is accumulated in the Cyanex 302, poisoning the carrier. So, this carrier and HCl as a strip solution can not be recommended for the HLM processing of WPA. Using the AHLM separation system, heavy metals such as cadmium and copper, were successfully recovered from the WPA: average fluxes (in mol/m2sec) to the strip phase were 3.7 × 10-7 and 6.7 × 10-7. Satisfactory dynamic selectivity to copper (&ap;16) was found. The transfer of iron and zinc species to the strip phase was below detection levels during the 144 hrs of the AHLM experiment. Contamination of the feed and strip solutions by PVSA was not observed during the 500 hrs of processing. The membranes were completely regenerable and used repeatedly. In comparison with the HLM and other liquid membrane systems, the AHLM can potentially provide many operational, economic and environmental advantages, some of them being higher transport rates, low carrier losses and low contamination of the feed and strip solutions by water-soluble polymers, long membrane lifetime, ambient temperature, low energy requirements and chemical consumption, no special requirements for membrane resistance in organic solvents, and fewer limitations on the variability of membrane pore sizes, etc. [ABSTRACT FROM AUTHOR]

Ozone was incorporated into an ultrafiltration system to produce higher quality reclaimed water from domestic laundry wastewater. Characteristics of the wastewater for initial washing waste were 488~2,847 mg/L COD, 62~674 mg/L MBAS, and 38~857 mg/L SS. The wastewater was contacted with ozone in a 10L storage tank and circulated through the membrane module for inner pressurized cross-flow filtration. The concentrate was returned back to the contact tank. The membrane used in this experiment was hollow fiber polysulfone UF membrane with MWCO 5,000 and 10,000. It has an effective filtration area of 0.06m2. The experiment was carried out in two phases with either continuous or intermittent ozone injection. For intermittent ozone injection, the mode of injection interval was changed to 5 min./5 min. and 5 min./10 min. for injection/idling. Ozone was dosed at the concentration of 1.5 mg/L. The permeate quality of UF (MWCO 5,000) was 57 mg/L as COD and 5 mg/L as MBAS at continuous ozone injection with removal of 95% in COD and 96.9% in MBAS. Using UF with MWCO 10,000, it was 93.7% and 95.5% of COD and MBAS, respectively. And using intermittent ozone injection, the removal efficiency was 93% in COD and 93.5% in MBAS without any deterioration in COD and MBAS removal. It could reduce the treatment cost. Using ozone injection, fouling of the membrane was also controlled by increasing organic degradation. The flux of UF (MWCO 5,000 and 10,000) was 0.13 and 0.20 m/d for 3 hour filtration (TMP 40&ap;45 kPa) without ozone injection. It was increased to 0.18 and 0.24m/d by ozone injection. The reclaimed water quality could be estimated well enough to reuse for rinsing purposes. [ABSTRACT FROM AUTHOR]

The feasibility of ultra- and microfiltration depends strongly on the achieved net flux. Direct treatment of surface water frequently results in low net fluxes and high cleaning frequencies. Experiments have been conducted at constant flux on surface water with an ultrafiltration pilot plant with direct filtration, in-line coagulation and pre-coating (EPCE) with ferric hydroxide flocs. The aim was to control and reduce the rate of fouling. With the use of a pre-coat at the beginning of the filtration cycle a stable ultrafiltration process was obtained. This is contrary to the use of in-line coagulation and direct filtration which resulted, due to a very high fouling potential of the feed water, in high fouling rates at low fluxes. The result of this exploratory research is an important step towards a higher feasibility of micro- and ultrafiltration. [ABSTRACT FROM AUTHOR]

Combining the Kubota Tank-Submerged type Ceramic Membrane filtration (TSCM) system with ozonation and powdered activated carbon (PAC) treatment, a new water treatment system, the advanced TSCM system, has been developed. It is a compact, maintenance-free and high performance system for water treatment plants in the small and medium size cities. In this paper, experimental results using 5.5m3/day demonstration plants are shown, and treatment performance is discussed mainly focussing on removal of disinfection by-products. Turbidity, color, total iron, total manganese and ammonia nitrogen were removed to below their detection limits. Accumulation of suspended solids including PAC at a maximum concentration of 8000 mg/L in the membrane tank increases biological degradation activity in the TSCM system. Furthermore, it was confirmed that the advanced TSCM system was able to effectively eliminate disinfection by-products and other organic contaminants in raw water. 75% of T-THMFP removal rate was obtained at only 10 mg-PAC/L dosage without the effect of ozone dosage, while the PAC treatment at 50 mg-PAC/L dosage prior to coagulation results in T-THMFP removals ranging from 40 to 60% (The Adoption Guideline of Advanced Water Treatment Facility, 1988). Therefore, it is concluded that the advanced TSCM system saves on the amount of PAC and is more cost-effective than the other conventional water treatment systems due to efficient utilization of the adsorption capacity of PAC. [ABSTRACT FROM AUTHOR]

This paper presents a concept for relating complex solute mixture characteristics to membrane fouling using field flow fractionation. Field flow fractionation has traditionally been used in separation science as a tool for characterizing aquatic system components in terms of size based on Stokes-Einstein theory. This work begins to assess the possibility of using flow field flow fractionation (FlFFF) as a tool for ‘fingerprinting’ complex solute mixtures and obtaining effective mixture properties, such as effective size, effective diffusion coefficient, and interaction potential between solutes and ultrafiltration membrane materials. The effect of applied field and changes in solute ionic strength are demonstrated for polystyrene sulfonate latex colloids. Quantitative analysis is also considered for the residence time distribution (commonly referred to as a fractogram) of solutes from FlFFF measurements. The results here suggest that FlFFF may effectively fingerprint a complex solute mixture based on solute characteristics and membrane solute interactions. [ABSTRACT FROM AUTHOR]

The City of San Diego was awarded a grant from the Bureau of Reclamation to investigate the feasibility of using Membrane Bioreactors (MBRs) for water reclamation. Based on the findings of the first phase of the project (Adham et al., 1998), the project team concluded that a parallel comparison of commercially available MBR systems needed to be evaluated at pilot-scale. Two submerged MBR systems were evaluated at the Aqua 2000 Research Center in Escondido, California. The project was designed to evaluate the MBR performance treating municipal wastewater and the feasibility of using the MBR permeate as a feed source for thin film composite reverse osmosis (RO) membranes. The first part of the project was dedicated to operating both MBRs in a nitrification and denitrification mode. After completion of Part 1 of the project, both MBR systems were retrofitted and operated in a nitrification only mode. Throughout both parts of the study, the effluent from each MBR was fed to two separate, single stage RO pilot systems. Both MBR systems showed high BOD removal with values below the detection limit, and significant TOC reduction. The effluent turbidities from the MBRs were consistently less than 0.1 NTU. Both MBRs also produced a high quality effluent that could be used by thin film composite RO membranes with minimal fouling. [ABSTRACT FROM AUTHOR]

The design standards of the iron and manganese removal system by membrane filtration were investigated. The membrane filtration after pre-chlorination could remove iron completely and could remove around 70% of manganese. In the case of manganese concentration being more than 0.05 mg/l, the membrane filtration could remove them completely after the deposition in the oxidation tank. The concentrations of iron and manganese were reduced more than 90% by the oxidation tank and were reduced until under the detection limit after the membrane filtration. The economic comparisons between the rapid sand filtration system and this membrane filtration system in a small capacity and a middle capacity were investigated. The 15 years total costs of the membrane filtration system were estimated to be smaller than or around equal to the rapid sand filtration system in both capacities. [ABSTRACT FROM AUTHOR]

The development of the ‘ Fyne’ Process, from an operator's perspective, for the removal of humic colour from water is described. The use of cellulose triacetate nanofiltration membranes in both tubular and spiral wound configurations has been a major advancement in the treatment of highly variable coloured upland water sources in Scotland. Membrane technology has enabled full compliance with all drinking water legislation even at the remote island sites. Operational and cost comparisons of nanofiltration with conventional treatment technologies for colour removal are discussed. [ABSTRACT FROM AUTHOR]

Dead-end ultrafiltration has proven itself as a technique for reclamation of backwash water of sand filters and as a pre-treatment step for spiral wound reverse osmosis. A direct result of dead-end filtration is a decreased flux caused by the accumulation of material in the membrane pores and on the membrane surface. Different cleaning techniques are used to remove this accumulated material. Recently a new technique has been introduced, the AirFlush. This technique makes use of air to create higher turbulence as compared to a water flush. At Delft University of Technology research has started into the fundamentals of the combined air- and water-flush. First a series of experiments has been carried out to determine the different flow patterns, followed by experiments to determine which air- and water-velocities give the best cleaning. Finally head loss experiments have been done to get information about the energy dissipated in the system. The results of the head loss experiments have been compared to the theoretical head loss calculated with the theory of heterogeneous two-phase flow. [ABSTRACT FROM AUTHOR]

Three UF-systems were tested for the drinking water treatment of karstic spring water. The pilot plants were operated in Marmagen, which is located in North-Rhine Westfalia (Germany). One of the two tested capillary UF-systems (in-out mode) was backflushed with filtrate and with filtrate and hydrogen peroxide. The other one could be flushed with air and raw water (forward flush) before a backflush was performed. The third system was an immersed UF-system (out-in) where negative pressure on the filtrate side is applied. The capillary UF-systems were operated at high flux and long backflush intervals when turbidity was low. At high turbidities the permeability declined and could only be restored by chemical cleaning. For this particular water the immersed system showed also a permeability decline, which was less pronounced. The permeability recovered in normal operation, when the turbidity declined. Using SEM and EDX the fouling could be mainly attributed to small particles (well below 3 μm). Qualitative element analysis showed that silica is the main component of the formed cake layer. [ABSTRACT FROM AUTHOR]

The efficacy of ultrafiltration (UF) and microfiltration (MF) membranes was assessed for the concentration of actual waterworks sludges using crossflow tubular membranes operated at constant trans-membrane pressure. The MF membrane gave higher initial fluxes than the UF membrane but after 10 min of filtration the flux value and its decline tended to be very similar for both membranes operating under the same conditions. All membranes gave permeate product water of <0.2 NTU and <100ppb coagulant at all times. For both membranes mechanical cleaning, with sponge balls, was at least as effective as acid chemical cleaning, indicating that no significant permanent internal fouling occurred for these membrane materials. Hydraulic resistance data indicated a significant difference in the dynamic layer resistance between the two membranes. Whilst the UF membrane had a hydraulic resistance 3.7 times that of the MF membrane, the dynamic layer formed on the UF membrane during operation displayed a maximum hydraulic resistance almost nine times lower than that of the MF membrane operating under the same conditions. Correlation of cake resistance R versus feed solids concentration C for all the data generated for t>0 demonstrated reasonable agreement with the expression R∝ca where a=0.37 in the current study. This trend has been recorded in previous reported studies, a varying between 0.33 and 0.62 depending on sludge dewaterability. [ABSTRACT FROM AUTHOR]

Comparisons between rejection characteristics of natural organic matter and inorganic salts by nanofiltration in a very low operational pressure (<0.5 MPa) are demonstrated here. A nanofiltration pilot plant has been operated at the Tamagawa water purification plant, which is located downstream on the Tama River, Japan. The molecular weight distribution and concentration of natural organic matter that account for trihalomethane formation potential, and concentrations of various inorganic salts have been measured in the raw water from the Tama River, microfiltration pretreated effluent and nanofiltration production during 6-months operation of the pilot plant. This paper discusses the water quality and rejection mechanism of natural organic matter and inorganic salts. In addition, the steric-hindrance pore (SHP) model and the Nernst-Planck equation have been used to evaluate membrane pore structures and rejection of ionic solutes, respectively. [ABSTRACT FROM AUTHOR]

Low permeate flux due to clogging is the major challenge of hollow fiber microfiltration systems in wastewater treatment because of the high content of suspended solids. In this research, a new crossflow microfiltration module to overcome this problem was studied. The module consisted of a housing, a cylindrical membrane and a bundle of hollow fibers. The nominal pore size of the cylindrical membrane was 5 μm and that of the hollow fibers was 0.1 μm. The cylindrical membrane worked as a screen for pretreatment to remove those particles larger than its pore size and therefore to reduce the solid loading to the hollow fibers. Extensive experiments were conducted under various transmembrane pressures, influent concentrations, and crossflow velocities with polydispersed suspensions prepared with ISO Test Particles of size ranging from less than 0.5 μm to 10 μm. The quantity and quality of the filtrate were monitored over time. The results have shown that the cylindrical membrane worked effectively in reducing solid fouling, and therefore increasing permeate production for the hollow fiber microfiltration system. [ABSTRACT FROM AUTHOR]

In order to achieve an efficient and economical advanced drinking water system, the relationship between operating conditions, membrane filtration flux and water qualities was examined by using a pilot plant consisting of ozonation facility, ozone resistant microfiltration (MF) module and granular activated carbon (GAC) tower. It was found that keeping residual ozone of more than 0.3 mg/L on the membrane surface is necessary in order to obtain a high membrane filtration flux. Withy ozone dosage of 3 mg/L by ejector and filtration by dead-end mode, this new system provided a high membrane filtration flux of more than 5 m3/m2/day. Quality of the produced water from the GAC tower was meeting the drinking water standard. [ABSTRACT FROM AUTHOR]

Experimental studies on chemical and biological pretreatments in membrane filtration processes were carried out to removal manganese contained in raw water and to prevent membrane fouling due to manganese. Two types of the pretreatment reactors, i.e. the fluidised-bed and fixed-bed configurations, were compared in the biological pretreatment experiments. New synthetic media (tubular polypropylene, I.D. 3 mm, O.D. 4 mm, length 5 mm) were used in all three experiments as a manganese-oxidising catalyst. The chemical pretreatment using sodium hypochlorite was effective in manganese removal and controlling membrane fouling; more than 0.8 mg-Cl2/L of chlorine dose was necessary to bring the manganese concentration from 0.4 mg/L in raw water to less than 0.05 mg/L. The biological pretreatment for manganese removal required a long start-up period of more than 40 days. The fixed-bed biological pretreatment was superior in manganese removal and in control of membrane fouling to the fluidised-bed biological pretreatment, which showed wash-out of the attached bacteria resulting in membrane fouling. The linear velocity and the empty-bed retention time required for the treatment of 0.14 mg-Mn/L in the fixed-bed biological pretreatment was 206 m/d and 8.0 minutes, respectively. [ABSTRACT FROM AUTHOR]

The effects of operational modes on the removal of a synthetic organic chemical (SOC) in natural water by powdered activated carbon (PAC) during ultrafiltration (UF) were studied, through model simulations and experiments. The removal percentage of the trace SOC was independent of its influent concentration for a given PAC dose. The minimum PAC dosage required to achieve a desired effluent concentration could quickly be optimized from the C/C0 plot as a function of the PAC dosage. The cross-flow operation was not advantageous over the dead-end regarding the SOC removal. Added PAC was re-circulated as a suspension in the UF loop for only a short time even under the cross-flow velocity of gt; 1.0 m/s. The cross-flow condition did not contribute much to the suspending of PAC. The pulse PAC addition at the beginning of a filtration cycle resulted in somewhat better SOC removal than the continuous PAC addition. The increased NOM loading on PAC which was dosed in a pulse and stayed longer in the UF loop could possibly further decrease the adsorption rate. [ABSTRACT FROM AUTHOR]

Membrane filtration allows safe retention of microorganisms when treating filter backwash water from conventional drinking water filters. The permeate of the membrane plant can thus be reused to produce drinking water. The benefits are a higher yield of the drinking water treatment plant and a minimised wastewater production. This paper discusses the results of a pilot study, cost data and full-scale operation experiences concerning the application of ultrafiltration to treat filter backwash water. The effectiveness of ultrafiltration was assessed with regard to flux, cost and permeate quality. [ABSTRACT FROM AUTHOR]

The authors have proposed a novel biofilm-membrane reactor (BMR) in which nitrifying biofilm is fixed on the surface of the rotating membrane disk. With this reactor, both strict solid-liquid separation and oxidation of ammonia nitrogen can be simultaneously performed. Based on the knowledge obtained in the previous experiments with the bench-scale equipment, a pilot scale study was conducted using river water at a water purification plant. The results obtained in the pilot study can be summarized as follows: (1) A pilot scale BMR worked successfully even with the water from an actual drinking water source. By implementation of pre-treatment (coagulation and sedimentation), the filter run could be continued for more than 8 months without any chemical washing. (2) The sponge cleaning developed in this study was found to be very effective. This indicates that the accumulated cake resistance is dominant in the practical situation. (3) Sufficient nitrification was observed when water temperature was high. (4) In addition to the nitrification, biological oxidation of AOC and manganese can be expected with the BMR. In this study, both AOC and manganese concentration in the permeate decreased to a level less than 10 μg/L. [ABSTRACT FROM AUTHOR]

DBP control by nanofiltration was studied. Nine DBPs including THMs were selected as target compounds. Reagent DBPs and DBPs formed as a result of chlorination of NOM contained in river water, were applied to nanofiltration to examine DBPs removal. Humic acid was used to investigate DBP precursor control. Nanofiltration was conducted for about one day for each experiment using two 2-inch membrane modules. NF showed high performance for DBP precursor removal but little effect on the removal of DBPs themselves, especially THMs. In the experiment of DBP precursors removal, rejections were 98 to 100 percent for most of the DBPFPs. For NOM in river water, rejections of THMFPs were 96 to 99 percent. This shows nanofiltration is effective even at low concentrations of NOM. Nanofiltration can remove almost all organic substances with the apparent molecular weight of greater than 2,000 daltons. Molecular weight distribution patterns of the permeate were different depending on NF. THMs showed decreasing rejection as filtration proceeded because of adsorption on the membrane surface. THM rejections at 24 hours after filtration started were about 10 and 30 percent in Module-A and B, respectively. Haloacetic acids showed relatively higher rejections. It is recommended, from the viewpoint of THM control, that NOM removal using NF is much preferable to the removals of THM, formed by chlorination, using NF. [ABSTRACT FROM AUTHOR]

In this paper, we briefly discussed the utilization of the PP HFM contactor for removal of gaseous ammonia directly from mixed gas in ammonia manufacture in order to find an economic way to substitute conventional soft-water washing absorption technology. The Hollow Fiber Membrane (HFM) contactor has very high elimination (99.9%) under optimum operation conditions. 5.1 Nm3/(m2.h) gas treatment ability can be achieved, as outlet and inlet ammonia concentration are 0.02 g/Nm3 and 20.0 g/Nm3 respectively and gas stay-time in the contactor is only 0.04 s when acid supply around the membrane is sufficient. The results of industrial tests indicated that this technology could be used to eliminate mixed gas containing ammonia and replace conventional soft-water washing and absorption techniques. [ABSTRACT FROM AUTHOR]

Membrane filtration plants for drinking water typically use pressure decay testing in conjunction with particle counting and turbidity to monitor membrane integrity. Pilot plants offer the capability of monitoring permeate quality with both intact and intentionally compromised membranes. We compare data from a particle counter, a pressure decay test and a laser turbidimeter on pilot plants from two different manufacturers of microfiltration membranes. [ABSTRACT FROM AUTHOR]