By J. S. Baker and L. Y. Dudley, PermaCare® ABSTRACT The use of microbicides, particularly chlorine may be advantageous to operation but can also exacerbate biofouling problems. Microorganisms subjected to low levels of biocides often exude large amounts of Extracellular Polysaccharides (EPS) as a protection; it is this EPS material that forms the biofilm. This paper examines the causes and effects of obstinate biofilms in membrane elements. In these cases, problems of increased differential pressure have proven difficult to correct during routine cleaning cycles. Consequently, regrowth rates, as indicated by this differential pressure for such biofilms, have been rapid. Experimental data has been taken from more than a hundred membrane autopsies around the world. These autopsies have been undertaken when such problematic biofilms are encountered. They confirm that biofilms formed in process and membrane systems are comparable except that membrane biofilms contain greater numbers of fungi. Some operating systems have reported the cessation of chlorination with a significant reduction of biofouling. This paper considers possible causes for this. It also considers the increasing use of proprietary non-oxidising microbicides. Conclusions are that biofouling is endemic within membrane systems, yet many systems operate satisfactorily even with a biofilm. Foulant layers can be ‘conditioned’ or ‘hardened’ by the repeated use of cleaning programmes. There is a strong case for alternating cleaners and biocides as used by the cooling water treatment industry. INTRODUCTION Biofouling and its control remains a major operating problem for many reverse osmosis (RO) plants, particularly those in tropical and sub-tropical regions. Presented at Membranes in Drinking and Industrial Water Production, Amsterdam, September 1998 Figures 1 and 2 — Biofouled membrane (top) and cartridge filter (bottom) The sequence of biofouling and its effect on membrane systems is well documented and has been the subject of some informative reviews.1 Biofouling occurs despite the use of pre-treatment systems and the addition of disinfectants such as chlorine (Figures 1 and 2). Biofilms occurring in membrane systems may cause severe loss of performance and the use of costly cleaning procedures to maintain output and quality. The Reprint R-790 Biofouling in Membrane Systems – A Review fouling is often so severe that acceptable operation cannot be maintained and membrane replacement is needed. Bacteria are capable of colonising almost any surface and have been found at extreme conditions such as temperatures from –12°C to 110°C and pH values between 0.5 and 13.2,3 Bacteria embedded in a biofilm are more resistant to biocides than the same bacteria in a dispersed state. This applies to many toxic substances commonly used in membrane processes.4,5 The important influences on the rate of biofilm development are the carbon:nitrogen:phosphorus ratio, temperature, redox potential and pH.6 Extracellular polysaccharides (EPS), substance responsible for the slimy nature of biofilms and a product of the microorganisms themselves, are produced to a greater or lesser extent by a variety of genera such as Pseudomonas. Despite a great quantity of published information, some of the questions posed by Characklis in 19817 whilst discussing biofilms in cooling water treatment applications remain unanswered. Some of these are particularly appropriate to membrane applications: Figure 3 — Biofilm blockage of feed channel spacer In a spirally wound element, there is the possibility for some sections of the flow channel to become blocked such that the water flow will be concentrated in other parts of the feed channel. Channelling problems also arise in hollow fibre bundles when the individual fibres become bound together by foulant. These conditions have often been observed during autopsies of severely fouled membranes and are difficult to clean as the cleaning solution fails to reach the fouled membrane leaves or fibres. For this reason, it is essential that biofouling is detected and dealt with during its early stages as biofilm removal is more difficult when it occurs to this extent. • How do biofilm properties influence energy losses? • How do biofilm properties change when biocides are applied continuously? • How does the inorganic content of the water influence biofilm properties? If these questions could be answered with any degree of accuracy, it would be a good start in the formulation of anti-biofouling strategies for membranebased systems. Such an understanding would be a clear and positive advance in the understanding of biofilm control. Channelling causes rapid salt concentration in the affected areas. This leads to the precipitation of sparingly soluble salts such as calcium carbonate and calcium sulphate, the latter being a specific problem with sulphuric acid dosed feedwaters. For example, a mixed foulant recovered from a blocked sea water membrane feed channel was found to be 83% calcium sulphate. Calcium sulphate would not ordinarily be a scale problem in a low recovery sea water plant, but this can occur as a result of biofilm blockages. Biofouling of the feed channels and spacers in spiral wound elements (Figure 3) often causes a significant increase in differential pressure (∆P), which is often detected before normalised product flow or quality is affected. Such fouling phenomena in any channel are known to influence frictional energy losses. With membrane biofilms, increased hydraulic resistance (i.e., ∆P) has often been attributed to the visco-elastic properties of the biofilm itself. Clearly these properties are dependent upon the biofilm composition, which is itself dependent upon environmental factors. The physical structure of biofilms found in membrane systems can be compact and ‘gel’ like or ‘slimy and adhesive’ with some consisting of a large ratio of polysaccharide slime to viable microorganisms. Other biofilms contain higher proportions of microorganisms. Between 106 and 108 colony forming units (cfu) of bacteria per cm2 of membrane are common. PermaCare have performed many destructive autopsies from RO plants around the world. In many of these, the plant operators have experienced difficulty in recovering plant performance by routine cleaning so these autopsies were undertaken to characterise the foulant and evaluate alternative cleaning procedures. The objective of this paper is to document the main chemical, physical and microbiological characteristics found during these autopsies and to discuss the available options for biological control. 2 LABORATORY CHARACTERISATION OF MEMBRANE BIOFILMS • >5% Fe as iron oxide when treating brackish water • high microbiological counts (>106 cfu/cm2) including bacteria, fungi and sometimes yeasts BIOFILM COMPOSITION Table 1 summarises analytical results from a number of biofouled systems producing potable quality water for municipal or process use. MICROBIOLOGICAL ENUMERATION AND IDENTIFICATION In summary, these laboratory studies and others performed within the last year have shown that a typical biofilm has the following characteristics: The dilution pour plate method is used to enumerate the numbers of viable bacteria, fungi and yeasts present as sessile organisms on the membrane surface, plastic spacer material and permeate carrier. The viable counts are expressed as cfu/cm2 for membrane, plastic spacer and permeate carrier samples. Table 2 gives the ranges of viable bacteria enumerated from autopsied membranes. • >90% moisture content • of dried deposit, >50% total organic matter • up to 40% humic substances as % of total organic matter in high coloured waters • low inorganic content Table 1 — Summary results of recent autopsies Plant Location Size (m3/hr) Water Source Major Foulants Netherlands 18 brackish 44% organics 30% Fe 10% SiO2 89% Canary Islands 63 sea water 63% organics 4.7% MgCO3 1.7% CaSO4 92% Spain 12 brackish 66% organics 14% alumina 3.4% SiO2 94% Italy 36 brackish 26% organics 36% Fe 13% SiO2 85% Argentina 160 brackish 44% organics 37% SiO2 5% Fe 93% Spain 20 brackish 90% organics 4% phosphate 1.9% Fe 0.5% SiO2 94% Germany 22 brackish 76% organics 7.1% SiO2 5.1% CaPO4 85% Spain 1000 brackish 67% organics 13% SiO2 4.5% alumina 90% Egypt 200 brackish 50% organics 39% Fe 2.9% CaSO4 92% USA 125 brackish 63% organics 10% alumina 85% 3 Foulant Moisture Content Table 2 — Typical microbiological activity in biofouled spiral wound elements Microbiological identification found only two species of Pseudomonas present on the membrane surface and spacer. This is unusual in biofouled membrane systems where more diverse microbial assemblages are normally encountered. Range of viable Range of bacteria counts fungal counts cfu/cm2 cfu/cm2 Fouled membrane 1 x 102 - 1 x 108 2 Plastic spacer material* 4 x 10 - 5 x 10 Permeate carrier 6 < 102 - 1 x 106 Bacteria: Pseudomonas vesicularis, Pseudomonas fluorescens 0 - 1 x 103 3 0 - 1 x 10 None Microbiological Counts *viable bacteria enumerated per cm2 of the spacer mesh Identification of microorganisms in RO biofilm have been carried out. Several species of bacteria were present in the majority of biofilms investigated. Table 3 details genera commonly encountered. Corynebacterium, Pseudomonas, Bacillus, Artrobacter, Flavobacterium, Aeromonas Fungi Penicillium, Trichoderma, Mucor, Fusarium, Aspergillus Yeasts Occasionally identified in significant numbers 2.9 x 106 cfu/cm2 bacteria Plastic spacer 1.81 x 105 cfu/cm2 bacteria Permeate carrier <102 cfu/cm2 bacteria This particular biofilm, when viewed with SEM (Figures 4 and 5), after careful preparation, was found to contain relatively few bacteria, in contrast to large amounts of apparently EPS material. It may be significant to note that the feed to this membrane was undergoing chlorination and dechlorination in the pre-treatment train, which may have been responsible for the large quantities of EPS present in this system. Table 3 — Common microorganisms identified in membrane biofilm Bacteria Membrane surface Although good kill rates were obtainable with oxidising (100%) and non-oxidising biocides (99.9%), laboratory cleaning tests indicated that this biofilm was particularly difficult to remove using standard cleaning procedures. The only success in removing the foulant was found with a surfactant at pH 13. This high pH exceeded that normally advised by the membrane manufacturer but approved for occasional intensive cleaning. The most commonly occurring are the bacterium Pseudomonas and Corynebacterium and the fungal genera Penicillium and Trichoderma. The autopsy procedure investigates the efficacy of biocides on viable organisms isolated from the autopsied membrane. This allows an indication of the potential effectiveness of a biocide agent in specific systems. In practice, this cleaning programme, used in conjunction with regular maintenance cleaning and careful monitoring, limited the increase in ∆P caused by the biofilm accumulation and growth. This has allowed the site to minimise the detrimental effects of increases in ∆P to within tolerable levels. CASE STUDIES MUNICIPAL DRINKING WATER PLANT, SOUTHERN EUROPE, BRACKISH FEEDWATER This plant produces 1,000 m3/h operating between 70 and 80% recovery. There is an extensive pre-treatment system comprising sand filtration, chlorination/ dechlorination, acid and phosphonate-based antiscalant dosing. High ∆P was evident for this plant, which was proving difficult to control. An autopsy of an 8" spiral wound membrane produced the following data: POTABLE WATER PLANT, NORTHERN EUROPE The SEM shown in Figures 6 and 7 reveal the fouling layer taken from a potable plant treating a brackish surface water where biofouling had caused an overall 30% decrease in membrane performance. Foulant composition Moisture content: 86% water Organic content: 73%, of which 34% was humic substances Inorganic content: iron as iron oxide (17 %) silica (3%) calcium as calcium sulphate (2%) alumina (0.6%) Foulant Composition Moisture content: 96% water Organic content: 76%, of which 44% was humic substances Inorganic content: silica (7%) calcium as calcium sulphate (3%) iron as iron oxide (3%) 4 Figures 4 and 5 — SEM views of the biofouled membrane (municipal drinking water plant) Figures 6 and 7 — SEM views of the biofouled membrane (potable water plant) lowed by an off-line non-oxidising biocide to control biogrowth. Significant aspects of the biofilm included high organics, iron and humic acid content. Microbiological analysis of the fouled membrane showed 7.3 x 105 cfu/cm2 of bacteria present on the separating surface and 4.3 x 105 cfu/cm2 on the spacer material. In addition, high numbers of fungi and yeasts were enumerated on both the membrane (1.51 x 103 cfu/cm2) and plastic spacer (7.44 x 103 cfu/cm2). The numbers of fungi and yeast observed were significant. Factors such as CIP tank size, the ability to clean stages separately and the provision of temperature control affect cleaning performance and the rate of microbiological regrowth. Practical experience of biofouled systems has shown that a three-stage cleaning procedure such as detailed below gives the best cleaning efficiency. The following types of microorganism were identified on the membrane surface: Cleaning recommendation Bacteria: Rod-shaped bacteria, Corynebacterium and Arthrobacter Fungi: Penicillium Yeasts: Candida Stage 1: Alkaline surfactant and chelating agent to condition and break down the organic fouling. Cleaning conditions: pH 10.5, 30°C, 4 hour recirculation and soaking. It was recognised that the major influence on fouling was due to the presence of iron bound with organics and EPS. For this reason, the most suitable cleaning product recommended was an alkaline surfactant fol- Stage 2: Broad-spectrum non-oxidising biocide to eliminate microbiological growth. Cleaning conditions: 30°C, 30 minute recirculation. 5 Stage 3: Alkaline and chelating surfactant, to remove microorganisms and organic debris. reported after chlorine use was discontinued in a Mediterranean sea water plant.8 In this case, cleaning frequency was reduced from biweekly to yearly. Cleaning conditions: pH 10.5, 30°C, 4 hour recirculation and soaking. The reasons for this improvement are not completely clear and it is probable there are several contributory factors: An optional acidic clean may be performed to follow step 3 to remove remaining traces of inorganic scale. The sequence of applying cleaning chemicals is important. For example, certain humic acids can become difficult to remove if subjected to acid conditions. In such cases where the exact nature of the foulant is not known, it is always advisable to commence cleaning with an alkaline product. • It is inevitable that some bacteria will survive disinfection protected by EPS or as part of a planktonic community. The latter could be sloughed biofilm that has become detached from pre-treatment equipment. Bacteria arriving in the membranes in this compromised condition may produce EPS as a defence mechanism which, in turn, will make the bacteria resistant to the biocide and more difficult to eliminate. The emergence of a oxidant resistant biofilm can cause problems since this may be composed of a high proportion of EPS, which can result in significant energy losses. Membrane compatibility is essential for all cleaners and biocides used in the cleaning process and desirable characteristics are: • • • • • • non-oxidising stability at pH 2-12 nonionic or anionic nonfilming compatible with other cleaning products stable at 20 – 35°C • It is probable that nonviable microorganisms in the water supply following chlorination can act as a nutrient source. • Oxidising biocides may break down humic acids into smaller components, which may become available as a nutrient to the microbiological flora. CHEMICAL BIOCIDES OXIDISING BIOCIDES Many oxidising biocides are available for use in industrial water treatment processes: • • • • • • • SODIUM BISULPHITE SHOCK DOSING The efficacy of NaHSO3 as a biocide for sea water is dependent on the use concentration, the exposure time and the type of microorganisms present.9 With an exposure time of 30 minutes at a concentration of 500 ppm, kill rates up to 99% have been reported for sea water microflora. Whilst these values appear impressive, other data presented for aerobic marine bacteria indicated a greater resistance to sodium bisulphite with only a 75% kill obtained after 4 hours of contact at 500 ppm. It is probable that certain microorganisms such as sulphate-reducing bacteria would adapt to these anaerobic conditions. chlorine bromine chloramine chlorine dioxide hydrogen peroxide peroxyacetic acid ozone The use of oxidising biocides requires caution since they are incompatible for long-term contact with polyamide-based membranes. Where compounds like chlorine are used in the pre-treatment system, it is essential to remove any residual chlorine with sodium bisulphite well in advance of the membrane array. When chlorine or ozone are dosed to sea water, hypobromous acid is formed, which will eliminate bacteria but also cause membrane damage. NON-OXIDISING BIOCIDES Oxidising biocides such as peroxyacetic acid/ hydrogen peroxide and chloramine are used in composite membranes, provided the pH, temperature and contact time are strictly controlled. However, the advantages of non-oxidising biocides are evident. CHLORINATION—FRIEND OR FOE? It has long been standard practice to control biological growth in the feedwater by using chlorine. Current theory and practical experience indicates that this is not always successful in controlling biofouling. In fact, it has been found on occasion to worsen the biofouling potential. Significant improvements in DuPont hollow fibre RO plant performance have been Many non-oxidising biocides such as formaldehyde, glutaraldehyde and quaternary ammonium compounds etc. are used by the water treatment industry, although non-oxidising biocides are not approved for potable applications. 6 Several proprietary biocides have been developed for use in membrane systems. These are membrane compatible, suitable for discharge and biologically effective. Many of these compounds are used intermittently at low dose rates, and are a cost-effective means of maintaining a clean membrane surface. Low molecular weight compounds can pass through the membrane to treat the product water side. Experience with many of these biocides in process applications has shown that their long-term use often causes microbial resistance. Continuous low dose rates should be avoided and it is advised they be ‘shock dosed’ to the system water. For instance, one such proprietary non-oxidising biocide is recommended at 200-300 ppm for 1/2 – 1 hour. This treatment may be necessary daily, weekly or monthly depending upon rates of regrowth and cleaning efficiency. In process applications, it is common practice to employ a secondary non-oxidising biocide to prevent the emergence of resistant strains. This is a technique we suggest be used in membrane applications. CONCLUSIONS Many autopsies have shown that some biofilms are difficult to eliminate by routine cleaning procedures. The biofilm becomes ‘conditioned’ and ‘resistant’ to chemical cleaning procedures due to the establishment of resistant microorganisms or the massive production of protective slimes. In the latter case, high energy losses will usually be evident. Resistant fouling necessitates cleaning outside recommended temperature and pH limits, which is inadvisable. Very often the only option is to try and maintain conditions at a tolerable level before an irrecoverable position is reached and the membrane must be replaced. Consideration should then be given to improved control procedures and more effective membrane biocides and cleaners. The following points are important: • Membrane biofouling is endemic. • Although most biofilms have common characteristics, their structure varies greatly. • The impact of biofilm upon plant performance depends on the composition and structure of the biofilm. Many plants with biofilm work satisfactorily. • The resistance of the biofilm to chemical cleaning and biocides will increase if inefficient control measures are employed. • A biofilm formed under natural conditions may prove easier to control in the long term than one formed where biocides are in use. • Non-oxidising biocides should always be applied as a ‘shock’ dose. Consideration should be given to the use of two non-oxidising biocides in a complementary biogrowth control programme. REFERENCES 1. Flemming H-C. (1993). Mechanistic aspects of RO membrane biofouling and prevention. In Reverse Osmosis: Membrane Technology, Water Chemistry, and Industrial Applications. Ed. Amjad Z. Publ. Van Nostrand Reinhold, New York. 2. Lessel T., Motsch H. and Hennig E. (1975). Experience with a pilot plant for the irradiation of sewage sludge. In Radiation for a Clean Environment. Publ. International Atomic Energy Agency, Vienna. 3. Characklis W. G. and Marshall K. C. Eds (1990). Biofilms. Publ. John Wiley, New York. 4. Nichols P. (1989). Susceptibility of biofilms to toxic compounds. In Structure and Function of Biofilms. Eds. Characklis W. G. and Wilderer P. A. Publ. John Wiley, New York. 5. LeChevalier M.W., Cawthon C. C. and Lee R. G. (1988). Inactivation of biofilm bacteria. Applied and Environmental Microbiology 54 2492-2499. 6. Cullimore D. R. (1992). Practical Manual of Groundwater Microbiology. Publ. Lewis Michigan. 7. Characklis W. G. (1981). Fouling biofilm development: A process analysis. Biotechnology and Bioengineering. 23 1923-1960. 8. Hamida A. B. and Moch I. (1996). Controlling biological fouling in open sea intake RO plants without continuous chlorination. Desalination and Water Reuse. 6 (3) 40-45. 9. Applegate L. E. and Erkenbrecher C. W. (1987). Monitoring and control of biological activity in Permasep seawater RO plants. Desalination. 65 331-359. Ondeo Nalco Company Ondeo Nalco Center • Naperville, Illinois 60563-1198 SUBSIDIARIES AND AFFILIATES IN PRINCIPAL LOCATIONS AROUND THE WORLD PermaCare is a Registered Trademark of Ondeo Nalco Limited Printed in U.S.A. 4-02
© Copyright 2026 Paperzz