mainte-To ensure delivery of a high-quality municipal potable water supply to each sumer, managers of public water supply systems must be continually vigilant for anyintrusions of contam
Trang 1CHAPTER 18 MICROBIOLOGICAL QUALITY CONTROL IN DISTRIBUTION
SYSTEMS
Edwin E Geldreich, M.S.
Consulting Microbiologist Cincinnati, Ohio
Mark LeChevallier, Ph.D.
Research Director American Water Works Service Co.
Voorhees, New Jersey
The purpose of a water supply distribution system is to deliver to each consumersafe drinking water that is also adequate in quantity and acceptable in terms of taste,odor, and appearance Historically, the initial network of pipes was a response topresent community needs that eventually created a legacy of problems of inade-quate supply and low pressure as the population density increased (Frontinus, 1973;Baker, 1981) To resolve the problems caused by increasing water demand along thedistribution route, reservoir storage was created Pressure pumping to move water tofar reaches of the supply lines and standpipes was incorporated to afford relief fromsurges of pressure In some areas, population growth exceeded the capacity of awater resource, so other sources of water were incorporated and additional treat-ment plants were built to feed into the distribution network Another response was
to consolidate neighboring water systems and interconnect the associated tion pipe networks
distribu-GENERAL CONSIDERATIONS
FOR CONTAMINATION PREVENTION
Today (1999), community expansion plans are more fully developed and include theengineering of utility service so that careful consideration is given to meeting futureprojected water supply needs Advanced planning provides the opportunity to
18.1
Trang 2design the pipe network as a grid with a series of loops to avoid dead ends Theobjective is to produce a circulating system capable of supplying high quality water
to all areas while at the same time permitting any section to be isolated for nance, repair, or decontamination without interrupting service to all other areas
mainte-To ensure delivery of a high-quality municipal potable water supply to each sumer, managers of public water supply systems must be continually vigilant for anyintrusions of contamination or occurrences of microbial degradation in the distribu-tion network This job is complicated by the very nature of the distribution system—
con-a complex network of mcon-ains, fire hydrcon-ants, vcon-alves, con-auxilicon-ary pumping, chlorincon-ationsubstations, storage reservoirs, standpipes, and service lines Following the intrusion
of microbial contamination, any of these component parts may serve as a habitatsuitable for colonization by certain microorganisms in the surviving flora The per-sistence and possible growth of organisms in the pipe network are influenced by avariety of environmental conditions that include physical and chemical characteris-tics of the water, system age, type of pipe materials, and the availability of sites suit-able for colonization (often located in slow-flow sections, dead ends, and areas ofpipe corrosion activity)
ENGINEERING CONSIDERATIONS
FOR CONTAMINATION PREVENTION
Many public water utilities make substantial efforts to expand their distribution works to keep up with continuing suburban growth Urban renewal and highwayconstruction projects may at times require the relocation or enlargement of portions
net-of the distribution network Corrosion, unstable soil, faulting, land subsidence,extreme low temperatures, and other physical stresses often cause line breaks andnecessitate repair or replacement of pipe sections To avoid possible bacteriologicalcontamination of the water supply during these construction projects, a rigorousprotective protocol must be followed
Distribution System Construction Practices
The American Water Works Association has developed standard procedures thatare used, with variations, by most of the water supply industry for disinfecting watermains (AWWA, 1986) In essence, these recommendations recognize six areas ofconcern: (1) protection of new pipe sections at the construction site; (2) restriction
on the use of joint-packing materials; (3) preliminary flushing of pipe sections;(4) pipe disinfection; (5) final flushing; and (6) bacteriological testing for pipe disin-fections
Pipe sections, fittings, and valves stockpiled in yard areas or at the constructionsite should be protected from soil, seepages from water or sewer line leaks, stormwa-ter runoff, and habitation by pets and wildlife (Becker, 1969; Russelman, 1969) Each
of these contamination sources may deposit significant fecal material in the interior
of pipe sections awaiting installation Septic tank drain fields, subsurface water inareas of poor drainage or high water table, and seasonal or flash flooding may alsointroduce significant contamination into unprotected pipe sections Fecal materialintroduced by contaminating sources may become lodged in pipe fittings and valves.Thus, such sites become protected habitats from which coliforms and any associatedpathogens in the contaminated material may be released into the bulk flow of water
Trang 3supply Common-sense protective measures include providing end covers for thesepipe materials, drainage of standing water from trenches before pipeline assembly,and flushing of new construction or line repairs to remove all visible signs of debrisand soil (Suckling, 1943; Davis, 1951).
Pipe-Joining Materials
Gasket seals of pipe joints can be a source of bacterial contamination in new pipes(Hutchinson, 1971) Annular spaces in joints provide a protected habitat for contin-ued survival and possible multiplication of a variety of bacteria in the distributionnetwork In these instances, although the heterotrophic plate count (HPC) and anycoliform occurrences may be temporarily reduced by main disinfection, bacteriasoon become reestablished from the residual population harbored in some joint-packing materials In this regrowth process, the variety of organisms and dominance
of strains change, often restructuring the bacterial flora to a predominant population
of Pseudomonas aeruginosa, Chromobacter strains, Enterobacter aerogenes, or
Kleb-siella pneumoniae Thus, where the pattern of organisms present is predominantly
one bacterial strain, a search for a protective habitat in joint-packing materials orimpacted material in pipe sections should be made (Calvert, 1939; Adam and Kings-bury, 1937; Taylor, 1950, 1967–1968; Schoenen, 1986; Schubert, 1967) Nonporousmaterials such as molded or tubular plastic, rubber, and treated paper products arepreferable Lubricants used in seals must be nonnutritive to avoid bacterial growth
in protected joint spaces Efforts to develop bacteriostatic lubricants have resulted
in the inclusion of various quaternary ammonium compounds that minimize tamination from pipe joint spaces (Hutchinson, 1974) The National SanitationFoundation (NSF International, Ann Arbor, Michigan) has proposed a test methodfor evaluating the biological growth potential of materials that contact drinkingwater (Bellen et al., 1993) Although the method has not yet received final approval,some manufactures have submitted their products for testing, and in many cases theresults are available from NSF upon request
con-Water Supply Storage Reservoirs
Water use in a community varies continuously as a reflection of the activities of thegeneral public and local industries While industrial uses of potable water may bemore predictable, expecting water treatment operations to gear production to thosefrequent and sudden changes in water demand from all consumers is impractical.For this reason storage reservoirs are an essential element of the distribution net-work These water supply reserves supplement water flows in distribution duringperiods of fluctuating demand on the system, providing storage of water during off-peak periods; equalize operational water pressures; and augment water supply fromproduction wells that must be pumped at a uniform rate Storage reservoirs also pro-vide a protective reserve of drinking water to guard against discontinuance of watertreatment during chemical spills in the source water, flooding of well fields, trans-mission line breaks, and power failures An important secondary consideration isproviding adequate storage capacity for fire emergencies
Finished water reservoirs may be located near the beginning of a distribution tem, but most often they are situated near the extremities of the system Localtopography plays an important part in determining the use of low-level or high-levelreservoirs Underground storage basins are usually formed by excavation, while
Trang 4sys-ground-level reservoirs are constructed by earth embankment.The sides and bottom
of such reservoirs are lined with concrete, Gunite, asphalt, or with a plastic sheet toprevent or reduce water loss in storage (Harem, Bielman, and Worth, 1976) Inearthquake zones, reinforced concrete or a series of flat-bed steel compartments ismandatory Reinforced concrete is often selected because of its minimal rate of dete-rioration from water contact Elevated storage tanks and standpipes are constructed
of steel, with an interior coating applied to retard corrosion (Wade, 1974)
Care must be taken to prevent potential contamination of the high-quality waterentering storage reservoirs and standpipes One area of concern is in the application
of coating compounds over the inner walls of tanks to maintain tank integrity.Organic polymer solvents in bituminous coating materials may not entirely evapo-rate even after several weeks of ventilation As a consequence, the water supply instorage may become contaminated from the solvent-charged air and from contact atthe sidewall Some of these compounds are assimilable (biodegradable) organicsthat support growth of heterotrophic bacteria during warm water periods (Schoe-nen, 1986; Thofern, Schoenen, and Tuschewitzki, 1987; Mackle, 1988; Bernhardt andLiesen, 1988) Liner materials, also used to prevent water loss, may contain bitumen,chlorinated rubber, epoxy resin, or a tar-epoxy resin combination that will eventu-ally be colonized by microbial growth and slime development (Schoenen, 1986).PVC film and PVC coating materials are other sources of microbial activity Non-hardening sealants (containing polyamide and silicone) used in expansion jointsshould not be overlooked as a possible source of microbial habitation
Water volumes in large reservoirs mix and interchange slowly with water that isactually distributed to service lines In some instances, water storage may becomestratified and experience a complete mixing only after a sudden change in ambientair temperature or as a consequence of a significant water loss in the system caused
by a major main break or intensive system flushing (Geldreich et al., 1992; Clark etal., 1996) Standpipes, in contrast, provide a fluctuating storage of water during adownsurge, thereby providing surge relief in the system Abrupt changes in waterflow that sometimes occur during surge relief can disturb sediment deposits, movingviable bacteria from biofilm sites into the main flow of water
Reservoirs of treated water should be covered whenever possible to avoid tamination of the supply from bird excrements (Alter, 1954; Fennel, James, and Mor-ris, 1974), air contaminants, and surface water runoff The health concern with bird
con-excrement is that this wildlife may be infected with Salmonella and protozoans
pathogenic to man Within the wildlife population in every area (as is true for anycommunity of people), a persistent pool of infected individuals exists that shedspathogenic organisms through fecal excretions Seagulls are scavengers and oftenare found at landfill locations and waste discharge sites, searching for food which isoften contaminated by a variety of pathogens At night, birds often return inland toroost in aquatic areas, such as source water impoundments and open finished waterreservoirs, thereby introducing pathogens through their fecal excrements (Fennel,James, and Morris, 1974) Pigeons roosting in elevated storage tanks were believed
to be the source of Salmonella typhimurium that contaminated the water supply of
a small community in Missouri (Geldreich et al., 1992; Clark et al., 1996)
Air pollution contaminants and surface water runoff can contribute dirt, ing leaves, lawn fertilizers, and accidental spills to a water supply that is not covered.Such materials increase productivity of the water by providing support to food chainorganisms and nitrogen-phosphate requirements for algal blooms This degrades thetreated water quality
decay-Covered distribution system storage structures also are subject to occasional tamination because of air movement in or out of the vents as a result of water move-
Trang 5con-ment in the structure During air transfer, the covered reservoir is exposed to fallout
of dust and air pollution contaminants from the inflowing air Vent ports or conduitsfrom the service reservoir to the open air should be equipped with suitable air filters
to safeguard the water quality from airborne contaminants Such air filters must bereplaced periodically to prevent a serious loss of air transfer or create an undesirablevacuum Birds and rodents may also gain access through air vents that have defectivescreen protection Bird or rodent excrement around the vents may enter the watersupply and become transported into the distribution system before dilution andresidual disinfection are able to dissipate and inactivate the associated organisms
FACTORS CONTRIBUTING TO MICROBIAL
QUALITY DETERIORATION
Factors contributing to deterioration of microbial quality may be associated withsource water quality, treatment processes, or distribution network operation andmaintenance The following sections review each of these areas
Source Water Quality
Bacteria in distributed water may originate from the source water High-qualitygroundwater can be characterized as containing <1 coliform per 100 mL and a het-erotrophic bacterial population that is often very sparse (less than 10 organisms permL), even in waters that reach the growth-stimulating temperature of 15°C or more(Alson, 1982) These microbial qualities show little fluctuation, because the ground-water aquifer is protected from surface contamination Some groundwaters, however,are not insulated from surface contamination (Allen and Geldreich, 1975) Agricul-tural fertilizer runoff can contribute nitrates and improperly isolated landfills mayintroduce a variety of organics, many of which are biodegradable In such situations,bacterial populations in the groundwater become excessive, resulting in 1000 to10,000 heterotrophic bacteria per mL Groundwaters containing a high concentration
of iron or sulfur compounds provide nutrients for a variety of nuisance bacteria thatmay become numerous and restrict water flow from a well Where groundwaters arepoorly protected from contamination by stormwater runoff and wastewater effluents,coliforms and pathogens (bacterial, viral, protozoan) may be introduced into the dis-tributed water unless a treatment barrier is provided (Craun, 1985)
Surface water sources are subject to a variety of microbial contaminants duced by stormwater runoff over the watershed and the upstream discharges ofdomestic and industrial wastes While impoundments and lakes provide water vol-ume and buffering capacity to dilute bacterial contamination, counterproductivefactors must be considered Stratification/destratification of lake waters, decayingalgal blooms, and bacterial nutrient buildup contribute to deteriorating water qual-ity that may interfere with treatment effectiveness
intro-Treatment Processes
Water supplies using a single barrier (disinfection) for surface water treatment willnot prevent a variety of organisms (algae, protozoa, and multicellular worms andinsect larvae) from entering the distribution system (Allen, Taylor, and Geldreich,
Trang 61980) While many of these organisms are not immediately killed by disinfectantconcentrations and contact times (C⋅T values) that control coliforms and viruses(Hoff, 1986), they eventually die because of lack of sunlight (algae) or adverse habi-tat (multicellular worms and insect larvae) Disinfection is also less effective on
a variety of environmental organisms that include spore-forming organisms
(Clos-tridia), acid-fast bacteria, gram-positive organisms, pigmented bacteria, fungi, yeast,
and protozoan cysts All of these more resistant organisms can be found in the pipeenvironment (Haas, Meyer, and Fuller, 1983; Bonde, 1977; Rosenzweig, Minnigh,and Pipes, 1983; Niemi, Knuth, and Lundstrum, 1982; Reasoner, Blannon, and Geld-reich, 1989)
Filtration is an important treatment barrier for protozoan cysts (Entamoeba,
Giar-dia, Cryptosporidium), being more effective than the usual disinfectant
concentra-tion and contact times applied in processing raw water (Logsdon, 1987) Improperlyoperated filtration systems have been responsible for releasing concentrated num-
bers of entrapped cysts (Giardia and Cryptosporidium) as a result of improper filter
backwashing procedures or filter bypasses and channelization within the filter bed(Amirtharajah and Wetgstein, 1980) Filter sand may become infested with nema-todes from stream or lake bottom sediments that shed into the process water andpass into the distribution system While nematodes are not pathogenic, they canshelter viable bacteria ingested from source water or filter media beds, and if thesefood-chain organisms are not digested quickly, provide a passage for escaping sur-vivors to reach the distribution system (Tracy, Carnasena, and Wing, 1966; Levy etal., 1984)
Properly operated water treatment processes are effective in providing a barrier
to coliforms and pathogenic microorganisms reaching the distribution system Thisdoes not, however, preclude the passage of all nonpathogenic organisms through thetreatment train Investigation of heterotrophic bacterial populations revealed that a4-log (99.99 percent) or better reduction can occur through conventional treatmentprocesses (raw water storage, coagulation, settling, rapid-sand filtration, and chlori-nation) for many of these organisms (Reasoner, Blannon, and Geldreich, 1989) Aless significant reduction occurs among the subpopulation of pigmented organisms
in the heterotrophic flora, however, so that these organisms may become the dominant bacteria after processing and then the dominant strains in distributedwater
pre-Application of powdered carbon, granular activated carbon (GAC) filtration, orbiological activated carbon (BAC) treatment introduces other opportunities forbacteria to enter the microbial community of processed water Highly adsorptivegranular or powdered carbon is often used in specific situations for removal oforganics, including taste- and odor-causing substances, and is changed periodicallywhen the carbon bed reaches near-saturation for expected removal rates In con-trast, the biological activated carbon process depends on the establishment of a per-manent biofilm for degradation of conversion products created by ozonation ofrecalcitrant organics in process water, as well as of biodegradable organic matteralready in the water
Several coliform species (Klebsiella, Enterobacter, and Citrobacter) have been
found to colonize GAC filters, grow during warm water periods, and discharge intothe process effluent (McFeters, Kippen, and LeChevallier, 1986; Camper et al., 1986;Camper et al., 1987; Stewart, Wolf, and Means, 1990) Activated carbon particleshave also been detected in finished water from several water plants using powderedactivated carbon or GAC treatment Over 17 percent of finished water samplesexamined from nine water treatment facilities contained activated carbon particlefines colonized with coliform bacteria (Ridgway and Olson, 1981) These observa-
Trang 7tions confirm that activated carbon fines provide a transport mechanism by whichmicroorganisms penetrate treatment barriers and reach the distribution system.Other mechanisms that could be involved in protected transport of bacteria into thedistribution system include release of aggregates or clumps of organisms from colo-nization sites in GAC/BAC filtration and by passage with unsettled coagulants.
Furthermore, heterotrophic bacterial densities in distributed water from a scale treatment train using GAC (Symons et al., 1981) were found to be significantlyhigher (Table 18.1) than in water from a similar full-scale treatment train that didnot employ GAC (Haas, Meyer, and Paller, 1983) Upon entering the pipe network,persistence and growth of these organisms will be influenced by the same factorsthat also affect disinfectant effectiveness: habitat locations, water temperature, pH,and assimilable organic carbon concentrations (Haas, Meyer, and Paller, 1983;Rittmann and Snoeyink, 1984)
full-Distribution Network Operation and Maintenance
Because the public health concern for the microbial quality of drinking water hasuntil recently been based solely on limiting total coliform occurrence (EPA, 1976),the acceptance of new or repaired mains has depended only on a laboratory reportthat no coliforms are detected in water held in the new pipe sections A more rigor-ous check on installed pipe cleanliness would include examination of water in thepipe section for elevated heterotrophic bacterial densities in addition to total coli-forms (Geldreich et al., 1972; AWWA, 1987) The HPC in this situation reflects themyriad of soil organisms that could have been introduced into the pipe section dur-ing construction or repair Soil deposits in new pipe sections may not only introduce
a variety of heterotrophic bacteria to the distribution network but also provide somemeasure of protection to associated bacteria from disinfection exposure Some ofthe poor disinfection results attributed to chlorine applied in these situations may betraced to excessively dirty line sections or joints
In Halifax, Nova Scotia, a new supply line was found to contain pieces of woodused during construction work embedded in some pipe sections (Martin et al., 1982)
TABLE 18.1 Treated Water Bacterial Populations Following Various Water Treatment Processes UsingStandard Plate Medium or R-2A Medium with Extended Incubation Times (organisms/mL)*
Lime-Softened water Sand-Filter effluent GAC adsorber effluentSampling SPC, 2 SPC, 6 R-2A, 6 SPC, 2 SPC, 6 R-2A, 6 SPC, 2 SPC, 6 R-2A, 6data days days days days days days days days daysInitial 120 350 510 890 1,200 1,500 < 1 140 220
* All cultures incubated at 35 ° C SPC, standard plate count (SPC agar); N.D., not done.
Trang 8An environmental Klebsiella associated with the wood forms adjusted to the
distri-bution pipe environment and colonized the exposed wood surfaces The flow ofwater transported this coliform into the bulk flow, resulting in consistently unsatis-factory test results Since the wood debris was difficult to remove from the pipeline,the problem was resolved by the addition of more than 5 mg/L lime to the processwater, which elevated the water pH to 9.1 in the distribution system At this pH,
Klebsiella were either inactivated or entrapped in the pipe sediment, and the
prob-lem was eliminated
Upon completion of a new pipeline or after emergency repairs are made to aline break, flushing water through the pipe section at a minimum velocity of 10 ft/s(76.2 cm/s) to remove soil particles is advisable (Buelow et al., 1976) In lines withdiameters of 16 in (4.1 cm) or more, this velocity may not be attainable or may beineffective, requiring polypig or foam swab applications to be considered Followingflushing, disinfectant should be introduced into the new sections and the water heldfor 24 to 48 h to optimize line sanitation Bacteriological tests for stressed coliforms(McFeters, Kippin, and LeChevallier, 1986) and the HPC should then be performed
If the results of these tests are satisfactory (<1 coliform/100 mL;<500 HPC/mL), theline may be placed in service If not, the line should again be flushed and refilled withdistributed water dosed with 50 mg/L free available chlorine Chlorine levels shouldnot decrease below 25 mg/L during the 24-h holding period before the line is flushedand bacteriological testing is repeated In pipes free of extraneous debris, free avail-able chlorine (1 to 2 mg/L), potassium permanganate (2.5 to 4.0 mg/L), or coppersulfate (5.0 mg/L) have been used to meet coliform requirements (Martin et al.,1982; Harold, 1934; Hamilton, 1974) Only free available chlorine, however, wasfound to eliminate large numbers of heterotrophic bacteria (Buelow et al., 1976)
MICROBIAL QUALITY OF DISTRIBUTED WATER
Because of the difficulties in isolating and identifying a broad spectrum of organismswith widely differing growth requirements, little in-depth information is available inthe literature on the identity of all heterotrophic organisms found in water supplies.However, recent research into fatty acid identification of these organisms is begin-ning to yield a better characterization of the many ill-defined bacteria in distributionwater In the analysis of cellular fatty acids the cells are lysed, and the constituentacids methylated and then identified by gas chromatography (Briganti and Wacker,1995) Bacteria are identified by comparing the profile of fatty acids from the drink-ing water organism isolates to a library of profiles from known organisms Previousstudies have generally been limited to identification of those organisms that wereassociated with consumer complaints about taste, odor, and color (Hutchinson andRidgway, 1977) Other studies have explored spoilage problems in the food, beverage,cosmetic, and drug industries, which use large quantities of potable water in produc-tion processes (Alson, 1967; Tenenbaum, 1967; Dunnigan, 1969; Borgstrom, 1978).The following section is a brief profile of organisms found in distribution systems
Bacterial Profiles
Heterotrophic organisms in water supplies most often originate in the source water,survive the rigors of treatment processes, and adapt to the environment of the waterdistribution network (Geldreich, Nash, and Spino, 1977) Trace organic nutrients
Trang 9already present in the bulk water or accumulated in reservoir and pipe sedimentscan support a diverse population (Table 18.2) of surviving organisms (heterotrophs).Habitat sites that are successfully colonized almost always invoke the mixed growth
of organisms that are attached to each other or to particles, sediments, and porousstructures in the pipe tubercles or sediments The spectrum of organisms may
include many negative bacteria (such as coliforms and Pseudomonas),
gram-positive organisms, sporeformers, acid-fast bacilli, pigmented organisms, mycetes, fungi, and yeast In addition, various protozoans and nematodes that feed
actino-on these microbial populatiactino-ons can be found in protected areas of low-flow sectiactino-ons
or dead ends
Coliforms. Total coliform bacteria counts are used primarily as a measure of watersupply treatment effectiveness and as a measure of public health risk These gram-negative bacteria are occasionally found in water supplies Data in Table 18.3 showsthe wide range of coliform species that are encountered (Geldreich, Nash, andSpino, 1977; LeChevallier, Seidler, and Evans, 1980; Olson and Hanami, 1980; Her-son and Victoreen, 1980; Reilly and Kippin, 1981; Clark, Burger, and Sabatinos, 1983;Staley, 1983) While coliform bacteria are chlorine sensitive, they may be protectedfrom inactivation by associated particles originating in source water turbidity, acti-vated carbon fines released from GAC filtration (Camper et al., 1985), and inorganicsediments in the contact basin, as well as by inadequate conditions for disinfectionaction (contact time, water pH, and temperature) Coliforms may also enter the dis-tribution system as injured organisms that pass through treatment barriers or byintroduction into water line breaks, cross-connections, uncovered finished waterstorage, or low water pressure (<20 psi) For these reasons, more attention should begiven to the search for injured coliforms in an effort to provide a more sensitivemeasurement of water quality (LeChevallier and McFeters, 1985; McFeters, 1990;Bucklin, McFeters, and Amirtharaja, 1991)
Coliform colonization of the pipe network may occur in areas where porous iments accumulate (Allen, Taylor, and Geldreich, 1980; Tuovinen et al., 1980) Thesesediments develop from the action of corrosion or are the accumulation of particlesthat passed through the treatment works because of inadequate processing and thensettled in slow-flow and dead-end areas Such sites are attractive to bacterial colo-nization because these deposits adsorb trace nutrients from the passing water andprovide numerous surface areas for bacterial attachment against the flow of water
sed-Among the coliform bacteria, Klebsiella pneumoniae, Enterobacter aerogenes,
Entero-bacter cloacae, and CitroEntero-bacter freundii are the most successful colonizers
Encapsu-lation by these coliforms provides protection from the effects of chlorine or otherdisinfectants Once total coliforms become established in an appropriate habitat,growth can occur and result in occasional sloughing of cells into the flowing water.This condition can persist until either shearing effects of water hydraulics limitcolony growth or elevated disinfectant residuals penetrate the protective habitatand inactivate the microbial population
Antibiotic-Resistant Bacteria. Heterotrophic bacteria in a water supply that areresistant to one or more antibiotics may pose a health threat if these strains areopportunistic pathogens or serve as donors of the resistant factor to other bacteriathat could be pathogens Antibiotic-resistant (R factor) bacteria may originate insurface water sources used for public water supplies (Armstrong, Calomiris, and Seid-ler, 1982; Bedard et al., 1982) Polluted waters acquire bacteria with R factors fromthe fecal wastes of man and domestic animals in wastewater effluents, and storm-water runoff from farm pasturelands and feedlots Farm animals in particular may
Trang 10Water plant filter effluent and clearwell Distribution water
Enterobacter aerogenes Escherichia coli Aeromonas hydrophila Citrobacter freundii
Pseudomonas cepacia Pseudomonas putida Pseudomonas aeruginosa Bacillus sp.
Actinomycetes sp.
Trang 11Water plant filter effluent and clearwell Distribution water
Pseudomonas aeruginosa Klebsiella rhinoscheromatis Serratia liquefaciens Serratia marcescens
Bacillus sp.
Corynebacterium sp Micrococcus sp.
Nitrococcus sp.
Source: Data from Geldreich et al., 1977.
Trang 12receive continuous doses of antibiotics in animal feed and become constant generators
of a variety of antibiotic-resistant bacteria.Although treatment processes inactivate or
remove antibiotic-resistant organisms (Aeromonas, Hafnia, and Enterobacter) in the source water, a shift of this transmissible factor to other heterotrophic bacteria (Pseu-
domonas/Alcaligenes group, Acinetobacter, Moraxella, Staphylococcus, and cus) may occur (Armstrong et al., 1981).
Micrococ-Water supply treatment processes apparently act as a mixing chamber for R factortransfers with surviving organisms, which then acquire multiple resistances to differentantibiotics Many of the transformations occur in the biofilm established on activatedcarbon and sand filters (Armstrong et al., 1981) The disinfection process may alsohave a major impact on the selection of drug-resistant bacteria The reason for thecommon occurrence of streptomycin resistance among bacteria that survive chlorina-tion is not known Multiple antibiotic-resistant bacteria passing through water treat-ment are more tolerant to metal salts (i.e., CuCl2, Pb(NO3)2, and ZnCl2) (Armstrong etal., 1981) Examination of bacteria for multiple antibiotic-resistance from two sites in
a distribution system indicates a dynamic state of fluctuation (16.7 percent R factororganisms at one site and 52.4 percent at the other location) In a typical population
of 100 heterotrophic bacteria per mL of water from the distribution system, 40 to 70
of these organisms could be expected to have some antibiotic-resistance factors (El-Zanfaly, Kassein, and Badr-Eldin, 1987) What health risk this represents, particu-larly when the heterotrophic bacterial population is above the 500 organisms or moreper mL limit suggested for potable water, is not clearly understood
Mycobacteria. Origins of mycobacteria in water supply can be found in sourcewaters, open finished water reservoirs, soil contaminants introduced in groundwa-ters, and line repairs and new line construction Densities of environmental strains ingroundwater supplies ranged from 10 to 500 organisms per 100 mL, while densities
of these acid-fast bacteria in polluted surface waters may approach 104organismsper 100 mL In water supplies processed from surface waters, mycobacteria reduc-tion ranges from 60 organisms per 100 mL to only a few per liter Those aquatic
mycobacteria of health concern (Mycobacterium avium, M gordonae, M flavescens,
M fortuitum, M chelonae, and M phlei) may colonize susceptible humans
(immuno-compromised patients, surgery cases, individuals on kidney dialysis) by a variety ofroutes including water supply (Haas, Meyer, and Fuller, 1983; duMoulin andStottmeier, 1986; duMoulin et al., 1985)
In water treatment, the most significant reductions for these organisms occur ing sand filtration For example, during an 18-month study of two water systems,
dur-TABLE 18.3 Coliforms Identified in 111 PublicWater Supply Distribution Systems*
* Published data (Geldreich et al., 1977; LeChevallier et al., 1980; Olson and Hanami, 1980; Herson and Victoreen, 1980; Reilly and Kippin, 1981; Clark et al., 1983; Staley, 1983) from various distribution systems in six states (United States) and Ontario Province (Canada).
Trang 13reductions in the concentration of acid-fast bacteria by rapid sand filtration rangedfrom 59 to 74 percent Final disinfection, including the presence of free residual chlo-rine, did not have a statistically significant effect on the residual densities of mycobac-teria leaving the treatment plant (Haas, Meyer, and Fuller, 1983; Pelletier, duMoulin,and Stottmeier, 1988) Even the presence of a free chlorine residual at a low pH (5.9
to 7.1) did little to reduce the number of these organisms in the distribution systems.The ability of mycobacteria to survive in water distribution systems may be influ-enced by the protective waxy nature of the cell wall, which helps these organismsresist 1.5 mg free chlorine over 30 m contact time Moreover, some increase in num-bers were noted in the pipe environment at the ends of the distribution lines, wherechlorine residual disappeared, increased total organic carbon concentrationsoccurred (providing bacteria nutrients), and pipe sediments or tuberculation (sitesfor bacterial attachment) accumulated Further amplification of the acid-fast bacte-ria can be expected in some building plumbing systems and their associated attach-ments (Bullin, Tanner, and Collins, 1970; Haas, Meyer, and Paller, 1983)
Pigmented Bacteria. A characteristic of some bacteria that may be present in tributed water is the ability to form brightly colored pigments Little is known abouttheir health significance from ingestion; however, some strains have been the cause
dis-of gastroenteritis while others have been associated with pyrogenic reactions andsepticemia (Quarles et al., 1974) The seasonal occurrence of pigmented bacteria in
a treated water supply suggests that these bacteria originated from the source watersupply at some previous point in time, or from line breaks and repairs to the distri-bution lines (Reasoner, Blannon, and Geldreich, 1989)
More pigmented bacteria were found at a distribution sampling site 25 mi (40 km)from the water treatment plant than in river source water, the presedimentationbasin, after flocculation, in chlorinated influent to rapid sand filters, or in the filteredwater This observation suggests that pigmented bacteria surviving treatmentbecame adapted to the distribution environment and grew when conditions werefavorable
The proportion of yellow, pigmented bacteria in the source water was lowest inthe autumn and highest in the summer The mean proportion of pigmented bacteriawas calculated to be 26 percent of the heterotrophic population Orange-pigmentedbacteria comprise less than 10 percent of the plate count in any season (average 5.8percent) and pink or red organisms, less than 3 percent Pigmented bacteria such aspurple, black, and brown organisms ranged from 0.2 to 1.5 percent
In contrast, at the distribution sampling site, pigmented bacteria comprised from
65 to nearly 90 percent of the heterotrophic population, depending on the season.Lowest and highest percentages occurred in the winter (5 percent) and autumn (70percent), respectively Orange-pigmented bacteria, on the other hand, appeared as
82 percent of the total pigmented population in the winter and only 10 percent of thepopulation in the autumn Pink bacteria ranged from <1 percent (winter, spring, andfall) to 9 percent (summer) Other pigmented bacteria (purple, black, and brown)were rarely encountered in the distributed water A mixed population of pigmentedbacteria was generally present in the distributed water, with surges to dominanceamong these groups when environmental factors were favorable Wolfe, Ward, andOlson (1985) found that red-pigmented bacteria were resistant to 0.75 mg/L freechlorine for 30 to 60 m, but sensitive to 1.0 mg/L chloramines for 60 m Fatty acidanalysis has shown that many of the red-pigmented bacteria belong to the genus
Rodococcus and possess lipids that make them resistant to disinfection Because
these organisms are found in drinking water, pigmented bacteria have been reportedfrom water used in hospital therapy machines, and on other devices that use watersupply (Favero et al., 1974; Favero et al., 1975; Herman, 1976)
Trang 14Disinfectant-Resistant Bacteria. This bacterial group comprises a wide range oforganisms of very limited health significance whose presence is a reflection of disin-fection treatment effectiveness or is a depiction of the heterotrophic bacterial flora
of an untreated public water supply that contains many spore-forming organisms.Comparison of bacterial genera present in two different public water supplies insouthern California by Olson and Hanami (1980) indicated that variation in thediversity of the bacterial population can also be related to source water (i.e., surfaceversus groundwater) The predominant genera in one untreated groundwater supply
were Acinetobacter and Pseudomonas Klebsiella was also found intermittently in
very low numbers in this supply Chlorinated distributed water from a surface water
source contained Acinetobacter, Pseudomonas/Alcaligenes, and Flavobacterium as
predominant genera
Chlorination of water creates a strong selective pressure on bacterial populations.Ridgway and Olson (1982) found that bacteria isolated from a chlorinated surfacewater distribution system were more resistant to both combined and free forms ofchlorine than members of the same genera isolated from an unchlorinated ground-water system These results may have also been influenced by attachment to particles
in the chlorinated surface water system, although the overall water chemistry of thetwo public water supplies was similar The most resistant microorganisms isolatedfrom either water system (gram-positive, spore-forming bacteria, actinomycetes, andsome micrococci) were able to survive exposure to 10 mg/L free chlorine for 2 min(LeChevallier, Evans, and Seidler, 1981) The most chlorine-sensitive bacteria iso-
lated from these two water distribution systems (Corynebacterium/Arthrobacter,
Klebsiella, Pseudomonas/Alcaligenes, Flavobacterium/Moraxella, Acinetobacter, and
most gram-positive micrococci) were readily killed by a chlorine concentration of1.0 mg/L or less The apparent contradiction in occurrence of genera that aregrouped as both chlorine-sensitive and chlorine-resistant reflects variations ofstrains within a genus, physical aggregation of cells, and protective associations withparticulate matter
Actinomycetes and Other Related Organisms. These microorganisms create some
of the objectionable taste and odors reported in water supplies Actinomycetes insource waters of temperate climates may be reduced by approximately tenfold withstorage in holding basins prior to other treatment measures Some drinking watertreatment process configurations that include slow sand filtration or GAC filteradsorbers may support an increase in density and species that enter the distribution
network Nocardia strains were predominant in finished water from water treatment
trains consisting of aeration and filtration or aeration, sand filtration, ozonation, and
activated carbon adsorption Micromonospora made up the greatest percentages in
finished water from treatment chains consisting of flocculation, settling, and slowsand filtration or aeration and granular bed filtration Some growth may also beexpected on PVC-coated walls in finished water reservoirs (Dott and Waschko-Dransmann, 1981) and in the distribution lines where organic material accumulates
in sediments (Burman, 1965; Bays, Burman, and Lavis, 1970) In addition to theirpresence in cold tap water, thermophilic actinomycetes and mesophilic fungi werefound in several hot water samples in three municipalities in Finland (Niemi, Knuth,
and Lundstrom, 1982) Thermoactinomyces vulgaris was the predominant
actino-mycete found in 11 of 15 water distribution systems examined In two studies oftreated water supplies in England, chlorination alone was not effective in eliminating
Streptomycetes Median densities were 2 Streptomycetes per 100 mL of distributed
water Taste and odor complaints involving Actinomycetes often were from waters that had Streptomycetes or Nocardia counts greater than 10 organisms per 100 mL.
Trang 15Fungi. Although many fungi have been found in the aquatic environment (Nagyand Olson, 1982), focus on these organisms in water supply has been limited to theiractive degradation of gasket and joint materials and association with taste and odorcomplaints by consumers (Burman, 1965; Bays, Burman, and Lavis, 1970; Niemi,Knuth, and Lundstrom, 1982) This focus is somewhat surprising because certainfungi are also the cause of nosocomial infections in compromised individuals, andresponsible for allergenic or toxigenic reactions through inhalation in water vapor
or by body contact in showering or bathing
Fungi occur at relatively low densities in ambient waters (Cooke, 1986) Sourcewater storage in temperate climates and chemical coagulation prior to filtration anddisinfection have been found to improve the removal efficiency somewhat, but pro-vide no absolute treatment barrier to these organisms Fungi present in sourcewaters may pass through sand filtration and disinfection treatment processes (Bur-man, 1965; Bays, Burman, and Lavian, 1970; Hinzelin and Block, 1985) In one study(Niemi, Knuth, and Lundstrom, 1982), these organisms were detected in 29 of 32treated water samples
Breakthrough in the treatment barriers, soil contamination from line repairs, andairborne particulates entering storage reservoirs and standpipes are pathways bywhich fungi enter the distribution network (Table 18.4) Fungi occurrences in dis-tributed water are more frequent during summer water temperature conditions
Aspergillus fumigatus was the predominant species detected in the distribution
sys-tem of 15 water supplies in Finland (Niemi, Knuth, and Lundstrom, 1982) A variety
of fungi (Cephalosporium sp., Verticillium sp., Trichodorma sporulosum, Nectria
veridescens, Phoma sp., and Phialophora sp.) were identified in water from service
mains in several water supplies in England (Ridgway and Olson, 1982; Dott andWaschko-Dransmann, 1981) Fungi densities in these drinking waters were usuallyless than 10 organisms per 100 mL
TABLE 18.4 Fungi Densities in Water Supply*
Average density perSystem† Sampling point Number of samples 100 mL
* Data from five small water systems revised from Rosenzweig et al., 1983.
† BG, Bradford Glen water system; BL, Brooklawn water system; MW, Marshallton Woods water system; SR, Spring Run water system; WH, Woodbury Heights water system.
Trang 16Water supplies with fungal densities of 10 to 100 organisms per 100 mL are quently responsible for customer complaints of bad taste and odor (Burman, 1965).The four most frequently occurring genera of filamentous fungi in two distribution
fre-systems (chlorinated and unchlorinated supplies in southern California) were
Peni-cillium, Sporocybe, Acremonium, and Paecilomyces (Nagy and Olson, 1982) In the
unchlorinated system, Penicillium and Acremonium represented approximately 50 percent of 538 colonies identified among 14 genera, while Sporocybe and Penicil-
lium accounted for 56 percent of 923 fungal strains distributed within 19 genera that
occurred in the chlorinated supply The majority of filamentous fungi appeared to benonpathogenic saprophytes The mean density of fungi from the unchlorinated andchlorinated system was 18 and 34 organisms per 100 mL, respectively Conidia of
Aspergillus fumigatus, A niger, and Penicillium oxalicum isolated from distribution
systems of three small water supplies in Pennsylvania showed a greater resistance tochlorine inactivation than yeast (Rosenzweig, Mironigh, and Pipes, 1983)
Yeasts are another category of fungi found in the aquatic environment Whilechemical coagulation and sedimentation will remove 90 to 99 percent of yeasts fromsource water, and granular bed filtration will remove another 90 percent, disinfec-tion is less effective in further yeast reductions The resistance of yeast to free avail-able chlorine is primarily a result of the thick and rigid cell wall, which presents agreater permeability barrier to chlorine Specific species identified in distribution
waters include Candida parapsilosis, C famata, Cryptococcus laurentis, C albidus,
Rhodotorula glutinis, R minuta, and R rubra (Hinzelin and Block, 1985;
Engel-brecht and Haas, 1977) Densities of yeasts reported in finished drinking water age 1.5 organisms per liter While these initial densities are low, yeasts slowlycolonize water pipes and may become more numerous over time
aver-Disinfectant Stability During Water Distribution
Stability of disinfectants during water supply distribution is important, particularly toreduce or prevent colonization by surviving organisms and to inactivate bacteria asso-ciated with the intrusion of contamination in the pipe network Microbial colonizationmay lead to corrosive effects on the distribution systems and adverse aesthetic effectsinvolving taste, odor, and appearance Also, the regrowth of health-related oppor-tunistic organisms and their impact on coliform detection should not be dismissed as
a trivial problem The analysis of data (Table 18.5) taken from the National munity Water Supply Study of 969 public water systems (Geldreich et al., 1972),revealed that standard plate counts of 10 organisms or less were obtained in over 60percent of these distribution systems that had a measured chlorine residual ofapproximately 0.1 to 0.3 mg/L Protective sediment habitats and selective survival ofdisinfectant-resistant organisms were the reasons why residual chlorine concentra-tions greater than 0.3 mg/L chlorine produced no further decreases in the HPC In
Com-an extensive study involving 986 samples taken from the Baltimore Com-and Frederick,Maryland, distribution systems, the maintenance of a free chlorine residual wasfound to be the single most effective measure for maintaining a low standard platecount (Snead et al., 1980)
A disinfectant residual in the distribution system can be very effective in the tivation of pathogens associated with contaminants that are slowly seeping intolarge volumes of high-quality potable water Obviously, limitations to this protectivebarrier exist Free chlorine residuals in distributed water often range from 0.1 to 0.3mg/L, and chloramines may be found to be from 0.2 to 2.0 mg/L Such residuals willnot provide adequate protection against massive intrusions of gross contamination
Trang 17inac-characterized by odors, color, and milky turbidities Studies conducted in a smallabandoned distribution system on a military base indicate that at tap water pH 8,with an initial free chlorine residual of 0.7 mg/L and wastewater added to levels of
up to one percent by volume, 3 logs (99.9 percent) or greater bacterial inactivationwere obtained within 60 minutes Viral inactivation under these conditions was lessthan 2 logs (<99 percent) In laboratory reservoir experiments, where the residualchlorine is replenished by inflow of fresh uncontaminated chlorinated tap water,greater inactivation was observed at the higher wastewater concentrations tested.Furthermore, a free chlorine residual was more effective than a combined chlorineresidual in the rapid inactivation of microorganisms contained in the contaminatedsupply (Snead et al., 1980)
Distribution system problems associated with the use of combined chlorine ual or no detectable residual have been documented in several instances (Langelier,1936; McCauley, 1960; Stumm, 1960) In these cases, the use of combined chlorine ischaracterized by an initial satisfactory phase in which chloramine residuals are eas-ily maintained throughout the system and bacterial counts are very low Over aperiod of years, however, nitrification problems may develop through the applica-tion of an excess amount of free ammonia (>0.1 mg/L), which can promote thedevelopment of nitrifying bacteria (Wolfe et al., 1988; Ike, Wolfe, and Means, 1988).Nitrification also has the adverse impact of neutralizing chloramine residuals andpromoting growth of heterotrophic bacteria (Kirmeyer et al., 1995)
resid-Conversion of a system to free chlorine residual typically produces an initialincrease in consumer complaints of taste and odors resulting from oxidation of accu-mulated organic material, and it may become difficult to maintain a free chlorineconcentration at the ends of the distribution system With application of a systematicmain flushing program in these instances, a free chlorine residual will become estab-lished throughout the system, bacterial counts will decrease, and taste and odor com-plaints will decline (Brodeur, Singley, and Thurrott, 1976)
Ozone is typically applied as a primary disinfectant and because of its high tivity, it is not found in the distribution system, nor would its presence be desirable.Ozone (and other disinfectants like chlorine) can react with nondegradable naturalorganic matter, making it more biodegradable This biodegradable organic matter(BOM) can stimulate bacterial growth, resulting in detectable coliform occurrences,
reac-TABLE 18.5 Effect of Varying Levels of Residual Chlorine on the Total Plate Count
in Potable Water Distribution Systems*
Residual chlorine, mg/LStandard plate
count† 0.0 0.01 0.1 0.2 0.3 0.4 0.5 0.6
<1 8.1 14.6 19.7 12.8 16.4 17.9 4.5 17.91–10 20.4 29.2 38.2 48.9 45.5 51.3 59.1 42.911–100 37.3 33.7 28.9 26.6 23.6 23.1 31.8 28.6101–500 18.6 11.2 7.9 9.6 12.7 5.1 4.5 10.7501–1000 5.6 6.7 1.3 2.1 1.8 0 0 0
>1000 10.0 4.5 3.9 0 0 2.6 0 0Number of
* All values are percent of samples that had the indicated standard plate count.
† Standard plate count (48 h incubation, 35 ° C).
Trang 18taste and odor problems, microbially-influenced corrosion, and potential publichealth concerns.
MICROBIAL COLONIZATION FACTORS
Water mains, storage reservoirs, standpipes, joint connections, fireplug connections,valves, and service lines and metering devices have the potential to be sites suitablefor microbial habitation No pipe material is immune from potential microbial colo-nization once suitable attachment sites are established Given sufficient time,aggressive waters or microbial activity will initiate corrosion of metal pipe surfaces;water characteristics may change the surface structure of asbestos/cement mains;and biological activity creates pitting on the smooth inner surface of plastic pipematerials (See Chapter 17, Corrosion and Deposition Control.) Not all pipe sectionsshow evidence of deterioration even after years of active service; in some cases, thenature of the water chemistry and continuous movement of water under high veloc-ity conditions help to prevent the buildup of chemical and microbial species thatcontribute to corrosion and pitting
in both filtered and nonfiltered source water, unstable coagulants, activated carbonfines, and biological debris
Corrosion of pipe surfaces provides not only a habitat for bacterial proliferation,but also is a source of substrate and protection from chlorine disinfectant residuals
In drinking water systems, the occurrences of coliform bacteria in corrosion cles on iron pipes has been reported by a number of investigators (LeChevallier,Babcock, and Lee, 1987; Opheim, Grochowski, and Smith, 1988; Facey, Smith, andErnde, 1990; Emde, Smith, and Facey, 1992) Laboratory studies showed that thedensity of HPC and coliform group bacteria were 10 times higher when grown onmild steel coupons than on noncorroded polycarbonate surfaces (Camper et al.,1996) The increased surface area due to tuberculation of the pipe walls, the concen-tration of organic substances within the tubercles, and the secretion of organic com-pounds by iron-utilizing bacteria have been postulated as reasons why ironcorrosion stimulates bacteria growth
tuber-LeChevallier, Lowry, and Lee (1990) showed that the disinfection of biofilm ongalvanized, copper, or polyvinyl chloride (PVC) pipes was effective at 1 mg/L of freechlorine or monochloramine but disinfection of organisms on iron pipes was ineffec-tive even at free chlorine residuals as high as 5 mg/L for several weeks Follow-upstudies showed that a combination of the corrosion rate, the ratio of the molar con-
centration of chloride and sulfate to bicarbonate (known as the Larson index), the
chloramine residual, and the level of corrosion inhibitor could account for 75 percent
of the variation in biofilm disinfection rates for microorganisms grown on iron pipes(LeChevallier et al., 1993) Corrosion control through the manipulation of waterchemistry (i.e., pH and alkalinity; Langelier index) or application of phosphate and
Trang 19silicate-based corrosion inhibitors should be thought of as not only protecting thepipe materials, but also as a necessary component of a microbial control plan.
Turbidity and Particle Effects
Particles that cause turbidity in finished water also contribute to sediment lation in the dead ends of the system and within porous scale and tubercle forma-tions Because turbidity is only an indirect measurement of the particulate matter inwater, it provides no specific information regarding the type, number, and size ofparticles being detected Turbidity monitoring in the distribution system, however, is
accumu-a good quaccumu-ality-control praccumu-actice Vaccumu-alues in excess of 1 NTU maccumu-ay signaccumu-al the need toflush the distribution system and to search for areas of pipe corrosion that must bebrought under control
In general, inorganic particles such as clay and water flocculating agents may trap
a variety of organisms These particles, however, appear to have little, if any, tive effect against disinfection action because of the absence of organic demand sub-stances (Hiisverta, 1986) Thus, few viable cells are transported in these kinds ofparticles to potential habitat sites along the distribution system
protec-In contrast, organic particles and algal cell masses in seasonal blooms can be avehicle for the transport of microbial entities through some treatment processes,including disinfection Organic particulates of concern include fecal cell debris(Hoff, 1978), wastewater solids including aggregates of bacteria and virus (Hejkal etal., 1979), protective mats of algal cells, and activated carbon fines that provideattachment sites for associated heterotrophic bacteria (Allen, Taylor, and Geldreich,1980; Ridgway and Olson, 1981; Foster et al., 1980; Camper et al., 1987)
Passage of Microorganisms in Macroinvertebrates
Not commonly recognized as a problem in the quality of distributed water is theoccurrence of various larger, more complex biological organisms including crus-taceans (amphipods, copepods, isopods, ostracods), nematodes, flatworms, water mites,and insect larvae such as chironomids (Small and Greaves, 1968; MacKenthun andKeup, 1970; Geraldi and Grimm, 1982; Chang, Woodward, and Kabler, 1960; Levy etal., 1984; Levy, Hart, and Cheetham, 1986; Zrupko, 1988) While these organisms may
be present in the source water, most are removed by various treatment processes, butsome may succeed in becoming established in filter beds, releasing progeny that cansuccessfully survive disinfection and migrate into the distribution system (Cobb, 1918;George, 1966; Tombes et al., 1979; Mott and Harrison, 1983) In so doing, these inver-
tebrates may harbor and protect coliforms and Legionella from contact with the
dis-infectant at concentrations typically present in distribution systems (Tracy, Camarena,and Wing, 1966; Chang et al., 1960; Smerda, Jensen, and Anderson, 1971; Sarai, 1976).The ingested bacteria may not only survive but multiply within the invertebrate hostand be released at a later time when the host cell bursts or disrupts (Fields et al., 1984;Tyndall and Domingue, 1982) This phenomenon may account for some of the contin-ued release of coliform and other bacteria into the distribution system by passagethrough treatment barriers While some bacterial feeders (protozoans, nematodes,etc.) are inactivated immediately by the chlorine residual in the distribution system,others die slowly or become adapted to the available food sources (biofilms of bacte-ria) in selected pipe sediments or tubercles They may use these sites for attachment,and proceed to harvest the organisms passing in the free-flowing waters
Trang 20Key Factors in Microbial Persistence and Growth
Key factors in the establishment of microbial colonization within the distributionsystem are a source of nutrients, a protective habitat, and a favorable water temper-ature for rapid growth (Geldreich, Nash, and Spino, 1977; Water Research Centre,1977) Not all bacteria that enter the water supply distribution system persist or areable to adapt to this environment and grow (Reasoner, Blannon, and Geldreich,1989; Ridgway, Ainsworth, and Gwilliam, 1978; Victoreen, 1978; Vander Kooij and
Zoeteman, 1978) Within the total coliform indicator group, Klebsiella, Citrobacter, and Enterobacter strains are most often noted in distribution systems (Martin et al.,
1982; Geldreich, Nash, and Spino, 1977; Ptak, Ginsburg, and Willey, 1973) and can
grow with minimal nutrients Other bacteria that may grow include Pseudomonas,
Flavobacterium, Acinetobacter, and Arthrobacter These are especially troublesome
because of their potential interference to coliform detection and acknowledgedroles as opportunistic pathogens (Geldreich, Nash, and Spino, 1977; Hutchinson,Weaver, and Scherago, 1943; Fischer, 1950; Herman and Himmelsbach, 1965; vonGraevenitz, 1977; Herson, 1980)
Bacterial Nutrients. Essential nutritive substances, including those naturallyoccurring and man-made, containing phosphorus, nitrogen, trace metals, and carbonare introduced in varying concentrations from source waters Surface waters receive
a variety of organics discharged in municipal wastewater effluents, industrial wastes,and agricultural activities While some of the nutrient content in dissolved organiccarbon may be removed (20 to 50 percent) through conventional treatment, moreattention to treatment refinements is needed to further reduce trace organic residu-als Applying ozone coupled to GAC or other equivalent biological treatment pro-cesses (e.g., for improved disinfection by-product precursor control) will alsominimize the available organic materials and thereby provide fewer opportunitiesfor microbial biofilm development and coliform growth For those water utilitieswith a relatively clean surface water source that only rely on disinfection treatment,
a seasonal threat of organic contributions from natural lignins, algal blooms, andrecirculating bottom sediments during lake destratification will always exist Theseorganic materials pass into distribution pipe networks, where they stimulate growth
of a wide range of aquatic bacteria (Postgate and Hunter, 1962)
The greatest success in reducing the potential for microbial persistence andgrowth in the distribution system will only be achieved by further reduction of theorganic portion of the essential nutrient base, since the critical inorganic substancesare ubiquitous in the aquatic environment (Allen and Geldreich, 1977; Tuovinen andHsu, 1982).Various studies have shown that biological treatment can be highly effec-tive, with more than 90 percent removal of the biodegradable organic matter (Bour-bigot, Dodin, and Lherritier, 1982; Van der Kooij, Visser, and Hijnen, 1982; Van derKooij, 1987; Janssens, Meheus, and Dirickx, 1984; Pascal et al., 1986; Bablon, Ven-tresque, and Roy, 1987; LeChevallier et al., 1992) Treatment to reduce organic car-bon for bacterial growth suppression is also beneficial in the reduction of taste, odor,color, chlorine demand, and disinfectant by-product formation
Phosphorus. Phosphorus in the environment occurs almost exclusively asorthophosphate (PO4 −), which has a valence state of +5 Although some members of
the genera Bacillus, Pseudomonas, and Clostridium have been shown to reduce
orthophosphate to hypophosphite (PO2 −) and phosphite (PO3 −) under anoxic tions, these transformations are thought to be limited and quantitatively insignificant.Because phosphorus is not consumed by microbial activity (like organic carbon), the
Trang 21condi-turnover rate of phosphorus in aquatic habitats can overcome low levels of phate in the water column Although some researchers have suggested that certainwaters may be phosphate-limited (Herson, Marshall, and Victoreen, 1984; Haas, Bitter,and Scheff, 1988), Rosenzweig (1987) found that phosphate-based corrosion inhibitorsdid not significantly influence the growth of several strains of coliform bacteria Highlevels of Virchem 932, a zinc orthophosphate, showed inhibitory effects for certain coli-form species.
orthophos-Nitrogen. The basic requirements for nitrogen in the metabolic processes of teria can be satisfied in a variety of ways.Watershed conditions involving wastewatereffluents released upstream of a water intake, landfill operation in the vicinity ofgroundwater aquifers poorly protected by the soil barrier, and seasonal application
bac-of farm and garden fertilizers over the watershed are bac-often the major contributors bac-ofnitrogenous compounds
Water treatment practices must also be carefully controlled to minimize addition
of nitrogen For example, application of ammonia to form chloramines may tribute excess ammonia to water Ammonia is an electron donor for autotrophicbacteria and can promote bacterial growth in distribution systems Rittmann andSnoeyink (1984) found that ammonia concentrations in groundwater supplies werefrequently high enough to cause biological instability The proliferation of ammonia-oxidizing bacteria (nitrification) can lead to accelerated loss of chlorine residuals,increased nitrite levels, and stimulated growth of HPC bacteria As reviewed byCrowe and Bouwer (1987), biological treatment techniques are available for theremoval of ammonia and nitrate from water The exact role of nitrogen in growth of
con-coliform bacteria is unclear, especially because some strains of Klebsiella can fix
molecular nitrogen
Metal Ions and Salts. Trace amounts of metal ions (Fe++, Fe+++, Mg++, and others)available in salts appear to contribute to the nutrient base required for microorgan-isms Victoreen (1977, 1980, and 1984) indicated that iron oxide stimulated thegrowth of coliform bacteria Trace amounts of copper salts, Mg, and Mn ions areneeded by other heterotrophic bacteria in the development of their normal meta-bolic processes (Laskin and LeChevallier, 1977) within the biofilm consortium
Carbon. Organic carbon is utilized by heterotrophic bacteria for production ofnew cellular material (assimilation) and as an energy source (dissimilation)(LeChevallier, Welch, and Smith, 1996) Most organic carbon in water supplies isnatural in origin and is derived from living and decaying vegetation These com-pounds may include humic and fulvic acids, polymeric carbohydrates, proteins, andcarboxylic acids In the U.S EPA National Organic Reconnaissance Survey (Symons
et al., 1975), the nonpurgable total organic carbon concentration of finished ing water in 80 locations ranged from 0.05 to 12.2 mg/L, with a median concentration
drink-of 1.5 mg/L Because heterotrophic bacteria require carbon, nitrogen, and rus in a ratio of approximately 100:10:1 (C:N:P), organic carbon is often a growth-limiting nutrient
phospho-Measuring Biodegradable Organic Matter in Water
Two of the most common measurements of the biostability of water include mination of the assimilable organic carbon (AOC) content and the biodegradabledissolved organic carbon (BDOC) level Measurement of AOC is based on a bioas-
Trang 22deter-say of two test strains (Pseudomonas fluorescens strain P17 and Spirillum sp strain
NOX) Bacterial growth is monitored in the water samples (e.g colony counts, ATPmeasurements), and the maximum growth (Nmax) observed during the incubation isconverted into AOC by using growth yield of the bacteria from calibration curvesperformed on known concentrations of standard organic compounds (e.g acetate,oxalate) (EPA, 1976; Van der Kooij and Hijnen, 1985) Assimilable organic carbon isexpressed as micrograms of acetate- (or oxylate-) carbon equivalents Biodegrad-able dissolved organic carbon evaluates the reduction in DOC levels following incu-bation of the water sample with microorganisms (Pascal et al., 1986; Servais, Billen,and Hascoet, 1987; Joret, 1988; Servais, Laurent, and Randon, 1993) The differencebetween the initial and final DOC levels is the biodegradable organic carbon frac-tion
Protective Habitats in Pipe Networks. Porous sediments and tubercles in the pipenetwork appear to adsorb and concentrate nutrients Bacteriological colonizationfound in encrustations, taken from water main sections removed from four munici-pal water systems during line repairs, demonstrated bacterial densities that rangedfrom 390 to 760,000 bacteria per mL (Allen and Geldreich, 1977) Variations in bac-terial densities will depend on the tubercle characteristics and fraction examined(Tuovinen and Hsu, 1982) Using the scanning electron microscope to locate micro-colony sites in tubercles, microorganisms observed were predominantly found at ornear the surface This is the area where encrustation, water interface, nutrients, andoxygen are constantly present (Allen, Taylor, and Geldreich, 1980; Tuovinen et al.,1980)
To colonize surfaces in contact with a flowing stream of water, bacteria must
adhere tenaciously They do so by means of a mass of tangled fibers (called a
glyco-calyx) that extends from the cell membrane and adheres to surrounding surfaces or
other bacteria (Costeron, Geesey, and Cheng, 1978; Bitton and Marshall, 1980).These glycocalyx adhesions to sediment coatings, tubercles, pipe joints, and roughwall surfaces prevent most of the individual cells of the microcolony from beingswept away by the shearing force of flowing water Because these appendages arepolysaccharide materials, they may also serve as a protective barrier against thelethal effects of residual disinfectants
Not all areas of the distribution network are favorable for microbial tion High water velocity in smooth pipe sections makes microbial attachment diffi-cult, increases nutrient flux, and provides less protection from disinfectant exposure(Victoreen, 1978) In contrast, sediment accumulations, tubercle development, andscale formation in low-flow, dead-end, and rough-walled pipe sections and connect-
prolifera-ing joints are prime sites (Figures 18.1a to d) for microbial growth (Olson, 1982;
Allen, Taylor, and Geldreich, 1980; Victoreen, 1978; Allen and Geldreich, 1977; toreen, 1974; Victoreen, 1977; Ainsworth, Ridgway, and Guilliam, 1978; Lee, O’Con-nor, and Banerji, 1980) Therefore, finding patchiness in biofilm development alongpipe surfaces should be expected
Vic-Water Supply Storage. Finished water reservoirs, water storage tanks, and pipes cannot be excluded as potential sites for biofilm development Operationalconditions that may encourage bacterial colonization include prolonged storagetime, reduced flow velocities, and infrequent cleaning Water is often not drawn fromthe bottom of reservoirs, so accumulations of sediments may occur Sedimentbuildup in reservoirs is more of a problem for systems that use unfiltered surfacewaters and limit treatment to disinfection Filtration of surface source waters willremove biological materials (e.g., algae and vegetation debris) and other suspended
Trang 23stand-solids that otherwise pass through the disinfection process and ultimately late in low-flow pipe sections and finished water storage facilities These sedimentscan contribute significant amounts of assimilable organic complexes that supportbacterial colonization of the distribution network.
accumu-Slime or biofilm development in cement structures may develop on a cementmortar with plastic additives, on sealers such as epoxy resin, bitumen, and PVC film,and on areas of cement erosion (Schoenen, 1986) Metal structures are also subject
to microbial activity in areas with corrosion activity such as seams and joint struction bonds Porous materials such as brick and wood used in reservoirs are par-ticularly suited for microbial colonization and may be difficult to dislodge byflushing and disinfection treatment
con-FIGURE 18.1 (a) Crosssection of pipe tubercle showing loose surface material at water face and compaction of deposits near the pipe wall Sample magnification, 100× (b) Bacterial
inter-colonization site in porous tubercle encrustation Electron micrograph magnification, 3000×
Material supplied from the Cambridge, Massachusetts, distribution system pipe network (c)
Attachment site for bacterial growth in tubercle surface material Electron micrograph fication, 4000× Material supplied from the Cambridge, Massachusetts, distribution system pipe
magni-network (d) Protected habitat and microbial community in porous encrustation Electron
micrograph magnification, 11,000× Material supplied from the Cambridge, Massachusetts, tribution system pipe network
Trang 24Tanks constructed of redwood are common in the western United States, beingused by small communities, state and federal recreational areas, mobile home parks,and motels The problem with redwood used in some water storage tanks is the pres-ence of microbial colonization that impacts on compliance with the drinking watercoliform standard Research on coliform occurrences in redwood storage tanks
showed that environmental strains of Klebsiella colonize the wood tissues of trees The association begins at embryo fertilization of the tree seed Klebsiella is intro-
duced to the embryo from insect contact or windblown dust and then colonizes thewood pores, receiving water and nutrients from the xylem and phloem tissues of thetree (Knittel, Seidler, and Cake, 1977) This coliform metabolizes the leached-outwood sugars (cyclitols) from the staves as a source of nutrients (Seidler, Morrow, andBagley, 1977; Talbot and Seidler, 1979) As a consequence, new redwood tanks were
found to be the source of Klebsiella in the water supply and the cause of massive
biofilm development over the inner surfaces of the tank (Seidler, Morrow, andBagley, 1977) The problem is most acute in new redwood tanks that are unlined.Disinfection and scraping the wood staves were ineffective in eliminating the bacte-ria because the organisms persist deep inside the wood pores, until all the availablewood sugars (cyclitols) are leached away or biodegraded The problem can be con-trolled by maintaining a free chlorine residual of 0.2 to 0.4 mg/L until the availablesugar supply is leached away with tank usage over a two-year period (Talbot, Mor-row, and Seidler, 1979) More effective is the use of plastic or fiberglass liners to pre-vent leakage and release of bacterial colonization into the water supply
As a preventative measure to avoid water quality degradation in all types ofwater storage structures, regularly scheduled, systematic inspections are recom-mended for slime development on structural surfaces in contact with the water, andfor the accumulation of sediments that serve as protective sites for microorganisms.These growths and deposits should be removed to reduce sites where taste and odorproblems and microbial growth originate
Water Temperature Effects. Microbial growth is not only keyed to bacterial strainsthat quickly adjust to limited nutrient sources, but also to water temperature Watertemperature above 50°F (10°C) accelerates the growth of adapted organisms withslow generation times Low water temperatures result in a precarious balancebetween new-cell development and the death of old cells Data available from watersystems in geographical areas with pronounced seasonal temperature changes sug-gest that growth of many heterotrophic bacteria is more pronounced than the col-iform subset of this population, often providing abrupt surges in density during thesummer (Geldreich, Nash, and Spino, 1977; Howard, 1940) Renewed growth wasnot, however, correlated with temperature in southern California groundwater sys-tems (Olson, 1982), possibly because the water temperatures common to those sys-tems were consistently above 50°F (10°C) and available assimilable organics werevery low Accumulations of nutrient particles in pipelines during winter periods ofminimal microbial activity may be the key to summer bacterial growth surges in sur-face water systems
MONITORING FACTORS
A reliable community water supply system will have a record of no associated ease outbreaks This approach to demonstrating the acceptable quality of a publicwater supply provides no opportunity to intervene in potential contamination