Abstract: Treatment trials have been realized with a new compact process in order to meet the drinking water regulation newly implemented in the province of Quebec. This process is especially designed for the supply of small communities. It combines complementary treatments operated by a centralized computer: ozonation, membrane filtration, and biological filtration, thus reducing operation to basic tasks. The process functioned very well for the period under review, in spite of extreme conditions met: cold water and constant fluctuation of the raw water quality. The process considerably lowers the contents of organic compounds (TOC, DOC, and color), the turbidity, the chlorine demand, and the concentrations of trihalomethane precursors present in the raw water. For this specific application, the backwash wastewaters may be mixed and discharged directly into a river nearby.
Key words: drinking water, small communities, ozonation, membrane filtration, biological activated carbon, turbidity, trihalomethane precursors, TOC.
Resume: Des essais de traitement ont ete realises dans le cadre de l'accreditation d'un nouveau procede par le Comite sur les technologies de traitement en eau potable du gouvernement du Quebec. Le procede teste est compact et il est concu specialement pour alimenter les petites communautes. Il combine differentes etapes de traitement complementaires (ozonation, filtration sur membrane et filtration biologique) et commandes par un systeme de controle automatique. Le procede a tres bien fonctionne pendant la periode de suivi, malgre les conditions extremes rencontrees : eaux froides et fluctuation constante de la qualite de l'eau brute a traiter. Le procede abaisse considerablement les teneurs de composes organiques (COT, COD et couleur vraie), la demande en chlore, les concentrations de precurseurs de trihalomethanes (SDS-THM) et la turbidite presentes dans l'eau brute. Dans le cas experimente, les eaux de lavage du tamis, des membranes et des unites de filtration peuvent etre melangees et rejetees directement dans le milieu environnant.
Mots cles: eau potable, petites communautes, ozonation, membrane, traitement biologique, turbidite, precurseurs de trihalomethane, COT.
Stricter regulations concerning drinking water quality have been implemented in Canada. Small communities need to upgrade their existing drinking water treatment plants in order to respond to these new regulations. They necessitate compact treatment plants efficient and easy to operate. Premanufactured compact treatment plants are easily transported in a shipping container, which minimize construction costs, and they do not necessitate large on-site infrastructures.
In the province of Quebec, the new regulation requires that the treated water has trihalomethane levels below 80 [micro]g/L, that turbidity of the treated water is low (below 0.1 nephelometric turbidity units, NTU, for membrane treated water) and that high removal of pathogen microorganisms occurs (the treated water must not contain pathogen organisms or indicator organisms of a fecal contamination, such as fecal coliforms, Escherichia coli, enterococcus bacteria, and coliphage viruses; Government of Quebec 2001).A new compact process must aim to meet this regulation. It has to combine complementary treatments removing particles, trihalomethane precursors, and microorganisms. The combination of ozone (inactivating pathogens and oxidizing trihalomethane precursors), membrane filtration (removing particles), and biological filtration (reducing chlorine demand and removing biodegradable organic matter) seems to be appropriate to achieve these objectives. The coupling of ozonation and biological activated carbon filtration is very effective to remove trihalomethane precursors (Niquette et al. 1999). The addition of a clarification step would help to remove pathogens and organic matter, but it needs the addition of coagulants and it produces sludge. Such a clarification step is not easy to operate in small communities where qualified staff is rare and, often, there is no municipal sewer where the clarification sludge may be disposed of.
The objective of the present study is to evaluate the combination of these processes. The government of Quebec requires that any new water treatment process equipment must undergo a 3 month pilot-scale trial. An official committee verifies their efficiency and, if the trial is successful, they issue an official accreditation. This accreditation enables the implementation of the full-scale treatment plant.
Materials and methods
A trial was conducted with the abovementioned compact treatment process. This trial was undertaken in the autumn and winter, from 17 October 2003 to 20 January 2004, when the raw water quality is fluctuating. The raw water quality was very unstable during the period in question. The raw water was pumped from the Yamaska River at the municipality of Farnham. This surface water is very contaminated by organic compounds, particles, and pathogenic microorganisms.
The process tested was the pilot plant version of the DaguaFlo UMF. It is very compact and it does not require coagulants. The flow schematic of the DaguaFlo UMF, pilot version, is shown in Fig. 1. The pilot plant is designed to treat 28 L/min (7.4 US gal/min) and it is automatically operated in a continuous mode, 24 h a day, 7 days a week. This process is composed of the following modules: micro rotary sieve of 21 [micro]m, ozonation, ultra-micromembrane filtration, and biological activated carbon (BAC) filtration. A polyvinylidene fluoride (PVDF) venturi (Mazzei, model 978) insures ozone injection. An ozone residual concentration between 0.62 and 0.69 mg/L is maintained at the ozone contactor outlet, which has a detention time of 11.6 min. These conditions combined with the hydraulic efficiency of the ozonation contactor, which has four contact cells, insure a 3.09 log removal of Giardia cysts and a 6.67 log removal of viruses (MENV 2001).
[FIGURE 1 OMITTED]
The ultra-micromembrane is composed of inside-out PVDF capillary. The membrane was resistant to ozone and the presence of ozone on the membrane reduced the chemical backwashing frequency of the membrane. In fact, no chemical cleaning of the membrane was made during the 3 month trial because the transmembrane pressure of the membrane was stable (20 psig; 1 psi = 6894.76 Pa) and below the critical pressure (35 psig) recommended by the manufacturer. Only air-water backwashing was performed. The BAC filters had an empty bed contact time of 6.9 min. The adsorption capacity of the activated carbon (Calgon F400; effective size: 0.55-0.75 mm; uniformity coefficient: 1.9) had been saturated before the trial. The level of organic matter entering and exiting the BAC filter was followed with UV absorbance measurements during the start-up phase and the official study started when these levels were equivalent. Only adsorption can remove the organic UV absorbing compounds, which are double and triple bonded carbon compounds that autochthonous bacteria hardly remove. The industrial version of the process includes thereafter UV lamps insuring an additional 3 log removal of Cryptosporidium cysts and 3 log removal of Giardia cysts (to obtain the removal criteria of MENV 2001). This need for UV lamps would be eliminated if the government agrees with the latest tests with latex microspheres showing that the membrane is able to remove more than 5 log of these parasites. The industrial version also includes a chlorination step at the end of the process. So, the industrial system respects the disinfection regulation (Ct; C is disinfection concentration and t is effective contact time) for surface waters having a very strong contamination by pathogenic microorganisms (MENV 2001). This industrial version may serve small communities with a nominal drinking water production varying from 50 [m.sup.3]/d to 4 900 [m.sup.3]/d. Figure 2 shows an example of a full-scale plant in a container able to treat 820 [m.sup.3]/d (150 US gal/min) and able to supply about 1800 persons.
[FIGURE 2 OMITTED]
The analytical parameters were measured by a professional laboratory according to standard methods (APHA et al. 1995). They include total organic carbon (TOC; Method # 5310 C: Persulfate-Ultraviolet Oxidation Method), dissolved organic carbon (DOC; Filtration on 0.45 [micro]m followed by Method # 5310 C: Persulfate-Ultraviolet Oxidation Method), UV transmittance (Method # 5910 B: Spectrophotometric Method), true color (Method # 2120 C: Spectrophotometric Method), fecal coliforms (Method # 9222 D: Membrane Filter Procedure), total coliforms (Method # 9222A: Membrane Filter Technique), heterotrophic plate counts (HPC; Method # 9215), total iron (acid digestion followed by Method # 3120 B: Induction Coupled Plasma Analysis), total manganese (acid digestion followed by Method # 3120 B: Induction Coupled Plasma Analysis), ammonia (Method # 4500-N[H.sub.3] D: Ammonia-Selective Electrode Method), nitrites (Method # 4500-N[O.sub.2.sup.-] B: Colorimetric Method), nitrates (Method # 4500-N[O.sub.3.sup.-] D: Nitrate Electrode Method), chlorine demand (Method # 2350 B: Chlorine demand based on a residual chlorine of 0.5[+ or -]0.2 mg/L after 24 h), trihalomethanes formed in conditions of distribution (SDS-THM; Method # 5710 C with incubation at in situ pH and temperature, incubation time of 24[+ or -]1 h; residual chlorine of 0.5[+ or -]0.2 mg/L), total hardness (based on calcium and magnesium concentrations), calcium (acid digestion followed by Method # 3120 B: Induction Coupled Plasma Analysis), total alkalinity (Method # 2320 B: titration method), suspended solids (Method # 2540 D: total suspended solids dried at 103-105[degrees]C), conductivity (Method # 2510), bromides (Method # 4500-[Br.sup.-]: phenol red colorimetric method), and bromates (Method # 4110: Ion Chromatography). These parameters were measured with a frequency respecting the sampling protocol prescribed by the committee.
On-line parameters were also measured: turbidity (measured by three Wallace & Tiernan turbidimeters, model TMS 561), temperature (thermometer), pH (Hannah electrode), and ozone residual (Prominent, model OZE 3). The ozone electrode was calibrated using a colorimetric method (Hach, DR-2000). Table 1 summarizes the location of the samples taken and the results of the analyses performed during the 3 month trial.
Results and discussion
Parameters measured on the raw water and the treated water
The Daguaflo-UMF process functioned very well for the period under investigation, in spite of the extreme conditions encountered: cold water and constant fluctuation of the raw water quality, which was much degraded.
The process considerably lowers the content of organic compounds. The physical removal by the membrane (for the removal of particulate organic matter) and the combined action of the ozone contactor and the BAC (for the removal of biodegradable dissolved organic carbon) reduce many parameters linked to the organic content of the treated water: TOC, DOC, true color, SDS-THM, and UV Transmittance. The concentration profiles of TOC and COD are respectively shown in Figs. 3 and 4. Average removals of TOC and COD through the study where 38.5% and 31.4%, respectively. The true color of the treated water, shown in Fig. 5, was well below the aesthetic criteria of 15 true color units (TCU) prescribed by the Canadian recommendations (Government of Canada 2004). They were below the detection limit of the analytical method used (2.0 TCU). The concentrations of SDS-THM(Fig. 6) measured in the raw water, 135.5 [micro]g/L, were reduced below the maximum regulated in the province of Quebec, 80 [micro]g/L (Government of Quebec 2001), and the maximum recommended by the government of Canada (2004), 100 [micro]g/L. Another study undertaken with a similar process in similar conditions (results not shown) revealed that ozonation and BAC filtration considerably reduced the concentrations of SDS-THM in spite of the cold temperature of the treated water (Niquette 2003). Ultraviolet transmittance of the treated water increased with an average value of 88.9% after treatment (Fig. 7).
[FIGURES 3-7 OMITTED]
The turbidity of the raw water was considerably lowered by the process. Figure 8 shows the raw water turbidity and Fig. 9 illustrates the turbidity of the treated water. Even when the raw water turbidity reached 134 NTU, the membrane process constantly maintained turbidity below the 0.1 NTU prescribed for membrane filtration (Government of Quebec 2001) and treated water had a turbidity lower than 0.5 NTU (MENV 2001). This efficiency indicates that the membrane was working satisfactorily.
[FIGURES 8-9 OMITTED]
Bacteriological results showed that the use of the Daguaflo-UMF process followed by a final disinfection would meet the standards. In spite of the high levels of total coliforms (Fig. 10) and fecal coliforms (Fig. 11) detected in the raw water, no coliform was detected in the treated water. The high dose of ozone residual maintained in the ozone contactor, between 0.62 and 0.69 mg/L, is responsible for this efficiency.
[FIGURES 10-11 OMITTED]
No bromide was detected in the influent (below 0.1 mg/L) of the pilot process and no bromate was detected in the effluent. The bromate concentrations were below the detection limit of the analytical method used (2.0 [micro]g/L), thus satisfying the maximum concentration of 10 [micro]g/L (Government of Quebec 2001; Government of Canada 2004). Song et al. (1997) found that bromate ion formation is an important consideration for waters containing more than 0.10 mg/L bromide ion. The iron concentrations in treated water are very low (average of 0.03 mg/L).
The pH (close to neutrality), total hardness, calcium concentration, total alkalinity, and conductivity remained at stable values before and after treatment. The average concentrations of iron, ammonia, and nitrites were very low in the treatment effluent (corresponding average of 0.03 mg Fe/L, 0.07 mg N[H.sub.4]-N/L and 0.02 mg N[O.sub.3]-N/L). The measured temperatures of the raw water were very cold, between 2.0[degrees]C and 9.2[degrees]C. The biological filtration was thus functioning at it lowest capacity. The combined concentration of nitrates and nitrites of 0.89 mg/L was well below the regulated levels of 10 mg/L in Quebec (Government of Quebec 2001) and 45 mg/L in Canada (Government of Canada 2004). Except for one measurement on 28 October, manganese concentrations were below the aesthetic standard of 0.05 mg/L as set by the Government of Canada (2004). The Mn concentration of 0.063 mg/L measured in the effluent on 28 October is not a problem because the concentration of manganese measured in the raw water this day, 0.28 mg/L, is unusual for surface water. Iron and manganese removals could also be optimized by the adjustment of the ozone treatment that is transforming these metals into particles, which are removed by the following membrane filtration (Bellamy et al. 1991).
Parameters measured on the backwash wastewaters
The membrane, resisting ozone, did not require any chemical backwashing. The industrial version of the process will be equipped with a peroxide feed in case chemical cleaning would be necessary. The backwash wastewaters of the micro rotary sieve and of the membrane could be mixed in a settling pond (in the case of an industrial process), where the residual ozone will be dissipated and where the suspended matter will settle. This mixing will attenuate the impact of the discharge and small concentrations of suspended solids are thus released into the environment. A mixing test was carried out on the proportional mixture of the backwash wastewaters of the micro rotary sieve and the backwash wastewaters of the membrane (see Table 2). Their combination meets the maximum level for suspended solids, 20 mg/L (MENV 2001), that could be released directly in the environment. The spent backwash waters of the BAC filters satisfies this maximum limit, with an average load of suspended solids of 6 mg/L and a maximum load of 12 mg/L.
The official Quebec committee judged that the quality of the results obtained at the time of the process trial validated its use as a full-scale drinking water treatment. The application of the process remains limited to all the raw waters having characteristics that correspond to the obligatory parameters summarized in Table 3. The values of these regulated parameters were met during this study.
The high UV transmittance of the treated water (Fig. 7) and the low iron concentrations in the treated water facilitates the use of an UV lamp as post-disinfection of the treatment. The removal of coliforms obtained during this study showed that the ozonation step already inactivated pathogen microorganisms.
In spite of a turbidity peak reaching 134 NTU in the raw water, the water treated by the membrane had a turbidity (average of 0.03 NTU) lower than the standard of 0.1 NTU prescribed for membrane filtration (Government of Quebec 2001) and the treated water had a turbidity lower than 0.5 NTU (MENV 2001). No bromate was detected in the effluent of the pilot process, thus the standard of 10 [micro]g/L was met (Government of Quebec 2001; Government of Canada 2004). Regardless of the cold temperature of the treated water (minimum of 2.0[degrees]C), the biological activity in the activated carbon filter and the ozone contactor considerably reduced the concentrations of precursors of trihalomethanes. As for the other measured physicochemical parameters, the membrane process preceded by an ozonization removes iron and manganese present in the raw water without demineralizing the treated water. In fact, the alkalinity, the conductivity, the calcium concentration, and the hardness of the raw water were unchanged. The nitrite concentrations and the concentrations of nitrites and nitrates were clearly below the regulated levels.
Concerning the backwash wastewaters from this process, they could be mixed in a settling pit before disposal. The possible use of such a pit in future applications has to be determined by additional pilot tests. The amount of suspended solids present in the backwash wastewaters would determine if these backwash wastewaters could be released into a river or, at best, in a nearby sewer, which is often not available in rural communities.
Received 13 February 2005. Revision accepted 23 June 2006. Published on the NRC Research Press Web site at http://jees.nrc.ca/ on 15 March 2007.
APHA, AWWA, WEF. 1995. Standard methods for the examination of water and wastewater. 19th ed. American Public Health Association, Washington, D.C. 1-205.
Bellamy, W.D., Damez, F., Langlais, B., Montiel, A., Rakness, K.L., Reckhow, D.A., and Robson, C.M. 1991. Ozone in water treatment: application and design, engineering aspects. Edited by B. Langlais, D.A. Reckhow, and D.R. Brink. AWWA Research Foundation and Lewish Publishers, USA.
Government of Canada. 2004. Summary of guidelines for Canadian drinking water quality. Prepared by the Federal-Provincial-Territorial Committee on Drinking Water of the Federal-Provincial-Territorial Committee on Health and the Environment. p. 11.
Government of Quebec. 2001. Regulation about the drinking water quality. Law about the environment quality (L.R.Q. c. Q-2, a. 31, par. e, h.1 and h.2, a. 45, a. 45.2, par. a, a. 46, par. a, b, d, m, o, o.1 and o.2, a.87, par. a and b, a. 109.1 and a. 124.1). 29 May.
Niquette, P. 2003. Analysis and interpretation of results from a pilot study: DaguaFlo UMF process at Farnham. University of Quebec at Montreal. Report presented to the ministere de l'environnement and to the ministere des affaires municipales et de la metropole of the Quebec's government. p. 25.
Niquette, P., Prevost, M., Merlet, N., and Lafrance, P. 1999. Influence de facteurs controlant l'enlevement de la demande en chlore et de precurseurs de sous-produits de chloration dans des filtres biologiques. Water Res. 33(10): 2329-2344.
Quebec's ministry of environment (MENV). 2001. Official conception guide of drinking water installations.
Song, R., Weaterhoff, P., Minear, R., and Amy, G. 1997. Bromate minimization during ozonation. J. Am. Water Works Assoc. 89(6): 69-78.
P. Niquette (1) and R. Hausler. Ecole de technologie superieure, STation Experimentale des Procedes Pilotes en Environnement (STEPPE-ETS), 1100, Notre-Dame West, Montreal, QCH3C 1K3, Canada.
P. Lahaye and M. Lacasse. Dagua inc. 1281, Industriel boulevard, Granby, QC J2J 2B8, Canada.
Written discussion of this article is welcomed and will be received by the Editor until 31 July 2007.
(1) Corresponding author (e-mail: firstname.lastname@example.org).
Table 1. Analyses made during the 3 month trial. Water after Treated Parameters Raw water membrane water pH (on-site) 13 13 Temperature (on-site) 13 13 Fecal coliforms 13 13 Total coliforms 13 13 HPC 13 13 True color 13 13 Total organic carbon 13 13 Dissolved organic carbon 13 13 Turbidity 13 13 13 Suspended solids UV transmittance (254 nm) 13 13 Ammonia 3 3 Nitrites 3 3 Nitrates and nitrites 3 3 Chlorine demand 3 3 Total alkalinity 6 6 Calcium 6 6 Total hardness 6 6 Total iron 6 6 Total manganese 6 6 Dissolved solids 3 3 Suspended solids 3 3 Conductivity 3 3 SDS-THM 3 Bromates 3 Bromides 3 21 [micro]m Membrane Sieve spent BAC filter spent backwash spent backwash Parameters waters waters waters pH (on-site) 13 Temperature (on-site) 13 Fecal coliforms Total coliforms HPC True color Total organic carbon 3 Dissolved organic carbon 3 Turbidity 3 Suspended solids 3 3 3 UV transmittance (254 nm) Ammonia Nitrites Nitrates and nitrites Chlorine demand Total alkalinity Calcium Total hardness Total iron 3 Total manganese 3 Dissolved solids 3 Suspended solids 3 Conductivity SDS-THM Bromates Bromides Table 2. Suspended solids in the backwash wastewaters. Suspended solids (mg/L) Mixed backwashing waters 10 Backwashing waters after a settling of 15 min 6 Backwashing waters after a settling of 30 min 4 Table 3. Official chart for the use of DaguaFlo UMF as a full-scale drinking water plant. Measured parameters Raw water Turbidity (NTU) (95% of the samples) <55 Turbidity (NTU) (maximum) 134 TOC (mg/L) (95% of the samples) <8.7 TOC (mg/L) (maximum) 12 Color (TCU) (95% of the samples) <53 Color (TCU) (maximum) 56 Bromides ([micro] g/L) (maximum) 100 Other parameters measured at Farnham Fecal coliforms (CFU/100 mL) (maximum) 20 000 Temperature 2.0 to 9.2[degrees]C pH 6.9 to 7.6 Total alkalinity (mg/L Ca[CO.sub.3]) 39 to 50 SUVA (L [mg.sup.-1][cm.sup.-1]) 2.5 to 5.8…