Effect of Suspended Particles on Disinfection of a Physicochemical Municipal Wastewater with Peracetic Acid

Water Qual. Res. J. Can. 2008 Volume 43, No. 1, 47-54 Copyright 2008, CAWQ Effect of Suspended Particles on Disinfection of a Physicochemical Municipal Wastewater with Peracetic Acid Dario Falsanisi, 1 * Ronald Gehr, 2 Lorenzo Liberti, 1 and Michele Notarnicola 1 1 Technical University of Bari, Department of Environmental Engineering and Sustainable Development, Taranto 74100, Italy 2 McGill University, Department of Civil Engineering, Montreal H3A 2K6, Canada The aim of this study was to elucidate the role of total suspended solids (TSS, which are claimed to be responsible for protecting embedded microorganisms) during disinfection of wastewater with peracetic acid (PAA). In particular, the focus was on a physicochemical effluent, having been treated with ferric chloride for TSS and phosphorus removal. Batch disinfection tests with various PAA dosages and contact times were carried out on the effluent from Montreal s wastewater treatment plant which uses only chemical precipitation and primary sedimentation. In addition to these samples, disinfection of medium and highly filtered effluents, obtained by sequential filtration through 120- and 10-μm nominal pore size membrane laboratory filters, was investigated. Modified second-order and Selleck model kinetics were used with moderate success to describe disinfectant consumption and microbial inactivation rates, respectively. This study showed that the overall amount of protection afforded by TSS to the microbial indicator considered (i.e., fecal coliforms in this case) was approximately 1.9 logs. TSS size was a key variable in this protection; approximately 1.3 logs and an additional 0.6 logs was the protection afforded by TSS greater than 120 μm, and between 10 μm and 120 μm, respectively. Fecal coliform inactivation of 3.8 to 7.3 logs after 40 minutes contact time could be achieved with a PAA dosage of 2 or 8 mg/l, respectively, in a highly filtered effluent. Key words: wastewater disinfection, peracetic acid, particle associated bacteria, modelling microbial inactivation Introduction Total suspended solids (TSS) are usually blamed for impairing (waste)water disinfection with chemical (chlorine, ozone, peracetic acid) or physical (ultraviolet [UV]) methods because free-swimming bacteria (i.e., bacteria dispersed in solution, not embedded within particles) may be easily inactivated, whereas particle-associated bacteria (PAB) may resist even high disinfectant dosages (Dietrich et al. 2003; Bohrerova and Linden 2006). Regardless of the disinfection method applied, coliform bacteria are expected to be well protected by TSS larger than 10 μm; this is the conventional size threshold dividing free-swimming bacteria from PAB (Parker and Darby 1995; Emerick et al. 1999; Emerick et al. 2000). The specific role of TSS on the disinfection process, however, is not straightforward. For UV disinfection, secondary effluents filtered through <10-μm filters were more effectively disinfected than the corresponding unfiltered samples, where protection of bacteria usually increased with average TSS particle size (Jolis et al. 2001; Madge and Jensen 2006). There were similar findings with chlorination, where the capability of chlorine to diffuse into solid particles also depends on the type of disinfectant (i.e., chlorine or chloramine) (Stewart et al. 1990; De Beer et al. 1994). * Corresponding author: d.falsanisi@poliba.it The influence of TSS appears to be particularly controversial in wastewater disinfection with peracetic acid, CH 3 COOOH (PAA). In general, PAA effectiveness decreases with TSS concentration which, while highly relevant at 100 mg/l, is still appreciable at lower TSS concentrations (<10 mg/l) (Gehr et al. 2002; Kitis 2004; Carrasco and Turner 2006). Contradictory results, however, are reported when advanced disinfection is required for wastewater reuse. Indeed, to achieve the well known Californian standard for unrestricted wastewater reuse in agriculture (2 CFU of total coliforms per 100 ml) with a tertiary effluent characterized by low TSS concentration, Mandra et al. (1996) and Liberti et al. (1999) required very high PAA dosages (400 mg/l) and 20 minutes contact time, while Koivunen and Heinonen- Tansky (2005) found that 7 mg of PAA per litre and 30 minutes contact time was sufficient. On the other hand, PAA doses as low as 4 mg/l and 10 minutes contact time were reportedly able to meet the Italian microbial standard for unrestricted wastewater reuse in agriculture (10 CFU of Escherichia coli per 100 ml) with a good quality secondary effluent (Dell Erba et al. 2004). In general, microbial inactivation, together with disinfectant consumption, is the most important information required to properly design a disinfection reactor. To this aim, microbial inactivation is reproduced with models that represent a simplification and idealization of complex reactions between disinfectant and target organism(s). The most widely used microbial inactivation kinetic models used are shown in Table 1. As TSS is a gross parameter that does not provide specific information about the extent of microbial 47

Falsanisi et al. protection afforded, the objective of this study was to try to elucidate the role of TSS when PAA is used for advanced disinfection of municipal wastewater. To that aim, effluent samples from the Montreal Wastewater Treatment Plant (MWTP) were sequentially and sizeselectively filtered, and the germicidal efficacy of PAA against fecal coliforms was evaluated in laboratory batch tests. Materials and Methods The investigation focused on effluent from the MWTP, an advanced physicochemical primary plant which uses screening, grit removal, addition of ferric chloride (and/ or alum), and polyelectrolyte to precipitate phosphorus and improve settling prior to discharge into the St. Lawrence River. During the investigation, the dosage of ferric chloride ranged from 4 to 6 mg/l. Undisinfected, settled effluent samples were transferred in a cooler to the Civil Engineering Environmental Laboratory at McGill University and stored at 4 C until testing. The following analyses were performed according to Standard Methods (APHA et al. 1998): total iron (method 3111B), chemical oxygen demand (COD; 5220D), turbidity (2130B), suspended solids (TSS; 2540D), and ph. UV absorbance at 254 nm (UVA 254 ) was measured by an Ultrospec 3300-pro UV spectrophotometer (General Electric Bio- Sciences, Piscataway, N.J., U.S.A.) using a 1-cm cuvette. UVA 254, ph, and turbidity measurements were carried out within 24 hours of receiving the sample; the other analyses were conducted within 48 hours. PAA Analysis and Dosages Used Peracetic acid is the peroxide of acetic acid, produced by reaction of acetic acid with hydrogen peroxide in the presence of 1% sulphuric acid as a catalyst, yielding the following quaternary equilibrium mixture: A commercial solution containing approximately 12% PAA and 18% H 2 O 2 (Solvay-Interox Inc., Houston, Tex. U.S.A.) was used. The solution was diluted 1:250 daily, and the actual PAA concentration was obtained by titration using the CHEMetrics V-2000 Peracetic Acid Vacu-Vials kit (Calverton, Va., U.S.A.). Up to 8 ml of the stock PAA solution was added to samples according to the required dosages. As the CHEMetrics kit cannot be used for wastewater samples, analysis of PAA in the MWTP effluent (10-ml sample) was carried out by reaction with N,N-diethyl-pphenylenediamine catalyzed by KI (postassium hydride), according to the procedure described by Falsanisi et al. (2006). The reaction produces a red colour, and sample absorbance was measured at 530 nm using the abovementioned spectrophotometer. PAA concentration (in milligrams per liter) is proportional to sample absorbance (1) at 530 nm (Abs 530 ) with a typical calibration curve determined according to the Beer-Lambert law as: The mean standard deviation, calculated on seven values with each absorbance measurement replicated six times, was ±0.010 absorbance units and, from equation 2, the estimated mean experimental error on the PAA measurement was ±0.051 mg/l. A first set of batch tests was carried out to investigate PAA consumption kinetics and fecal coliform inactivation of the MWTP effluent. To that aim, a suitable amount of PAA stock solution was added to yield 2, 4, or 8 mg/l of PAA in three 500-ml beakers filled with 400 ml of MWTP effluent. Each beaker was maintained under gentle magnetic stirring to avoid sedimentation. Residual PAA concentrations as well as fecal coliform counts were measured after 1, 5, 10, 20, and 40 minute contact times. A second set of batch tests was carried out to investigate bacteria protection afforded by TSS. To that aim, MWTP effluent was sequentially filtered (see later), then dosed with the PAA stock solution to yield 2 and 8 mg of PAA per litre respectively. Residual PAA and fecal coliforms in the filtered effluents were measured after 1, 5, 10, 20, and 40 minutes contact time. PAA and H 2 O 2 residues were quenched in the samples taken for fecal coliform analysis; up to 0.21 ml of 1 N sodium thiosulfate (97% Na 2 S 2 O 3, Merck 106512) solution and 0.5 ml of a 1000 mg/l catalase (Sigma Chemical Co., St. Louis, Mo., U.S.A.) stock solution were added to 10-ml samples. The final concentrations achieved in these samples were then slightly less than 500 mg/l of sodium thiosulfate and 50 mg/l of catalase. Samples were then immediately stored at 4 C for subsequent fecal coliforms analysis. (2) 48

Effect of Particles on Peracetic Acid Disinfection Fecal Coliforms Fecal coliforms were enumerated according to the Standard Methods (APHA et al. 1998) membrane filtration procedure (Method 9222D) which was carried out in triplicate for each plate. The volume of samples taken from batches to perform this analysis ranged from 10 to 30 ml. Filtration In order to differentiate the various TSS fractions, MWTP effluent was submitted to sequential filtration through membrane filters with different average pore sizes, then dosed with 2 or 8 mg of PAA per litre (the intermediate 4 mg of PAA per litre dosage was omitted in these tests for the sake of simplicity). To that aim, two 300-mL aliquots taken from a 2-L beaker filled with 1.5 L of MWTP effluent, kept under gentle magnetic stirring, were filtered through two nylon membrane filters (Millipore, Billerica, Mass. U.S.A.) of 120- or 10- μm nominal pore size, yielding the so-called medium and highly filtered effluents respectively. Filtration occurred under low-vacuum conditions (approximately 100 mm Hg) to minimize particle break-up and fecal coliform release (which, however, cannot be completely avoided). The filters were changed frequently to prevent pore occlusion/restriction. All glassware and equipment were sterilized to avoid sample contamination. Particle Size Distribution The average TSS cumulative size distribution in the MWTP effluent samples before and after each filtration step was determined using a Lasentec PAR-TEC 100 Particle Size Analyzer (Mettler Toledo Inc., Columbus, Ohio, U.S.A.). Statistical Analysis and Modelling In previous studies on Montreal s effluent, Gehr et al. (2002) and Wagner et al. (2002) found that PAA consumption followed second-order kinetics. Based on this work, Falsanisi et al. (2006) included the effect of initial PAA demand, yielding the following equation: where PAA(t) is the residual PAA concentration (mg/l) after contact time t (min), PAA 0 is the PAA dose (mg/l), k 2 is a second-order rate coefficient [(mg L -1 min -1 ) 1/2 ] and D is the initial PAA demand (mg/l). Wagner et al. (2002) found that k 2 decreases with the disinfectant dose of PAA 0 according to: (3) where a and b are empirical coefficients. Fitting experimental results through model equations was carried out with the aid of Graph Pad Prism ver. 4.02 statistical software (Graph Pad Software, Inc., San Diego, Calif., U.S.A.) which provides best fit values and statistical indexes helpful to assess model suitability (i.e., standard error and 95% confidence intervals), together with goodness of fit coefficients (i.e., R 2, absolute sum of squares, and standard deviations of the vertical distances). Compared with the Chick (1908) and Chick- Watson (Watson 1908) models, the Hom (1972) model was successfully used in a recent investigation to model microbial inactivation occurring after dosing PAA in secondary and tertiary municipal effluents characterized by tailing behaviour (Falsanisi et al. 2006; Santoro et al. 2007). However, the Hom model reproduces with difficulty experimental data such as those found in the present investigation characterized by both tailing and shoulder behaviour (see Results and Discussion section for more details). Accordingly, the Hom model was abandoned in this case and the Selleck model (Selleck et al. 1978), reported to successfully reproduce shoulder and tailing phenomena (Metcalf and Eddy 2003), was considered. Indeed, in the presence of a finite TSS concentration, such as in the present situation, the fecal coliform inactivation rate has been described satisfactorily by the Selleck model, which considers the initial density of microorganisms, N 0 (Lee and Nam 2002). Combined with equation 3, the Selleck equation becomes: where n and h are empirical coefficients which vary with PAA 0. Results and Discussion Wastewater Quality and Particle Size Characterization As can be seen in Table 2, effluent characteristics from the MWTP are quite variable, particularly TSS and UVA 254 (usually associated with PAB), due to varying the FeCl 3 dosage during the investigation, as shown by residual Fe concentrations. As found in previous studies, high initial concentrations of fecal coliforms and TSS in MWTP effluent significantly influence the final disinfection effectiveness achieved (Gehr et al. 2002; Wagner et al. 2002; Gehr et al. 2003). Figure 1 illustrates the average TSS cumulative size distributions for the unfiltered and filtered samples on a typical experimentation day. Membrane filters of 10- and (4) (5) 49

Falsanisi et al. 120-μm nominal pore size appear capable of retaining approximately 70 and 10% of the TSS respectively. These are the highly and medium filtered effluents. Although filtration removes larger particles according to the size of the filter used, the removal efficacy was variable. It is well known that membrane filters do not have precise cut-off points defined by their nominal porosity. As shown in Table 3, indeed, up to 7% of particles >120 μm and up to 32% of particles >10 μm were found in the MWTP medium and highly filtered effluents respectively. Imperfect wastewater filtration through nylon filters was also reported by Qualls et al. (1985) and by Bohrerova and Linden (2006), who found that larger particles could pass 10- and 20-μm filters respectively, whilst Berman et al. (1988) found that approximately 30% of particles >7μm were still present after filtering primary sewage effluent through a 7-μm filter. A similar situation occurred with unfiltered, medium, and highly filtered samples as shown in Fig. 3. Apart from confirming second-order kinetics, results in Fig. 3 indicate that under the conditions investigated, PAA consumption was almost unaffected by filtration, i.e., by the variation of concentration and average size of TSS in the sequentially filtered MWTP effluent. These results allow one to speculate on the complex role played by the overall wastewater characteristics (i.e., physical, chemical, and microbial) on PAA consumption. Indeed, except in some cases, chemical disinfection (with chlorine, chlorine dioxide, ozone, etc.) is generally reported to follow first-order kinetics with the rate Disinfectant Consumption Experimental results of PAA consumption with contact time during batch disinfection of unfiltered MWTP effluent are shown in Fig. 2. The second-order model kinetic reproduces PAA consumption reasonably well according to equation 3, with best fitting values of the constants reported in Fig. 2. The very low values of rate coefficient k 2, which decreases with initial PAA concentration according to equation 4 (see Table 4), account for the quasi-asymptotic decrease of PAA over time. Fig. 1. TSS cumulative size distribution curves for filtered and unfiltered effluents. Fig 2. PAA consumption during disinfection of unfiltered effluent. Best fit values for equation 3: a = 0.17; b = 2.14; D = 0.65. 50

Effect of Particles on Peracetic Acid Disinfection controlling step being the release of the active chemical species (hydroxyl radicals, elemental oxygen, etc.) from the disinfectant (Haas and Karra 1984; Hoigne and Bader 1994; Mitrakas et al. 2007). This was also confirmed for PAA in recent investigations on disinfection of Italian primary and secondary municipal wastewaters (Antonelli et al. 2006; Falsanisi et al. 2006; Rossi et al. 2007). In the present study on Montreal s advanced primary effluent, however, a first-order kinetics model failed to describe the experimental data in Fig. 2 and 3; they were better described by second-order kinetics in agreement with Gehr et al. (2002) and Wagner et al. (2002) in studies with this same effluent. A major difference among wastewater characteristics in the investigations referred to above is the excess Fe 3+ ion resulting from the physicochemical treatment in Montreal s case. As shown by equation 1, the disinfectant solution is a quaternary equilibrium mixture and the Fe 3+ ion is known to affect PAA stability, and this probably influenced PAA decomposition as well as the overall chemical equilibrium of the quaternary mixture (Bianchini et al. 2002; Kitis 2004). Further experiments are planned to better assess the influence of Fe 3+ as well as other wastewater characteristics (COD, TSS, etc.) on PAA reaction kinetics. Microbial Inactivation Figure 4 reports the microbial inactivation results obtained in the first series of batch tests. Shoulder behaviour is evident at the lowest dose of 2 mg of PAA 0 per litre, and tailing behaviour occurs at all dosages used. Figure 4 shows solid lines representing the Selleck model according to equation 5 and reports the best fitting values. It can be seen that equation 5 may acceptably describe tailing, but it fails to predict shoulder behaviour, particularly relevant at the lowest PAA dosage in the current situation. Since tailing and shoulder phenomena suggest that disinfection is influenced by PAB, the results of the second series tests of batch tests, carried out on sequentially filtered effluents, must be considered. In addition to particle size distribution, COD, UVA 254, and initial fecal coliform concentrations were measured in unfiltered, medium, and highly filtered samples. COD and UVA 254 values slightly decreased with filtration (92, 88, 83 mg of COD per litre and 0.736, 0.726, 0.715 UVA 254 in unfiltered, medium, and highly filtered samples, respectively). No significant variation of fecal coliform concentration was noticed among unfiltered, medium, and highly filtered samples as shown in Fig. 5. One could expect a significant variation of fecal coliforms with filtration as found in a similar investigation that considered Mycobacterium as the target microorganism (Bohrerova and Linden 2006). However, it should be considered that approximately 99% of the overall fecal coliform population in wastewater is free-swimming (i.e., <10-μm size) and only approximately 1% or less is PAB (Qualls et al. 1983, 1985). Furthermore, since the traditional microbial detection technique based on membrane filtration (such as the 9222D Method in Standard Methods [APHA et al. 1998] used in this investigation) is more sensitive to free-swimming fecal coliforms easily passing through the filter sizes investigated (Parker and Darby 1995; Madge Fig. 3. PAA consumption during disinfection of sequentially filtered effluent. Fig. 4. Fecal coliform inactivation during PAA disinfection of unfiltered effluent (N 0 3,200,000 CFU/100 ml). 51

Falsanisi et al. and Jensen 2006), one should not expect a significant variation of their overall concentration. However, even the presence of 1% PAB (i.e., 10 3 to 10 4 CFU per 100 ml in the present situation) may prevent achieving the very strict microbial standards usually required for municipal wastewater reuse (10 1 to 10 2 CFU per 100 ml). In contrast with PAA consumption, the microbial inactivation rate was strongly affected by filtration, as indicated in Fig. 6 where the interpolation curves were obtained using Selleck s model (equation 5), with the best fitting values reported in Table 5. The following considerations further emerge from the experimental results reported in Fig. 6. It can be observed that the final (i.e., after 40 minutes contact time) -log N/N 0 value depends on the initial fecal coliform concentration in the MWTP effluent, N 0, which varied widely in the present investigation (see Table 2). As shown in Fig. 4 and 6, in fact, the final -log N/N 0 value for unfiltered MWTP effluents disinfected with similar PAA dosages decreased from approximately 5.5 to approximately 5.0 (at the 8 mg per litre PAA dosage) and from approximately 3.2 to approximately 2.3 (at Fig. 5. Fecal coliform concentration in filtered and unfiltered effluents. the 2 mg per litre PAA dosage) when N 0 decreased from approximately 3.2 10 6 to approximately 1.25 10 6 CFU per 100 ml, respectively. Accordingly, this variation should be attributed to the variation of initial microbial concentration in the samples, N 0, which is duly accounted for by the Selleck model. Furthermore, it is interesting to assess whether disinfection effectiveness varies with the size of particles removed in medium and highly filtered samples that, as discussed above, were characterized by initial fecal coliform concentrations similar to unfiltered samples. Indeed, PAB could associate in a different fashion with different sizes of particles; the results of sequential filtration allow one to assess semiquantitatively the protection to bacteria afforded by these different sizes. Comparing -log N/N 0 for sequentially filtered effluents with the corresponding unfiltered effluent in Fig. 6, it may be argued that the protection afforded by >120-μm TSS (medium filtered sample) amounted to approximately 1.3 log at both PAA dosages investigated. The additional protection afforded by TSS between 10 and 120 μm (the highly filtered sample) amounted to a further 0.6 log. Accordingly, the overall protection offered by TSS particles to fecal coliforms in the present situation may be estimated to amount cumulatively to approximately 1.9 logs at the PAA dosages investigated. The hypothesis that typical shoulder and/or tailing disinfecting behaviour should be related to PAB, however, has not been completely ascertained since filtration inefficiency obviously impairs disinfection effectiveness results. As shown in Table 3, indeed, the total wastewater particle removal through the 10-μm-size filter was not achieved. Accordingly, the sequential filtration of MWTP effluent did not eliminate shoulder or tailing behaviour in the present situation. From a practical point of view, however, by reference to the typical 40 minutes contact time in common chicane disinfection reactors, it may be concluded that (partial) removal of PAB by filtration will yield fecal coliform inactivation up to 3.8 or 7.3 logs at 2 or 8 mg of PAA per litre dosages, respectively, thus attaining the strict microbial standards required for unrestricted municipal wastewater reuse in agriculture (U.S. EPA 1999). Fig. 6. Fecal coliforms inactivation with PAA of sequentially filtered effluent (N 0 1,250,000 CFU/100 ml). 52

Effect of Particles on Peracetic Acid Disinfection Conclusions The results of this study on PAA disinfection of an advanced primary effluent coagulated with ferric chloride offered evidence of the protection afforded by TSS to the microbial indicator considered (fecal coliforms), amounting to close to 2 logs under the conditions investigated. Evidence was further obtained that the size of suspended particles plays a key role in this disinfection, as the increasing protection afforded by TSS of greater than 120 μm, and between 10 μm and 120 μm was approximately 1.3 logs and 0.6 logs, respectively. The relatively low PAA dosage (8 mg/l) and 40 minutes contact time resulted in a very low fecal coliforms concentration (<10 CFU per 100 ml) in the so-called highly filtered (i.e., 10 μm) effluent. Accordingly, a properly designed filtration unit before disinfection appears mandatory to achieve the strict agriculture wastewater reuse standards. As for reaction rate modelling, PAA consumption has been shown to follow modified second-order kinetics, instead of the expected first-order kinetics, probably due to excess Fe 3+ ion concentration in the effluent as a residue of physicochemical treatment. Microbial inactivation kinetics, affected by both tailing and shoulder behaviour, followed only in part the Selleck kinetic model. Further investigation aimed at developing more sophisticated models capable to account in greater detail for the role of concentration and size of TSS and related particle-associated bacteria is planned. Acknowledgments Useful theoretical discussions with Domenico Santoro, research engineer at Trojan Technologies, (Ontario, Canada) as well as analytical assistance by Daina Brumelis, Francois Martin, Wonjae Chan, and Dongxin Chen (McGill) and by staff at the Montreal Wastewater Treatment Plant are sincerely acknowledged. References Antonelli M, Rossi S, Mezzanotte V, Nurizzo C. 2006. Secondary effluent disinfection: PAA long term efficiency. Environ. Sci. Tech. 40:4771-4775. APHA, AWWA, WEF. 1998. Standard methods for the examination of water and wastewater. 20 th Edition. American Public Health Association, Washington D.C. Berman D, Rice EW, Hoff JC. 1988. Inactivation of particle associated bacteria by chlorine and chloramines. Appl. Environ. Microbiol. 54(2):507-512. Bianchini R, Calducci L, Caretti C, Lubello C, Pinzino C, Piscicelli M. 2002. An EPR study on wastewater disinfection by peracetic acid, hydrogen peroxide and UV irradiation. Anal. Chem. 92:783-793. Bohrerova Z, Linden KG. 2006. Ultraviolet and chlorine disinfection of Mycobacterium in wastewater: effect of aggregation. Water Environ. Res. 78:565-571. Carrasco L, Turner CD. 2006. Evaluation of disinfection techniques in the treatment of advanced primary treated wastewater for Ciudad Juarez, Mexico. Water Environ. Res. 78(1):49-58. Chick H. 1908. An investigation of the laws of disinfection. J. Hygiene 8:92-158. De Beer D, Srinivasan R, Stewart PS. 1994. Direct measurement of chlorine penetration into biofilms during disinfection. Appl. Environ. Microbiol. 60:4339-4344. Dell Erba A, Falsanisi D, Liberti L, Notarnicola M, Santoro D. 2004. Disinfecting behaviour of peracetic acid for municipal wastewater reuse. Desalination 168:435-442. Dietrich JP, Basagaoglu H, Loge FJ, Ginn TR. 2003. Preliminary assessment of transport process influencing the penetration of chlorine into wastewater particles and the subsequent inactivation of particle-associated organisms. Water Res. 37:139-149. Emerick RW, Loge FJ, Ginn T, Darby JL. 2000. Modelling the inactivation of particle associated coliform bacteria. Water Environ. Res. 72(4):432-438. Emerick RW, Loge FJ, Thompson D, Darby JL. 1999. Factors influencing ultraviolet disinfection performance part II. Association of coliform bacteria with wastewater particles. Water Environ. Res. 71(6):1178-1197. Falsanisi D, Santoro D, Dell Erba A, Notarnicola M, Liberti L. 2006. Kinetics of wastewater PAA demand exertion and its implications on microbial inactivation. Water Qual. Res. J. Can. 41(4):398-409. Gehr R, Cochrane D, French M. 2002. Peracetic acid as a disinfectant for municipal wastewaters: encouraging performance results from physicochemical as well as biological effluents. In Disinfection 2002. Health and Safety Achieved through Disinfection. Proceedings of the WEF Disinfection Specialty Conference, 17-20 February. St Petersburg, FL. Gehr R, Wagner M, Veerasubramanian P, Payment P. 2003. Disinfection efficiency of peracetic acid, UV and ozone after enhanced primary treatment of municipal wastewater. Water Res. (37):4573-4586. Haas CN, Karra SB. 1984. Kinetics of wastewater chlorine demand exertion. J. Water Pollut. Control Fed. 56:170-173. Hoigne J, Bader H. 1994. Characterization of water quality criteria for ozonation processes. Part II: lifetime of added ozone. Ozone: Sci. Eng. 16(2):121-134. Hom LW. 1972. Kinetics of chlorine disinfection in an ecosystem. J. Sanit. Eng. Div. Proc. ASCE 98(SA1):183-194. Jolis D, Lam C, Pitt P. 2001. Particle effects on ultraviolet disinfection of coliform bacteria in recycled water. Water Environ. Res. 73(2):233-236. 53

Falsanisi et al. Kitis M. 2004. Disinfection of wastewater with peracetic acid: a review. Environ. Int. 30:47-55. Koivunen J, Heinonen-Tanski H. 2005. Peracetic acid (PAA) disinfection of primary, secondary and tertiary treated municipal wastewaters. Water Res. 39:4445-4453. Lee Y, Nam S. 2002. Reflection on kinetic models to the chlorine disinfection for drinking water production. J. Microbiol. 40(2):119-124. Liberti L, Lopez A, Notarnicola M. 1999. Advanced treatment and disinfection for municipal wastewater reuse in agriculture. Water Sci. Technol. 40:235-245. Madge BA, Jensen JN. 2006. Ultraviolet disinfection of fecal coliform in municipal wastewater: effect of particle size. Water Environ. Res. 78(3):294-304. Mandra V, Lazarova V, Dumoutier N, Audic JM. 1996. Étude comparative de la désinfection des eaux résiduaires urbaines par l acide peracetique, l irradiation UV et l ozone. J. du traitement des eaux et des nuisances de Poitiers 61:1-12. Metcalf L, Eddy H. 2003. Wastewater engineering: treatment and reuse. Metcalf & Eddy Inc. 4th Edition. McGraw-Hill Inc., New York. Mitrakas M, Patsos A, Sikalides C, Tzimou-Tsitoundou R. 2007. Effect of temperature on CT value and bromate formation during ozonation of bottled water. Fresnius Environ. Bull., 17(3):341 346. Parker JA, Darby JL. 1995. Particle associated coliforms in secondary effluents: shielding from ultraviolet disinfection. J. Water Pollut. Control Fed. 55(10):1006-1011. Qualls RG, Flynn MP, Johnson JD. 1983. The role of suspended particles in ultraviolet disinfection. J. Water Pollut. Control Fed. 55(10):1280-1285. Qualls RG, Ossoff SF, Chang JCH, Dorfman MH, Dumais CM, Lobe DC, Johnson JD. 1985. Factors controlling sensitivity in ultraviolet disinfection of secondary effluents. J. Water Pollut. Control Fed. 57(10):1006-1011. Rossi S, Antonelli M, Mezzanotte V, Nurizzo C. 2007. Peracetic acid disinfection: a feasible alternative to wastewater chlorination. Water Environ. Res. 79(4):341-350. Santoro D, Gehr R, Bartrand TA, Liberti L, Notarnicola M, Dell Erba A, Falsanisi D, Haas C. 2007. Wastewater disinfection by PAA: assessment of models for tracking residual measurements and inactivation. Water Environ. Res. 79(7):775-787. Selleck RE, Saunier BM, Collins HF. 1978. Kinetics of bacterial deactivation with chlorine. J. Environ. Eng. Div. ASCE 104(6):1197-1212. Stewart MH, Wolfe RL, Means EG. 1990. Assessment of the bacteriological activity associated with granular activated carbon treatment of drinking water. Appl. Environ. Microbiol. 56(12): 3822-3829. U.S. EPA. 1999. Disinfection profiling and benchmarking guidance manual. EPA/815/R/99/013, U. S. Environmental Protection Agency, Athens, Ga. Wagner M, Brumelis D, Gehr R. 2002. Disinfection of wastewater by hydrogen peroxide or peracetic acid: development of procedures for measurement of residual disinfectant and application to a physicochemically treated municipal effluent. Water Environ. Res. 74:33-50. Watson HE. 1908. A note on the variation of the rate of disinfection with change in the concentration of the disinfectant. J. Hygiene 8:536-542. Received: 10 August 2007; accepted 1 February 2008 54