In this study water samples of different origins (subalpine lake, artificial lake and river) were treated by pre-oxidation, coagulation/flocculation, adsorption on granular activated carbon and disinfection. Different laboratory-scale tests were carried out to evaluate the treatment impact on ClO2 consumption in disinfection and on the formation of disinfection by-products (trihalomethanes, adsorbable organic halogen, chlorite and chlorate). The results showed that coagulation/flocculation and activated carbon adsorption have the most significant impact on reducing disinfectant consumption. Pre-oxidation of artificial lake water with KMnO4 and NaClO determines the highest ClO2 consumption. Regardless of the water source, the amount of chlorite produced after disinfection with ClO2 is 40–60% lower using NaClO as the pre-oxidant rather than KMnO4 or ClO2. Otherwise, NaClO leads to a high formation of adsorbable organic halogens and trihalomethanes in artificial lake water (up to 60 μg/L and 20 μg/L respectively), while in the case of ClO2 oxidation, trihalomethane formation is 98% less compared to NaClO. Further, adding ferrous ion in coagulation/flocculation improves the removal of chlorite produced during pre-oxidation, with a 90% removal, mainly due to the reduction of chlorite to chloride. Finally, activated carbon adsorption after pre-oxidation and coagulation/flocculation removes adsorbable organic halogens and trihalomethanes respectively by 50–60% and 30–98%, and completes the chlorite and chlorate removal.

INTRODUCTION

Chlorine is the most common disinfectant used in drinking water treatment. However, during the disinfection treatment, chlorine can react with naturally occurring organic material to produce several by-products, such as total trihalomethanes (TTHMs). Chlorine dioxide (ClO2) is considered one alternative to chlorine for TTHM control since it implies a lower TTHM formation (Sorlini & Collivignarelli 2005a; Zhao et al. 2006; Hua & Reckhow 2007; Badawy et al. 2012; Yang et al. 2013a,  2013b). However, ClO2 can generate, through secondary reactions, both organic and inorganic disinfection by-products (DBPs), in particular chlorite (ClO2) and chlorate (ClO3), which may lead to hemolytic anemia in humans at low levels of exposure, and to oxidative stress resulting in changes in the red blood human cells at higher levels of exposure (Korn et al. 2002; WHO 2011). According to the Italian regulations, the TTHM and the ClO2 maximum allowable concentrations are 30 μg/L and 700 μg/L, respectively (Legislative decree 31/2001), while for ClO3 no limit is fixed; however, the World Health Organization (WHO) suggests a guideline value both for ClO2 and ClO3 of 700 μg/L (WHO 2011).

Several studies investigated different processes for ClO2 removal, such as the addition of reduced-sulphur compounds like sulphur dioxide and sodium sulphite (Gonce & Voudrias 1994), the addition of ferrous chloride and ferrous sulphate (Henderson et al. 2001; Katz & Narkis 2001; Sorlini & Collivignarelli 2005b; Shao-xiu et al. 2012) or the use of powdered (PAC) or granular (GAC) activated carbon (Gonce & Voudrias 1994; Sorlini & Collivignarelli 2005c).

While the control of ClO2 and ClO3 produced in pre-oxidation can be ensured by applying these processes, the control of the DBP formation in the final disinfection stage is more difficult since no additional treatments are generally applied after final disinfection. For this reason, it is necessary to optimize the type and operation of the processes applied in the treatment plant in order to improve the water quality before final disinfection. This allows the amount of disinfectant required to be reduced, and, consequently, to minimize the DBP formation in the final disinfection step and in the distribution system. The disinfectant dose required for final disinfection depends on both the chemical (organic and inorganic compounds) and microbiological pollutants present in the water, and on the disinfectant residual, which should be guaranteed in the distribution system. Increased dissolved organics represent a larger quantity of DBP precursors when disinfectant is applied (USEPA 1999). Therefore, enhancing water treatment performance before final disinfection represents an interesting option for minimizing DBP formation. Coagulation, especially enhanced or optimized coagulation, is considered an interesting treatment option to remove DBP precursors, such as natural organic matter (NOM) (Matilainen et al. 2010); for example, coagulation/flocculation can remove 29–70% of dissolved organic carbon (DOC) (Matilainen et al. 2002; Kaleta & Elektorowicz 2009). Some researchers report that conventional coagulation followed by chemical oxidation-biodegradation processes is a reasonable alternative to enhance coagulation for the removal of total organic carbon (TOC) and dissolved organic halogenide (DOX) precursors (Speitel et al. 2000).

GAC is effective in adsorbing a wide range of organic compounds, removing taste and odours, controlling the formation of chlorinated pollutants, and reducing bromate; therefore, it has been widely used in drinking water treatment (Zhang et al. 2011). In particular, GAC is effective in NOM removal (Swietlik et al. 2002; Cheng et al. 2005; Velten et al. 2011); some researchers report a NOM adsorption yield on microporous carbons close to 70% (Newcombe et al. 2002), a trihalomethane formation potential (THMFP) removal close to 85% (Iriarte-Velasco et al. 2008) and an AOC (assimilable organic carbon) and TOC removal, respectively, close to 66 and 30% (Matilainen et al. 2010). Other studies show that GAC absorption can remove 70–90% of organic carbon (Hu et al. 1999; Polanska et al. 2005; Zhao et al. 2009) and NOM (Uyak et al. 2007). BAC (biological activated carbon) filtration can remove biodegradable organic matter, reduce chlorine dosage, remove bad-smelling substances or other pollutants (TOC, DOC, UV254 absorbance and AOC) and improve drinking water taste (Huang et al. 2004; Zhang et al. 2011; Lou et al. 2012; Trang et al. 2014); moreover, the combined use of O3-BAC can remove up to 90% of AOC (Hu et al. 1999; Chen et al. 2007; Lou et al. 2009,2012). Adsorption of dissolved organic matter (DOM) onto activated carbon depends on the pore size and surface chemistry of the adsorbent and on the initial concentration, type and molecular size distribution of the DOM (Newcombe 1999; Swietlik et al. 2002; Cheng et al. 2005; Schreiber et al. 2005).

Other researchers suggest biofiltration following intermediate ozonation as the most effective process for the removal of TTHM and haloacetic acid precursors. In addition, in order to reduce the DBP formation, ozone can be used to remove recalcitrant organic substances, to increase biodegradability and to reduce the taste of natural organic substances (Chaiket et al. 2002; Selcuk et al. 2005, 2007; Chen et al. 2007; Lou et al. 2012).

With regard to the DBP formation, water pre-oxidation can influence the TTHM formation during subsequent final disinfection with chlorine: while UV/Vis pre-oxidation does not have an effect on TTHMs formed by chlorine, pre-oxidation with O3 often leads to a lower TTHM formation with an unaltered chlorine demand, and pre-oxidation with ClO2 reduces both the TTHM formation and the chlorine demand (Gallard & von Gunten 2002). Some researchers reported an 85% reduction in TTHMs (i.e., from 30 to 5 μg L−1) treating lake water samples with ClO2 instead of chlorine (Volk et al. 2002).

Water pre-oxidation can also influence the formation of other DBPs, such as the adsorbable organic halogens (AOX). For example, some researchers, by means of a pilot plant study, found that the ozonation treatment allows a 35% decrease of the AOX concentration (Vahala et al. 1999).

In this study the aim was to evaluate the impact of different treatments applied in a drinking water treatment plant on the ClO2 consumption in final disinfection and on the formation of DBPs (TTHMs, AOX, ClO2 and ClO3) after each treatment. Laboratory tests in batch conditions were carried out on different water sources (subalpine lake, artificial lake and river), treated by pre-oxidation, coagulation/flocculation, GAC adsorption and final disinfection.

MATERIALS AND METHODS

Types of water tested

The following types of water were studied:

  1. subalpine lake water located in the north of Italy (hereinafter called water 1);

  2. river water located in the north of Italy (hereinafter called water 2);

  3. artificial lake water located in the south of Italy (hereinafter called water 3 and 4).

The main physico-chemical and organoleptic characteristics of these waters are shown in Table 1. Water sampling and storage were performed following the national standard methods (APAT/IRSA-CNR 2003a). Water sampling was performed by collecting five samples for each type of water (Table 1 shows the average values), during the same season for 2 weeks, in order to have homogeneous samples. Water samples were stored in a fridge at 4 °C in dark conditions. All the analyses were performed in a certified laboratory.

Table 1

Raw water quality characteristics

Water Subalpine lake water (water 1) River water (water 2) Artificial lake water (water 3) Artificial lake water (water 4) Dir. 98/83/EC 
pH 7.7 7.7 8.0 7.3 6.5–9.5 
UV254 absorbance (cm−10.011 0.057 0.099 0.086 – 
TOC (mg L−11.7 2.4 5.9 4.7 Without abnormal variations 
NH4+ (mg L−1<D.L. 0.3 0.28 0.2 0.5 
Turbidity (NTU) 0.5 4.2 7.1 4.9 Without abnormal variations 
Total coliforms (CFU 100 mL−125 11,000 200 250 
Escherichia coli (CFU 100 mL−1390 200 
Colony count at 22 °C (CFU mL−1290 760 1,700 350 Without abnormal variations 
Colony count at 37 °C (CFU mL−130 1,200 300 65 Without abnormal variations 
Water Subalpine lake water (water 1) River water (water 2) Artificial lake water (water 3) Artificial lake water (water 4) Dir. 98/83/EC 
pH 7.7 7.7 8.0 7.3 6.5–9.5 
UV254 absorbance (cm−10.011 0.057 0.099 0.086 – 
TOC (mg L−11.7 2.4 5.9 4.7 Without abnormal variations 
NH4+ (mg L−1<D.L. 0.3 0.28 0.2 0.5 
Turbidity (NTU) 0.5 4.2 7.1 4.9 Without abnormal variations 
Total coliforms (CFU 100 mL−125 11,000 200 250 
Escherichia coli (CFU 100 mL−1390 200 
Colony count at 22 °C (CFU mL−1290 760 1,700 350 Without abnormal variations 
Colony count at 37 °C (CFU mL−130 1,200 300 65 Without abnormal variations 

D.L. = detection limit.

Both artificial lake waters show higher TOC and turbidity values than the other types of water. Moreover, as expected, the river water shows high concentrations of microbiological contaminants.

Experimental tests

The experimental tests were performed at laboratory scale in batch conditions (Table 2). All the water samples were treated with chemical pre-oxidation followed by final disinfection. During the first series of tests, a subalpine lake water sample was treated by pre-oxidation alternately with KMnO4, NaClO and ClO2, and subsequently disinfected with ClO2. During the second series of tests, a river water sample and two artificial lake water samples, which showed a poorer quality than the subalpine lake water, were treated by pre-oxidation alternately with KMnO4, NaClO and ClO2 and subsequently treated by coagulation/flocculation with aluminum polychloride (PACl) and FeCl2 and finally disinfected with ClO2. During the third series of tests, two artificial lake water samples, that showed a higher TOC content than river water, were treated by pre-oxidation alternately with KMnO4, NaClO and ClO2 followed by coagulation/flocculation with PACl and FeCl2, adsorption on GAC and final disinfection with ClO2. Water samples were stored in a fridge at 4 °C in dark conditions. All tests were performed in a certified laboratory at a temperature ranging from 20 to 25 °C.

Table 2

Operating conditions of the laboratory-scale experimental tests

  Treatment
 
Water Pre-oxidation Coagulation/flocculation GAC adsorption (column test) Disinfection 
Subalpine lake water (water 1) KMnO4 (0.30 mg L−1) or NaClO (0.50 mg L−1) or ClO2 (0.30 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
River water (water 2) KMnO4 (0.50 mg L−1) or NaClO (1.60 mg L−1) or ClO2 (1.40 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
PACl (5.71 mg L−1
ClO2 (1.40 mg L−1PACl (5.71 mg L−1) + FeCl2 (3.0 mg L−1
Artificial lake water (water 3) KMnO4 (0.45 mg L−1) or NaClO (1.50 mg L−1) or ClO2 (0.90 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
PACl (5.71 mg L−1
ClO2 (0.90 mg L−1PACl (5.71 mg L−1) + FeCl2 (2.0 mg L−1
KMnO4 (0.45 mg L−1) or ClO2 (0.90 mg L−1PACl (5.71 mg L−1Q = 30 mL min−1; EBCT = 10 min 
ClO2 (0.90 mg L−1PACl (5.71 mg L−1) + FeCl2 (2.0 mg L−1Q = 30 mL min−1; EBCT = 10 min 
Artificial lake water (water 4) KMnO4 (0.35 mg L−1) or NaClO (0.90 mg L−1) or ClO2 (0.50 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
PACl (5.71 mg L−1
ClO2 (0.50 mg L−1PACl (5.71 mg L−1) + FeCl2 (1.0 mg L−1
KMnO4 (0.35 mg L−1) or NaClO (0.90 mg L−1) or ClO2 (0.50 mg L−1PACl (5.71 mg L−1Q = 30 mL min−1; EBCT = 10 min 
ClO2 (0.50 mg L−1PACl (5.71 mg L−1) + FeCl2 (1.0 mg L−1Q = 30 mL min−1; EBCT = 10 min 
  Treatment
 
Water Pre-oxidation Coagulation/flocculation GAC adsorption (column test) Disinfection 
Subalpine lake water (water 1) KMnO4 (0.30 mg L−1) or NaClO (0.50 mg L−1) or ClO2 (0.30 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
River water (water 2) KMnO4 (0.50 mg L−1) or NaClO (1.60 mg L−1) or ClO2 (1.40 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
PACl (5.71 mg L−1
ClO2 (1.40 mg L−1PACl (5.71 mg L−1) + FeCl2 (3.0 mg L−1
Artificial lake water (water 3) KMnO4 (0.45 mg L−1) or NaClO (1.50 mg L−1) or ClO2 (0.90 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
PACl (5.71 mg L−1
ClO2 (0.90 mg L−1PACl (5.71 mg L−1) + FeCl2 (2.0 mg L−1
KMnO4 (0.45 mg L−1) or ClO2 (0.90 mg L−1PACl (5.71 mg L−1Q = 30 mL min−1; EBCT = 10 min 
ClO2 (0.90 mg L−1PACl (5.71 mg L−1) + FeCl2 (2.0 mg L−1Q = 30 mL min−1; EBCT = 10 min 
Artificial lake water (water 4) KMnO4 (0.35 mg L−1) or NaClO (0.90 mg L−1) or ClO2 (0.50 mg L−1n.a. n.a. ClO2 (1.0 mg L−1
PACl (5.71 mg L−1
ClO2 (0.50 mg L−1PACl (5.71 mg L−1) + FeCl2 (1.0 mg L−1
KMnO4 (0.35 mg L−1) or NaClO (0.90 mg L−1) or ClO2 (0.50 mg L−1PACl (5.71 mg L−1Q = 30 mL min−1; EBCT = 10 min 
ClO2 (0.50 mg L−1PACl (5.71 mg L−1) + FeCl2 (1.0 mg L−1Q = 30 mL min−1; EBCT = 10 min 

n.a. = not applied; EBCT = empty bed contact time; PACl = aluminum polychloride.

Chemical pre-oxidation

Pre-oxidation treatment was performed alternately with KMnO4 (solution of 1.00 g KMnO4 L−1), NaClO (solution of 1.00 g NaClO L−1) and ClO2 (solution of 1.38 g ClO2 L−1), after determining the oxidant demand for a contact time of 1 hour which is generally applied in the pre-oxidation basin in drinking water treatment plants (DWTPs). It can be noticed that the oxidant with the highest demand is NaClO for each water sample (Table 2). Moreover, the maximum demand with all the tested oxidants is registered for river water, which shows the worst quality characteristics in terms of NH4+, total coliforms, Escherichia coli and colony count at 37 °C (Table 1).

The pre-oxidation tests were performed using 2-L water samples in glass jars in dark conditions. Continuous slow mixing at 30 rpm with a magnetic stirring bar ensured a good contact between the reagent and the solution.

Coagulation/flocculation

The coagulation/flocculation process was performed by adding to the water sample an aqueous solution of PACl, containing 18% Al2O3. The optimal dosage of PACl (5.7 mg L−1) was determined by means of the Jar Test, performed according to ASTM D2035-08 (2008). The laboratory-scale coagulation/flocculation tests were performed using 2-L water samples in beakers, following the procedure outlined in Table 3. Other tests were performed on samples pre-oxidized with ClO2 and treated with PACl, to which ferrous ions were added in order to assess the impact of the Fe2+ on the ClO2 originating from the pre-oxidation. In this case, an aqueous solution of FeCl2 at 40% w/w (density = 9.27 g L−1) was employed. Ferrous ion dosage was based on the assumption that 3.31 mg of Fe2+ is required to deplete 1 mg of ClO2 (stoichiometric demand), following the reaction: 4Fe2+ + C1O2 + 10H2O → 4Fe(OH)3(s) + Cl + 8H+. During this process, ferrous ions (Fe2+) are oxidized to Fe3+ in the form of insoluble Fe(OH)3, which can easily be removed by means of sedimentation and/or filtration. During a coagulation/flocculation process a lower reagent dose can be applied due to the coagulation/flocculation effect of ferrous ions (Katz & Narkis 2001; Sorlini & Collivignarelli 2005b).

Table 3

Experimental conditions of the coagulation/flocculation tests

Step 1 - Coagulation Dosage of aqueous solution of PACl (18% Al2O3) and rapid mixing at 120 rpm for a contact time of 1 minute 
Step 2 - Flocculation Slow mixing at 30 rpm for a contact time of 2 hours 
Step 3 - Sedimentation Settling of flocs for 1 hour 
Step 1 - Coagulation Dosage of aqueous solution of PACl (18% Al2O3) and rapid mixing at 120 rpm for a contact time of 1 minute 
Step 2 - Flocculation Slow mixing at 30 rpm for a contact time of 2 hours 
Step 3 - Sedimentation Settling of flocs for 1 hour 

Adsorption with granular activated carbon

After sedimentation, during the coagulation/flocculation tests, the supernatant of each sample was filtered on GAC. The filtration was performed using an up-flow GAC column (Figure 1): a glass column (height = 50 cm, diameter = 3.5 cm) was filled with mesoporous mineral coal-based GAC (NORIT GAC 1240) with a particle size of D50 = 1.12 mm. The flow rate was set to 30 mL min−1 and the EBCT was 10 minutes. The carbon in the GAC column was replaced with virgin carbon after treating each water sample.
Figure 1

Laboratory-scale system employed for the GAC filtration.

Figure 1

Laboratory-scale system employed for the GAC filtration.

Disinfection

A dosage of 1 mg ClO2 L−1 (initial concentration) was added to 1-L dark glass flasks containing the water samples collected at the outlet of the GAC filter. After stirring manually for 1 minute, each 1-L sample was divided into ten 100 mL sealed dark glass jars. The glass jars had different water-disinfectant contact times (0.5, 1, 2, 4, 7, 22, 27, 44, 54 and 72 hours after the addition of ClO2) and were then opened for the analytical determinations. The different contact times were selected in a different way for each test.

PARAMETERS AND ANALYTICAL METHODS

The following parameters were analysed: pH, TOC, DOC, UV254 absorbance, deep-ultraviolet absorbance at 254 nm wavelength (DUV254 absorbance), turbidity, microbiological contaminants (colony count at 22 °C and 37 °C, total coliforms), residual oxidant, DBPs (ClO2, ClO3, TTHMs and AOX).

The following parameters were monitored at the end of the final disinfection, at each contact time: residual ClO2, ClO2 and ClO3.

The instruments and the analytical methods employed are reported in Table 4.

Table 4

Parameters and analytical methods

Parameter Instrument Analytical method 
Turbidity Hach 2100P turbidimeter – 
pH WTW pH340i pH-meter – 
UV254 and DUV254* absorbance Lambda 2 UV/Visible spectrometer Standard Methods (2011)  
TOC and DOC* TOC analyzer Carlo Erba Total Carbon Monitor 480 APAT/IRSA-CNR (2003d)  
Residual free and total chlorine Spectrophotometer Hach DR/2400 DPD method (APAT/IRSA-CNR 2003c)  
Residual ClO2 Spectrophotometer Hach DR/2400 CPR (chlorine phenol red) method (Fletcher & Hemmings 1985)  
Residual Mn Spectrophotometer Hach DR/2400 PAN - [1 - (2-pyridylazo)-2-naphtol] method (Goto et al. 1977)  
ClO2 and ClO3 Ion chromatograph Dionex DX320 APAT/IRSA-CNR (2003b)  
TTHMs Head space gas chromatograph Perkin Elmer 8600 Standard Methods (2010)  
AOX Euro Glass AOX analyzer ISO 9562:2004 (2004)  
Total coliforms and Escherichia coli Membrane filter (medium used for total coliforms: m-Endo Agar Les; medium used for Escherichia coli: mFC Agar) APAT/IRSA-CNR (2003e, 2003f)  
Colony count at 22 °C and 37 °C Seeding in agar-germs (medium used: Tryptone Glucose Extract AgarAPAT/IRSA-CNR (2003e, 2003g)  
Parameter Instrument Analytical method 
Turbidity Hach 2100P turbidimeter – 
pH WTW pH340i pH-meter – 
UV254 and DUV254* absorbance Lambda 2 UV/Visible spectrometer Standard Methods (2011)  
TOC and DOC* TOC analyzer Carlo Erba Total Carbon Monitor 480 APAT/IRSA-CNR (2003d)  
Residual free and total chlorine Spectrophotometer Hach DR/2400 DPD method (APAT/IRSA-CNR 2003c)  
Residual ClO2 Spectrophotometer Hach DR/2400 CPR (chlorine phenol red) method (Fletcher & Hemmings 1985)  
Residual Mn Spectrophotometer Hach DR/2400 PAN - [1 - (2-pyridylazo)-2-naphtol] method (Goto et al. 1977)  
ClO2 and ClO3 Ion chromatograph Dionex DX320 APAT/IRSA-CNR (2003b)  
TTHMs Head space gas chromatograph Perkin Elmer 8600 Standard Methods (2010)  
AOX Euro Glass AOX analyzer ISO 9562:2004 (2004)  
Total coliforms and Escherichia coli Membrane filter (medium used for total coliforms: m-Endo Agar Les; medium used for Escherichia coli: mFC Agar) APAT/IRSA-CNR (2003e, 2003f)  
Colony count at 22 °C and 37 °C Seeding in agar-germs (medium used: Tryptone Glucose Extract AgarAPAT/IRSA-CNR (2003e, 2003g)  

*The parameter was measured after filtering the sample with a 0.45 μm membrane.

RESULTS AND DISCUSSION

Impact of treatments on quality parameters

With regard to the TOC removal evaluated for artificial lake water (water 4), pre-oxidation is quite ineffective regardless of the type of reagent employed (Figure 2) and also the sequence of treatments composed of pre-oxidation followed by coagulation/flocculation shows low TOC removal yields (25–35%). Moreover, the addition of ferrous ion during coagulation/flocculation after pre-oxidation with ClO2 does not increase the TOC removal. A significant TOC removal is obtained only by adsorption on activated carbon, which increases the TOC removal yield by up 70% (40% on average) for the complete treatment sequence. In agreement with this result, other researchers found a 65% TOC removal yield using a bituminous coal-based activated carbon, with a 13 min EBCT in a full-scale filter in a drinking water treatment plant (Gibert et al. 2013).
Figure 2

TOC concentration after different treatments for artificial lake water (water 4) (TOC in raw water = 4.7 mg L−1; ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2; P = pre-oxidation, CF = coagulation/flocculation, GAC = granular activated carbon adsorption, DIS = disinfection).

Figure 2

TOC concentration after different treatments for artificial lake water (water 4) (TOC in raw water = 4.7 mg L−1; ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2; P = pre-oxidation, CF = coagulation/flocculation, GAC = granular activated carbon adsorption, DIS = disinfection).

DOC concentrations are not reported because they are very similar to the TOC values in all tests (on average the DOC ranges from 83 to 100% of the TOC). Moreover, the analysis of the effect of the different treatments on UV254 and DUV254 absorbance shows results comparable with TOC and DOC.

With regard to turbidity, which was measured 1 hour after the oxidation, for artificial lake water (water 4) (Figure 3) pre-oxidation is quite effective with removal yields ranging from 30% for NaClO to 50% for ClO2; this may be due to the precipitation of some compounds 1 hour after the oxidation, by means of incorporation of suspended colloids in the precipitates. Coagulation/flocculation increases the turbidity removal with an additional removal yield of 30% and 50–60% respectively after pre-oxidation with KMnO4 and ClO2/NaClO. A very low removal yield is obtained when ferrous ion is added during coagulation/flocculation with PACl after pre-oxidation with ClO2. GAC adsorption contributes to turbidity removal by about 50–70% and final disinfection by about 10–40%. The highest turbidity removal yields are obtained with a treatment sequence consisting of pre-oxidation with NaClO, coagulation/flocculation and adsorption on GAC, with a total removal yield of about 90%.
Figure 3

Turbidity values after different treatments for artificial lake water (water 4) (turbidity in raw water = 4.9 NTU; ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2).

Figure 3

Turbidity values after different treatments for artificial lake water (water 4) (turbidity in raw water = 4.9 NTU; ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2).

Figure 4 shows the effect of the pre-oxidation treatment of different water sources on the results for colony count at 37 °C, colony count at 22 °C and total coliforms. The highest concentrations of colony count at 37 °C and total coliforms in the raw water are registered for the river water (water 2), while the highest concentration of colony count at 22 °C is registered for the artificial lake water (water 3). Predictably, regardless of the water source, (Kim et al. 2002), the removal of microbiological contaminants during pre-oxidation is higher using ClO2 and NaClO than using KMnO4. In particular, the best results are obtained with ClO2 applied to river water, which enables a reduction of the colony count at 37 °C by 98–100%, the colony count at 22 °C by 92–100% and the total coliforms by 100%. Moreover, chemical pre-oxidation tests performed on river water (water 2) (Figure 5) confirm the effectiveness of chlorine, both as NaClO and as ClO2, in reducing colony counts at 37 °C (by 94% and 96% respectively) compared to the low reduction rates reached with KMnO4 (27%). Reduction of colony counts at 37 °C is significantly improved by coagulation/flocculation after pre-oxidation with KMnO4, with an additional reduction of about 70% due to coagulation (Figure 5). Conversely, the coagulation/flocculation process contributes to the reduction of colony counts by 4% and 6% using respectively NaClO and ClO2 in pre-oxidation, compared to the pre-oxidation process alone. This is due to the fact that KMnO4 has a low oxidation potential implying a negligible microbial inactivation, which is almost completely achieved by the subsequent coagulation/flocculation; in contrast, NaClO and ClO2 are efficient oxidants that completely ensure microbial removal, thus the subsequent coagulation/flocculation does not further improve the result.
Figure 4

Effect of the pre-oxidation of different water sources on the microbiological contaminant concentration: (a) colony count at 37 °C; (b) colony count at 22 °C; (c) total coliforms.

Figure 4

Effect of the pre-oxidation of different water sources on the microbiological contaminant concentration: (a) colony count at 37 °C; (b) colony count at 22 °C; (c) total coliforms.

Figure 5

Total colony count at 37 °C after pre-oxidation (P) and coagulation/flocculation (CF) for river water (water 2).

Figure 5

Total colony count at 37 °C after pre-oxidation (P) and coagulation/flocculation (CF) for river water (water 2).

In agreement with these results, Table 5 shows that for artificial lake water (water 3) ClO2 allows the best removal yield (94%) to be obtained, and the coagulation/flocculation contributes to a further removal of 5% when ClO2 is used in pre-oxidation; moreover, the ferrous ion addition slightly decreases the removal efficiency compared to the coagulation with PACl alone. As expected, activated carbon adsorption, applied to artificial lake water (water 3), slightly reduces the concentration of microbiological parameters in treated water (Table 5).

Table 5

Total colony count at 37 °C (CFU mL−1) after pre-oxidation (P), coagulation/flocculation (CF), GAC adsorption (GAC) and final disinfection (DIS) for artificial lake water (water 3) (total colony count at 37 °C in raw water = 300 CFU mL−1)

  Total colony count at 37 °C (CFU mL−1)
 
 Reagent used in pre-oxidation
 
Treatment KMnO4 NaClO ClO2 ClO2
Pre-oxidation 114 92 19 20 
Pre-oxidation + coagulation/flocculation 
Pre-oxidation + coagulation/flocculation + GAC adsorption 
Pre-oxidation + coagulation/flocculation + GAC adsorption + disinfection 
  Total colony count at 37 °C (CFU mL−1)
 
 Reagent used in pre-oxidation
 
Treatment KMnO4 NaClO ClO2 ClO2
Pre-oxidation 114 92 19 20 
Pre-oxidation + coagulation/flocculation 
Pre-oxidation + coagulation/flocculation + GAC adsorption 
Pre-oxidation + coagulation/flocculation + GAC adsorption + disinfection 

ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2.

Impact of treatments on disinfection by-products

The results represented in Figures 6 and 7 show that the highest concentrations of AOX and TTHMs are observed when artificial lake water (water 4) is pre-oxidized with NaClO, due to the high reactivity of this oxidant with DOM (Vahala et al. 1999; Sorlini & Collivignarelli 2005a; Zhao et al. 2006; Hua & Reckhow 2007; Badawy et al. 2012) to form these compounds. In particular, the results show that for the case of ClO2 oxidation, TTHMs are reduced by 98% compared to NaClO, as agreed by other researchers who have reported an 85% reduction in TTHM formation (i.e., from 30 to 5 μg L−1) when treating lake water samples with ClO2 instead of NaClO (Volk et al. 2002). The application of KMnO4 resulted in low TTHM formation compared to the other oxidants. Coagulation/flocculation has a negligible influence on the removal of AOX and TTHMs, except in the case when NaClO is used in pre-oxidation, as about 20% less AOX can be found in the water after coagulation. The adsorption with activated carbon contributes to the removal by 50–60% (depending on the pre-oxidant used) for AOX, and by 30–98% (depending on the pre-oxidant used) for TTHMs. Moreover, after the GAC treatment, as expected, disinfection does not affect the AOX and TTHM concentrations. Furthermore, the results show that AOX and TTHM concentrations slightly increase after pre-oxidation with KMnO4 or ClO2 followed by coagulation/flocculation and disinfection compared to pre-oxidation alone (Figures 6 and 7).
Figure 6

AOX concentration after different treatments for artificial lake water (water 4) (ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2).

Figure 6

AOX concentration after different treatments for artificial lake water (water 4) (ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2).

Figure 7

TTHM concentration after different treatments for artificial lake water (water 4) (ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2; detection limit of the single THM compound = 0.05 μg/L).

Figure 7

TTHM concentration after different treatments for artificial lake water (water 4) (ClO2* = addition of ferrous ion in coagulation/flocculation after pre-oxidation with ClO2; detection limit of the single THM compound = 0.05 μg/L).

Figure 8 shows the ClO2 formation over time for different water sources pre-oxidized alternately with KMnO4, NaClO and ClO2, and subsequently treated with ClO2 in order to simulate the influence of final disinfection. The results show that for water 1 the highest ClO2 formation rate is in the first 200–800 minutes after the ClO2 addition, while for water 2, 3 and 4 the ClO2 formation rate is high in the first 60 minutes after the ClO2 addition; during this time, the highest ClO2 formation in disinfection is obtained in the case of pre-oxidation with ClO2, which is to be expected as additional ClO2 is generated during the pre-oxidation process (Sorlini et al. 2014). At 60 minutes after the ClO2 addition, pre-oxidation using both KMnO4 and NaClO followed by disinfection with ClO2 determines a slower but still significant ClO2 formation in each water sample; this is due to the reactions of ClO2 with organic and inorganic pollutants which lead to the conversion of about 50–70% of the chlorine consumed to ClO2 (Korn et al. 2002). Overall, the ClO2 formation is 40–60% lower after the addition of ClO2 in the water samples oxidized with NaClO instead of KMnO4 or ClO2. Comparing the different water sources, the ClO2 formation in subalpine lake water (water 1) and in artificial lake water (water 4) is always below the current regulatory limit of 0.7 mg L−1, while in river water (water 2) and in artificial lake water (water 3) it exceeds the limit after 60 minute contact time. Moreover, more ClO2 is formed when KMnO4 is applied as pre-oxidant in the case of subalpine lake water (water 1) and artificial lake water (water 3); this result may be due to the fact that KMnO4 is more reactive with ClO2 precursors, such as NOM, compared to the other oxidants.
Figure 8

Chlorite formation after pre-oxidation alternately with KMnO4, NaClO and ClO2 followed by disinfection with ClO2 of different water sources: (water 1) subalpine lake water; (water 2) river water; (water 3) artificial lake water; (water 4) artificial lake water.

Figure 8

Chlorite formation after pre-oxidation alternately with KMnO4, NaClO and ClO2 followed by disinfection with ClO2 of different water sources: (water 1) subalpine lake water; (water 2) river water; (water 3) artificial lake water; (water 4) artificial lake water.

Figure 9 shows the effect of different treatments on ClO2 and ClO3 concentration in artificial lake water (water 3). The results show that coagulation/flocculation with PACl does not influence ClO2 formation, whereas the addition of ferrous ion after PACl during coagulation/flocculation has a markedly visible influence on the ClO2 depletion, with a removal yield close to 90%. In fact, ferrous ions efficiently decrease chlorite ions concentration by reducing them to chlorides. In agreement with this result, several researchers showed that ferrous ion addition to raw waters pre-disinfected with ClO2 completely removes chlorite ions by reducing them to chloride ions (Griese et al. 1991,  1992; Iatrou & Knocke 1992; Katz & Narkis 2001; Matilainen et al. 2006). The addition of GAC filtration after pre-oxidation, coagulation/flocculation and ferrous ion addition leads to a complete removal of the residual ClO2. However, final disinfection with 1 mg ClO2 L−1 leads to the re-formation of ClO2 up to a concentration of 0.03 mg L−1 after one hour contact time.
Figure 9

Chlorite and chlorate concentrations after pre-oxidation with chlorine dioxide combined with different treatments applied to artificial lake water (water 3) (D.L. = detection limit = 0.01 mg/L).

Figure 9

Chlorite and chlorate concentrations after pre-oxidation with chlorine dioxide combined with different treatments applied to artificial lake water (water 3) (D.L. = detection limit = 0.01 mg/L).

Regarding ClO3, Figure 9 shows that coagulation/flocculation, both with and without addition of ferrous salts, does not influence ClO3 formation. This is confirmed by other researchers who reported an efficient reduction of residual ClO2 and ClO2 to chloride ions by ferrous ions, while the concentration of ClO3 was almost unaffected (Griese et al. 1991,  1992; Iatrou & Knocke 1992; Katz & Narkis 2001; Matilainen et al. 2006). Conversely, the results of this study show that GAC filtration can completely remove ClO3. After final disinfection, further ClO3 is generated (0.02 mg L−1), as a consequence of the ClO2 addition.

Impact of treatments on chlorine dioxide consumption in final disinfection

The ClO2 consumption was calculated as the difference between the ClO2 dosed initial concentration and the residual ClO2 concentration after a certain contact time. The analysis of the ClO2 consumption in final disinfection in artificial lake water (water 4) samples shows that pre-oxidation with KMnO4 and NaClO determines the highest ClO2 consumption rates (Figure 10). Moreover, in the case of pre-oxidation with KMnO4, with respect to NaClO, consumption occurs more rapidly during the first 60 minutes.
Figure 10

Chlorine dioxide consumption after oxidation with ClO2 of artificial lake water (water 4) pre-oxidized alternately with KMnO4, NaClO and ClO2.

Figure 10

Chlorine dioxide consumption after oxidation with ClO2 of artificial lake water (water 4) pre-oxidized alternately with KMnO4, NaClO and ClO2.

The analysis of the ClO2 demand after 60 minutes, evaluated for artificial lake water (water 4) (Figure 11), shows that after pre-oxidation with KMnO4 water consumes the maximum ClO2 dosage, whereas for pre-oxidation with NaClO and ClO2 about 60% and 50% of the disinfectant dose, respectively, is consumed.
Figure 11

Chlorine dioxide consumption at 1 hour after oxidation with ClO2 of artificial lake water (water 4) pre-oxidized alternately with KMnO4, NaClO and ClO2.

Figure 11

Chlorine dioxide consumption at 1 hour after oxidation with ClO2 of artificial lake water (water 4) pre-oxidized alternately with KMnO4, NaClO and ClO2.

The addition of the coagulation/flocculation process significantly reduces the ClO2 consumption, by 75% to 95%, respectively, after pre-oxidation with KMnO4 and ClO2. This result agrees with other studies that proved an effective reduction of water permanganate oxidability (70–80%), DOC (25–67%), UV254 absorbance (44–77%), THMFP (25–66%) and turbidity (97%) by means of coagulation (Kaleta & Elektorowicz 2009; Matilainen et al. 2010).

Adsorption on GAC further reduces the ClO2 consumption by about 80–85% compared to the consumption observed for water treated only with pre-oxidation (Figure 11). This is due to the effective adsorption on GAC of NOM (Matilainen et al. 2006) and other THM precursors (Iriarte-Velasco et al. 2008). In particular, after treatment with GAC the highest ClO2 demand reduction is observed in the case of pre-oxidation with KMnO4, compared to the ClO2 demand for water treated with pre-oxidation followed by coagulation/flocculation. GAC is successfully employed in reducing the oxidant consumption after pre-oxidation with KMnO4, since this oxidant is less effective in the removal of organic substances compared to other oxidants tested (Chen & Yeh 2005).

CONCLUSIONS

The aim of this study was to evaluate the impact of different treatments on the ClO2 consumption and on the formation of DBPs (THMs, AOX, ClO2 and ClO3) in the final disinfection. Moreover, water quality parameters were evaluated, such as pH, ultraviolet absorbance at 254 nm wavelength, turbidity, colour, total organic carbon and microbiological parameters. The experimental tests were performed at laboratory-scale on four different water sources applying chemical pre-oxidation, coagulation/flocculation, adsorption on granular activated carbon and final disinfection.

The results show that the stability of ClO2 added in final disinfection depends on the type of oxidant used during pre-oxidation and on the type of treatments applied before the final disinfection. The effectiveness of coagulation/flocculation and GAC adsorption in reducing the ClO2 consumption in final disinfection is confirmed due to the removal and adsorption of the organic matter. Further, the addition of ferrous ion in coagulation/flocculation leads to the reduction of ClO2 concentration, due to the reduction of chlorite to chloride.

Finally, the results confirm that GAC treatment is also effective in adsorbing other disinfection by-products, such as AOX and TTHMs. Comparing the different water sources, the ClO2 formation in subalpine lake waters is always below the current regulatory limit of 0.7 mg L−1, while in river water and in artificial lake water (water 3) it exceeds the limit after 60 minute contact time.

Therefore, enhancing the treatment processes prior to final disinfection increases the efficiency in contaminant and DBP precursor removal and, consequently, decreases the disinfectant consumption.

ACKNOWLEDGEMENTS

The authors wish to thank the technicians at Caffaro S.p.A., especially Dr Mario Belluati for research collaboration and the water treatment plant operators who provided water samples. We also wish to thank the engineer Andrea Cotelli for performing the tests as part of his 3-year thesis. Sabrina Sorlini coordinated the experimental work, the discussion of results and the drafting of the paper, Francesca Gialdini carried out the experimental test, the data processing and the drafting of the paper; Michela Biasibetti and Maria Cristina Collivignarelli contributed to the drafting of the paper.

REFERENCES

REFERENCES
APAT/IRSA-CNR
2003a
Metodi analitici per le acque 29/2003. Metodo 1030
Metodi di campionamento
.
Roma
.
APAT/IRSA-CNR
2003b
Metodi analitici per le acque 29/2003. Metodo 4020 Determinazione di anioni mediante cromatografia ionica
.
Roma
.
APAT IRSA-CNR
2003c
Metodi analitici per le acque 29/2003. Metodo 4080 loro attivo libero
.
Roma
.
APAT/IRSA-CNR
2003d
Metodi analitici per le acque 29/2003. Metodo 5040 Carbonio organico disciolto
.
Roma
.
APAT/IRSA-CNR
2003e
Metodi analitici per le acque 29/2003. Metodo 7010 Coliformi totali
.
Roma
.
APAT/IRSA-CNR
2003f
Metodi analitici per le acque 29/2003. Metodo 7030 Escherichia coli
.
Roma
.
APAT/IRSA-CNR
2003g
Metodi Analitici per le Acque 29/2003. Metodo 7050 Conteggio delle colonie su agar a 36°C e 22°C
.
Roma
.
ASTM D2035-08
2008
Standard Practice for Coagulation-Flocculation Jar Test of Water
.
Badawy
M. I.
Gad-Allah
T. A.
Ali
M. E. M.
Yoon
Y.
2012
Minimization of the formation of disinfection by-products
.
Chemosphere
89
(
3
),
235
240
.
Chaiket
T.
Singer
P. C.
Miles
A.
Moran
M.
Pallotta
C.
2002
Effectiveness of coagulation, ozonation, and biofiltration in controlling DBPs
.
Journal of American Water Works Association
94
(
12
),
81
87
.
Chen
J. J.
Yeh
H. H.
2005
The mechanisms of potassium permanganate on algae removal
.
Water Research
39
(
18
),
4420
4428
.
Cheng
W.
Dastgheib
S. A.
Karanfil
T.
2005
Adsorption of dissolved natural organic matter by modified activated carbon
.
Water Research
39
(
11
),
2281
2290
.
Griese
M. H.
Hauser
K.
Berkemeier
M.
Gordon
G.
1991
Using reducing agents to eliminate chlorine dioxide and chlorite ion residuals in drinking water
.
Journal American Water Works Association
83
(
5
),
56
61
.
Griese
M. H.
Kaczur
J. J.
Gordon
G.
1992
Combining methods for the reduction of oxychlorine residuals in drinking water
.
Journal American Water Works Association
84
(
11
),
69
77
.
Hu
J. Y.
Wang
Z. S.
Ng
W. J.
Ong
S. L.
1999
The effect of water treatment processes on the biological stability of potable water
.
Water Research
33
(
11
),
2587
2592
.
Huang
W. J.
Peng
H. S.
Peng
M. Y.
Chen
L. Y.
2004
Removal of bromate and assimilable organic carbon from drinking water using granular activated carbon
.
Water Science and Technology
50
(
8
),
73
80
.
Iatrou
A.
Knocke
W. R.
1992
Removing chlorite by the addition of ferrous iron
.
Journal American Water Works Association
84
(
11
),
63
68
.
Iriarte-Velasco
U.
Alvarez-Uriarte
J. I.
Chimeno-Alanìs
N.
Gonzàlez-Velasco
J. R.
2008
Natural organic matter adsorption onto granular activated carbons: implications in the molecular weight and disinfection byproducts formation
.
Industrial & Engineering Chemistry Research
47
(
20
),
7868
7876
.
ISO 9562:2004
2004
Water quality -- determination of adsorbable organically bound halogens (AOX
)
.
Kaleta
J.
Elektorowicz
M.
2009
Removal of humic substances from aqueous solutions by the coagulation process
.
Environmental Technology
30
(
2
),
119
127
.
Kim
B. R.
Anderson
J. E.
Mueller
S. A.
Gaines
W. A.
Kendall
A. M.
2002
Literature review – efficacy of various disinfectants against Legionella in water systems
.
Water Research
36
(
18
),
4433
4444
.
Legislative decree 31
,
2 February
2001
Execution of Directive 98/83/CE Related to Drinking Water Quality
.
Lou
J. C.
Lin
C. Y.
Han
J. Y.
Tseng
W. B.
Hsu
K. L.
Chang
T. W.
2012
Comparing removal of trace organic compounds and assimilable organic carbon (AOC) at advanced and traditional water treatment plants
.
Environmental Monitoring and Assessment
184
(
6
),
3491
3501
.
Matilainen
A.
Lindqvist
N.
Korhonen
S.
Tuhkanen
T.
2002
Removal of NOM in the different stages of the water treatment process
.
Environment International
28
(
6
),
457
465
.
Matilainen
A.
Vieno
N.
Tuhkanen
T.
2006
Efficiency of the activated carbon filtration in the natural organic matter removal
.
Environment International
32
(
3
),
324
331
.
Matilainen
A.
Vepsäläinen
M.
Sillanpää
M.
2010
Natural organic matter removal by coagulation during drinking water treatment: a review
.
Advances in Colloid and Interface Science
159
(
2
),
189
197
.
Polanska
M.
Huysman
K.
Van Keer
C.
2005
Investigation of assimilable organic carbon (AOC) in Flemish drinking water
.
Water Research
39
(
11
),
2259
2266
.
Selcuk
H.
Vitosoglu
Y.
Ozaydin
S.
Bekbolet
M.
2005
Optimization of ozone and coagulation processes for bromate control in Istanbul drinking waters
.
Desalination
176
(
1–3
),
211
217
.
Selcuk
H.
Rizzo
L.
Nikolaou
A. N.
Meric
S.
Belgiorno
V.
Bekbolet
M.
2007
DBPs formation and toxicity monitoring in different origin water treated by ozone and alum/PAC coagulation
.
Desalination
210
(
1–3
),
31
43
.
Shao-xiu
L.
Wen-qin
X.
De-jun
Z.
Xiu-li
Y.
Zhi-hong
W.
Dong-mei
L.
Bao-hong
X.
2012
Removing algaecide byproduct chlorite by ferrous salt
.
Environmental Science and Technology
12
.
Sorlini
S.
Collivignarelli
C.
2005b
Chlorite removal with ferrous ions
.
Desalination
176
(
1–3
),
267
271
.
Sorlini
S.
Collivignarelli
C.
2005c
Chlorite removal with granular activated carbon
.
Desalination
176
(
1–3
),
255
265
.
Speitel
G. E.
Jr
Symons
J. M.
Mialaret
J. M.
Marie
M.
Wanielista
E.
2000
AOP/Biofilm processes for DOX precursors
.
Journal American Water Works Association
92
,
59
67
.
Standard Methods for the Examination of Water, Wastewater
2011
Part 5910/2011 UV-Absorbing Organic Constituents
.
Standard Methods for the Examination of Water, Wastewater
2010
Part 6232/2010 Trihalomethanes and Chlorinated Organic Solvents
.
Swietlik
J.
Raczyk-Stanislawiak
U.
Bilozor
S.
Ilecki
W.
Nawrocki
J.
2002
Adsorption of natural organic matter oxidized with ClO2 on granular activated carbon
.
Water Research
36
(
9
),
2328
2336
.
Trang
V. N.
Dan
N. P.
Phuong
L. D.
Thanh
B. X.
2014
Pilot study on the removal of TOC, THMs, and HAAs in drinking water using ozone/UV–BAC
.
Desalination and Water Treatment
52
(
4–6
),
990
998
.
USEPA
1999
Alternative Disinfectants and Oxidants Guidance Manual
.
Uyak
V.
Yavuz
S.
Toroz
I.
Ozaydin
S.
Genceli
E. A.
2007
Disinfection byproducts precursors removal by enhanced coagulation and PAC adsorption
.
Desalination
216
(
1–3
),
334
344
.
Velten
S.
Knappe
D. R. U.
Traber
J.
Kaiser
H. P.
Von Gunten
U.
Boller
M.
Meylan
S.
2011
Characterization of natural organic matter adsorption in granular activated carbon adsorbers
.
Water Research
45
(
13
),
3951
3959
.
Volk
C. J.
Hofmann
R.
Chauret
C.
Gagnon
G. A.
Ranger
G.
Andrews
R. C.
2002
Implementation of chlorine dioxide disinfection: effects of the treatment change on drinking water quality in a full-scale distribution system
.
Journal of Environmental Engineering and Science
1
(
5
),
323
330
.
WHO
2011
Guidelines for drinking-water quality, 4th edn. WHO Press, World Health Organization, Switzerland
.
Yang
X.
Guo
W.
Zhang
X.
Chen
F.
Ye
T.
Liu
W.
2013b
Formation of disinfection by-products after pre-oxidation with chlorine dioxide or ferrate
.
Water Research
47
(
15
),
5856
5864
.
Zhang
D. Y.
Li
W. G.
Zhang
S. M.
Liu
M.
Zhao
X. Y.
Zhang
X. C.
2011
Bacterial community and function of biological activated carbon filter in drinking water treatment
.
Biomedical and Environmental Sciences
24
(
2
),
122
131
.