Chemical solutions for greywater recycling

Available online at www.sciencedirect.com

Chemosphere 71 (2008) 147–155 www.elsevier.com/locate/chemosphere

Technical Note

Chemical solutions for greywater recycling Marc Pidou a, Lisa Avery a, Tom Stephenson b, Paul Jeffrey a, Simon A. Parsons a, Shuming Liu c, Fayyaz A. Memon c, Bruce Jefferson a,* a

Centre for Water Science, School of Applied Sciences, Cranfield University, Cranfield MK43 0AL, United Kingdom b School of Applied Sciences, Cranfield University, Cranfield MK43 0AL, United Kingdom c School of Engineering, Computer Science and Mathematics, University of Exeter, Exeter EX4 4QF, United Kingdom Received 15 August 2006; received in revised form 24 October 2007; accepted 24 October 2007 Available online 21 December 2007

Abstract Greywater recycling is now accepted as a sustainable solution to the general increase of the fresh water demand, water shortages and for environment protection. However, the majority of the suggested treatments are biological and such technologies can be affected, especially at small scale, by the variability in strength and flow of the greywater and potential shock loading. This investigation presents the study of alternative processes, coagulation and magnetic ion exchange resin, for the treatment of greywater for reuse. The potential of these processes as well as the influence of parameters such as coagulant or resin dose, pH or contact time were investigated for the treatment of two greywaters of low and high organic strengths. The results obtained revealed that magnetic ion exchange resin and coagulation were suitable treatment solutions for low strength greywater sources. However, they were unable to achieve the required level of treatment for the reuse of medium to high strength greywaters. Consequently, these processes could only be considered as an option for greywater recycling in specific conditions that is to say in case of low organic strength greywater or less stringent standards for reuse. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Coagulation; Greywater; Magnetic ion exchange resin; Recycling

1. Introduction Interest in wastewater recycling has been raised by the increase of water demand, water shortage due to low rainfall, economic and environmental issues (Eriksson et al., 2002). Among the different options for water reuse such as industrial, irrigation, and ground water recharge, water recycling within urban environments is the least developed. Urban recycling usually integrates the reuse of black, grey or rain waters. Greywater is defined as domestic wastewater excluding water from the toilet, and generally includes wastewaters from baths, showers, hand basins, washing machines, dishwashers and kitchen sinks. However, at small scale the heavily polluted sources such as washing

*

Corresponding author. Tel.: +44 0 1234 750111; fax: +44 0 1234 75167. E-mail address: b.jefferson@cranfield.ac.uk (B. Jefferson). 0045-6535/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.10.046

machines, dishwashers and kitchen sinks tend to be excluded whereas at larger scale all sources are used to maximize water savings. The most common application for greywater reuse is toilet flushing which can reduce water demand within dwellings by up to 30% (Karpiscak et al., 1990). However, other applications such as irrigation of parks, school yards, cemeteries and golf courses, vehicle washing, fire protection and air conditioning are practiced (Lu and Leung, 2003). The water quality standards for wastewater recycling depend on location and application but generally include parameters based on organic, solids and microbiological contents of the water. The most stringent criteria require a biochemical oxygen demand (BOD) of less than 10 mg l1, a turbidity below 2 NTU and a nondetectable level of either total or faecal coliforms (USEPA, 2004; Tajima, 2005). However, other standards which are less restrictive allow higher concentrations of the different parameters or do not include some of the parameters at all (USEPA, 2004; Gross et al., 2007).

148

M. Pidou et al. / Chemosphere 71 (2008) 147–155

A large range of technologies has been used for greywater recycling from simple 2-stage processes (coarse filtration and disinfection) to physical, physicochemical and biological processes (Jefferson et al., 2000). The latter, widely used in large building (Santala et al., 1998; Surendran and Wheatley, 1998; Nolde, 1999; Friedler et al., 2004) suffer from feed source variability and potential shock loading at smaller scale. Such problems are avoided with simple physical processes such as cartridge filters or depth filtration beds. However, whilst these are effective at removing the physical pollution within the greywater, they do not significantly alter the organic fraction (Jefferson et al., 2000). As such chemical processes such as coagulation and adsorption provide great potential for the removal of the dissolved organic fraction within greywater. Indeed, coagulation with metal salts remains the main process utilised in the potable water treatment field for the removal of high concentrations of dissolved organic carbon (DOC) (Parsons and Jefferson, 2006). In more recent times a novel magnetic ion exchange resin (MIEXÒ) has been trialled and is being used to reduce organic loads onto some water treatment works, to reduce coagulant demand or improve the structural properties of the flocs produced (Jefferson et al., 2004a). In opposition to traditional ion exchange resins, MIEXÒ has a magnetic component in its structure which facilitates agglomeration and settling. Moreover, with an average particle size of 180 lm, 2–5 times smaller than traditional ion exchange resins, MIEXÒ has a high surface area for adsorption. And finally, it is designed to be added to the water as slurry in a mixed reactor. The particles dispersed in the water maximise the contact with the organics reducing the contact time needed compare to a fixed-bed set up (Boyer and Singer, 2005). The aim of the present work is to assess the potential for utilise these chemical processes for greywater recycling. 2. Material and methods 2.1. Sampling Greywater was collected from a purpose built facility which diverts water from the bath, shower and hand basin of 18 flats within a student hall of residence located at Cranfield University. This source water was defined as the mixed source in the subsequent experiments. As an alternative, shower water was also collected from a single location. Care was taken to standardise the products used, with respect to their concentration and the duration of the shower between samples. Samples were taken on the morning that the shower was used with all the analyses performed on the same day.

was then adjusted with a plastic pipette and allowed to settle for another half hour. 1 l of the greywater to be treated was placed on a PB900 jar tester (Phipps and Bird, Virginia, USA) which was set to 150 rpm. The resin was shaken in the measuring cylinder and added to the water sample, and the residual was rinsed into the jar with deionised water. The test was carried out for various contact times varying between 10 and 90 min. At the end of each test, the treated water was filtered through 0.45 lm glass fibre filter papers. For the coagulation test, 1 L of the water to be treated was placed on the jar tester. Two speeds were used, a rapid mix at 200 rpm for 90 s, time during which the coagulant, either ferric sulphate (solution of FeSO4 (13%), Feripol XL, EA West, UK) or aluminium sulphate (solution of Al2(SO4)3, 14H2O (48%), Kemira Chemicals, UK) was dosed in the jar and the pH adjusted to the chosen value (4.5, 6 and 7). The sample was then flocculated for 15 min at 30 rpm and allowed to settle for an additional 15 min. Finally, both tests were coupled with the MIEXÒ resin prepared as explained above and added to the water, at the optimum conditions found under the previous tests. The jar tester was set up at 150 rpm for 10 min, after a settling period the treated water was filtered. The filtrate then underwent the coagulation experiments as described before for a range of concentrations and pH. All tests reported here were duplicated and carried out at room temperature. 2.3. Fractionation The raw water samples were fractionated into their hydrophobic and hydrophilic components with a method adapted from Malcolm and McCarthy (1992). The raw water was first filtered through a 0.45 lm filter and acidified to pH 2 using HCl (1 M). The acidified sample was then put through the XAD-8/XAD-4 column pair. The effluent from both columns contained the hydrophilic non-adsorbed fraction. XAD-8 and XAD-4 columns were eluted with NaOH (0.1 M) and the eluates were the hydrophobic acid fraction and the hydrophilic acid fraction, respectively. The organic content of each fraction was then determined by measuring the DOC with a recovery of 88% on average. For comparison, an anionic, a cationic and a non-ionic surfactants were separated into their hydrophobic and hydrophilic fractions. Synthetic solutions of the anionic surfactant sodium lauryl sulphate (Fisher Scientific, UK), the cationic surfactant cetyl trimethylammonium bromide (Sigma–Aldrich, UK) and the non-ionic surfactant TritonÒ X-100 (Acros Organics, UK) at a concentration of 1 mM were fractionated using the method previously described.

2.2. Experiments 2.4. Zeta potential and charge density MIEXÒ (Orica, Australia) resin was prepared by measuring the required dose in measuring cylinders and allowed to settle for approximately 1 h. If needed, the dose

Zeta potential was measured using a Malvern Zetasizer (Malvern, UK). The charge density of water samples was

M. Pidou et al. / Chemosphere 71 (2008) 147–155

determined by using the zeta potential and polydiallyl dimethyl ammonium chloride (PolyDADMAC) (Sigma– Aldrich, UK). The samples were placed in a 1-l beaker and stirred. A 0.1% solution of PolyDADMAC was dosed into the solution, the pH adjusted to 7 and the zeta potential measured until the point of zero charge or iso-electric point was reached. The charge density of the samples (meq g1 DOC ) was then deduced from the amount of PolyDADMAC (charge density: 6.2 meq g1) used (Sharp et al., 2004). 2.5. Analytical procedures DOC (mg l1) was measured using a total organic carbon analyser Shimadzu TOC-5000A (Shimadzu, Milton Keynes, UK). Turbidity (NTU) was measured with a turbidimeter Hach 2100N. Escherichia coli and total coliforms (MPN 100 ml1) were measured using the method Colilert 18 with quanti-tray 2000 (Idexx, UK) and faecal Enterococci (MPN 100 ml1) using the Enterolert with quantitray 2000 (Idexx, UK). BOD (mg l1) was measured using the procedure 5 day Biochemical Oxygen Demand from The Standard Methods for Examination of Water and Wastewater (APHA, 1992). Merck cell tests (VWR International, UK) were used for the following tests: COD, ammonia, nitrate, phosphate and total nitrogen (according to Korleff’s method). UV absorbance was measured with a 6505 UV/vis. Spectrophotometer (Jenway, UK) at a wavelength of 254 nm. 3. Results and discussion 3.1. Characteristics The two sources of greywater tested in the current investigation varied considerably in terms of their organic concentration (Table 1). For instance, the BOD and COD of the two sources were 39 ± 17 and 144 ± 63 mg l1 for the

149

mixed source and 166 ± 37 and 575 ± 98 mg l1 for the shower water. Published greywater strengths indicate that the organic load exerted by a greywater can vary considerably from one scheme to the next but normally fall within the range 50–300 mg l1 BOD (Jefferson et al., 2004b). Interestingly, the organic strength of the mixed source was the lowest level recorded in the literature for a real greywater source. Water production rates from the facility are roughly at the expected level and so the cause of the low strength is unclear but is probably a combination of product choice by the students and the residence time in the collection system. Comparison of the organic parameters indicates that the COD to BOD ratio is approximately 3.5 in both waters studied in the current investigation. This compares to 2.2 for typical domestic sewage and 3–10 for final effluent suggesting greywater contains more non-biodegradable material than sewage (Metcalf and Eddy, 2003). This confirms the limitation of biological processes to treat greywater and supports the case for investigation of non-biological treatment options. The higher nutrient concentrations and charge in the shower water than in the mixed source were due to the washing products choice. Characterisation of the greywaters in relation to parameters commonly used when describing coagulation reveal the DOC of the two sources to be 12 ± 4 and 56 ± 7 mg l1 for the mixed and shower sources, respectively. The former is equivalent to that of a potable water source with a high concentration of natural organic matter (NOM) (Ratnaweera et al., 1999) whilst the latter represents a very high strength and as such high coagulant doses are expected. The specific UV absorbance (SUVA) was between 2.5 and 3.5 l mg1 m1 for the mixed greywater and between 1.5 and 2.5 l mg1 m1 for the shower water suggesting both greywater sources contain mainly hydrophilic and low molecular weight (MW) compounds (Edzwald and Tobiason, 1999; Karanfil et al., 2002). Fractionation of the water confirmed this with between 60% and 70% of the organic matter being associated with the hydrophilic

Table 1 Water sources characteristics Greywaters BOD (mg l1) COD (mg l1) COD/BOD DOC (mg l1) Turbidity (NTU) TN (mg l1) 1 PO3 4 (mg l ) 1 þ NH4 (mg l ) 1 NO 3 (mg l ) pH Charge density (meq g1 DOC ) SUVA (l mg1 m1) Hydrophobic fraction (%) a

Wastewater

Natural water

Mixed (n = 14) (low strength)

Shower (n = 15) (high strength)

Metcalf and Eddy, 2003

Sharp et al., 2006

39 ± 17 144 ± 63 3.7 ± 1.2 12 ± 4 35 ± 16 7.6 ± 3.0 0.5 ± 0.2 0.7 ± 0.7 3.9 ± 1.6 6.6–7.6 0.6 ± 0.1 2.5–3.5 40

166 ± 37 575 ± 98 3.5 ± 0.4 56 ± 7 42 ± 9 16.4 ± 3.0 1.3 ± 0.1 1.0 ± 0.3 7.5 ± 1.2 7.3–7.8 2.4 ± 0.1 1.5–2.5 30

110–450 250–800 1–3 80–260 nr 20–70 nr nr 0 nr nr 1.5–2.7a nr

nr nr nr 7–14 nr nr nr nr nr nr 5–15 5 59–75

In secondary treated effluent (Jarusutthirak et al., 2002; Hu et al., 2003); nr: not reported.

150

M. Pidou et al. / Chemosphere 71 (2008) 147–155

fractions separated during the XAD resin fractionation process (Table 1). Similar values are found for river water sources containing a high degree of sewage effluent were the hydrophilic components can constitute around 40– 60% of the total organic strength in the water (Parsons and Jefferson, 2006). Surfactants, commonly used in household products and proved to be the biggest fraction of the organic content of greywater (Wiel-Shafran et al., 2006), are known to contain both hydrophilic and hydrophobic sections (Ikehata and El-Din, 2007). However, the fractionation of three surfactants individually, sodium lauryl sulphate, an anionic surfactant commonly used as a thickener, cetyl trimethylammonium bromide, a cationic surfactant used as an antiseptic and TritonÒ X-100, a non-ionic surfactant used as a detergent, confirmed the tendency of greywater to be mainly hydrophilic. Indeed, the results of the fractionation on XAD resins were similar for the three surfactants with a hydrophilic fraction of 92%, 92% and 88% for the anionic, cationic and non-ionic surfactants, respectively. In relation to the electrical character of the greywaters studied, the colloids were negatively charged in natural pH environments with a zeta potential of 13.2 ± 0.7 mV and 19.4 ± 3.1 mV for the mixed and shower sources, respectively. The corresponding charge densities of the two waters were 0.6 ± 0.1 and 2.4 ± 0.1 meq g1 DOC indicating that the fresher shower water contained considerably more charged components compared to the mixed source water. Comparison with other species commonly coagulated suggests that the colloids in greywater exert a relatively low charge demand to the water per unit of organic material. For instance, reported values of other systems are around 5 meq g1 DOC for NOM (Kam and Gregory, 2001; Sharp et al., 2006), 0.1–3.2 meq g1 DOC for algal organic matter (Henderson et al., 2006) and 0.1–1 meq g1 for inorganic colloids such as kaolin (Edzwald, 1993). Interestingly, reported charge densities for the hydrophilic components within NOM are 1 ± 0.6 meq g1 DOC which is in between the values obtained for greywater. Conversion of the charge densities to charge concentration reveals that the greywater will exert a charge load of 0.0072 meq l1 and 0.134 meq l1 for the mixed source and shower water, respectively. Comparative levels for NOM are around 0.02–0.05 meq l1 (Sharp et al., 2006) such that the mixed source water represents relatively low coagulant demand but the shower water exerts a very high coagulant demand which is around 25–70 times greater than a typical NOM rich water. Calculation of the neutralising charge of coagulants based on speciation data (Jiang and Graham, 1998) reveals that iron based coagulants provide 35.5, 30.7, 12.6, 1.7 meq g1 Fe at pH values of 4, 5, 6, 7, respectively. Similarly, over the same pH value aluminium coagulants provide 104.5, 45.8, 12.5, 1.9 meq g1 Al . Hence the charge neutralising capacity of both coagulants is greater under acidic conditions. To illustrate, the neutralising capacity reduces by a factor of 8 and 3 for alum and iron, respectively, as the pH increases from 4 to 6. Interestingly,

at near neutral pH levels as found in greywater the neutralising capacity of both systems is very similar. Differences in the neutralising capacity of each coagulant are based on the weight of the metal ion such that conversion to a meq basis reveals both systems to provide almost identical capacities (Sharp et al., 2006). 3.2. Coagulation Both greywater sources were trialled over a range of coagulant doses and pH with the results presented in terms of BOD as it represents the most common compliance parameter for urban reuse (Fig. 1). In the mixed greywater, the residual BOD concentration remained quite constant at around 1–5 mg l1 for alum and 1–7 mg l1 for Ferric representing a removal efficiency between 68% and 99%. The response of the system appeared independent of pH over the ranges tested. In contrast, in the case of the shower water the residual BOD decreased as a function of dose until it reached a plateau value (Fig. 1c and d). To illustrate, the BOD decrease from 166 ± 37 mg l1 to 23 and 30 mg l1 for alum and ferric, respectively, representing removal efficiencies of 85% and 79%. In both cases the effectiveness of the coagulants was improved under more acidic conditions although this was more pronounced in the case of alum than ferric. For example, when coagulating with 18 mg l1 of alum the residual BOD was 69, 39 and 20 mg l1 at pH 7, 6 and 4.5, respectively. This had also an impact on the dose required to reach the plateau as in the case of alum the minimum dose necessary was 24, 28 and 32 mg l1 for pH values of 4.5, 6 and 7, respectively. Comparison of the two coagulants revealed that the maximum level of removal was around 85% for both systems suggesting little difference in performance. Comparison of the required doses showed that more ferric was required by mass to achieve a set level of removal. Conversion to molar concentrations indicates the required minimum dose was 0.79 mM for ferric and 0.89 mM for alum indicating that in fact proportionally more alum was required per unit of treatment. Comparison with COD, DOC and UV254 revealed that the level of treatment was less dose and pH dependant in terms of these parameters. To illustrate, the most apparent case was in terms of UV254 where the level of removal remained around 74% in case of alum. Information on the types of organics that are removed is provided by comparing the SUVA before and after treatment. In the case of both coagulants the SUVA decreased from 2.5–3.5 in the raw water to 0.6 post-coagulation suggesting that the hydrophobic components have been predominately removed. 3.3. MIEXÒ Treatment of both greywater sources with MIEXÒ showed a similar pattern to coagulation. For the mixed greywater (Fig. 2a), the treatment appeared to be independent of the contact time and dose, with BOD residual con-

M. Pidou et al. / Chemosphere 71 (2008) 147–155

a

pH

7

6

4.5

pH

BOD (mg.l-1)

16

BOD (mg.l-1)

40

b

20

12 8

151

4 0

7

20 10

5

10

15

20

0

5

10

Al dose (mg.l-1)

7

6

4.5

BOD (mg.l-1)

pH

20

60

80

200

d

160

15

Fe dose (mg.l-1)

200

BOD (mg.l-1)

4.5

0 0

c

6

30

120 80

pH

160

7

6

4.5

120 80 40

40

0

0 0

10

20

30

40

50

0

60

20

-1

40 -1

Al dose (mg.l )

Fe dose (mg.l )

Fig. 1. Influence of coagulant dose and pH on biochemical oxygen demand (BOD) removal in the (a and b) mixed and (c and d) shower greywaters.

a

b Contact time (min)

10

20

30

45

250

60

Contact time (min)

200

BOD (mg.l-1)

60

-1

BOD (mg.l )

80

40

20

10

30

60

90

150 100 50

0

0 0

5

10

15

20

25

30

-1

MIEX dose (ml.l )

35

0

20

40

60

-1

MIEX dose (ml.l )

Ò

Fig. 2. Influence of MIEX dose and contact time on biochemical oxygen demand (BOD) removal in the (a) mixed and (b) shower greywaters.

centrations between 1 and 14 mg l1 over the range of conditions tested. Corresponding removal efficiencies varied between 80% and 99% which reflects the low level of BOD in the influent. In contrast, the residual BOD of the shower greywater decreased as both the contact time and dose of MIEXÒ increased (Fig. 2b). To illustrate, at a 10 ml l1 MIEXÒ dose, the BOD residual was 60.2, 32.7, 27.6 and 17.6 mg l1 for respective contact times of 10, 30, 60 and 90 min. Whereas, for a 10-min contact time, the BOD residual was 112.7, 85.7, 60.2 and 40.2 mg l1 for a MIEXÒ dose of 2, 5, 10 and 20 ml l1, respectively. However, the residual BOD reached a plateau with a removal efficiency of around 83% once the dose reached 20 ml l1 or above. Similar patterns were observed in terms of COD, DOC and UV254 except considerably less varia-

tion was observed between contact times in terms of UV254. For example, in the case of the shower water dosed at a MIEXÒ concentration of 10 ml l1 the residual UV254 was 0.158, 0.132, 0.134 abs at contact times of 10, 30 and 60 min, respectively. Comparison of the different parameters reveals that the feed water has a specific UV absorbance of between 2.5 and 3.5 suggesting the water is moderately hydrophilic in nature compared to the effluent which is always less than 1 suggesting the residual organics are mostly hydrophilic in nature such that the MIEXÒ process appears to be targeting the same organics as the coagulation process. This agrees with the work of Fearing et al. (2004) who studied the use of MIEXÒ for NOM removal and showed that only the very small, most hydrophilic material remains after treatment.

152

M. Pidou et al. / Chemosphere 71 (2008) 147–155

3.4. MIEXÒ + coagulation

confirms once more the preference of treatment of MIEX and coagulation for hydrophobic and high MW materials.

A series of experiments were conducted on a combined system for which the greywater was first treated with the optimum dose of MIEXÒ and then coagulated with different doses of either ferric or alum. As both treatment systems were capable of achieving a sufficient treatment according to the water quality standards for the low strength greywater the combined treatment was only tested on the shower water. Residual BOD concentrations varied between 20 and 40 mg l1 under all the conditions tested in terms of coagulant choice, dose and pH (Fig. 3). Comparison with the previous tests reveals that the combined system was not able to reduce the BOD concentration below the level previously obtained. However it was found that this treatment was achieved independently of the coagulant dose or pH. Similar trends to the one presented for BOD were found for COD, DOC and UV254. And removals of 64% COD, 53% DOC and 70% UV254 were observed. Similarly, Singer and Bilyk (2002) reported removals between 53 and 96% for UV254 and between 46% and 87% for TOC for the treatment of natural waters with MIEXÒ in combination with alum. The lower removals were observed for the raw waters with the lower SUVA. To illustrate, UV254 removals of 53% and 94% were recorded for waters with a SUVA of 1.4 and 4.5 l mg1 m1, respectively. This

a

250

pH

7

6

3.5. Comparison of the systems With an organic removal from 68% to 100%, both systems, coagulation and ion exchange, proved to be efficient to treat the mixed greywater to the most stringent standard for reuse with the lowest doses and contact times tested. However, it must be noted that the raw water strength was very low and it is perhaps not too surprising that the systems were capable of removing sufficient materials to meet the compliance standard. In the case of the shower water, optimum conditions for MIEXÒ was shown to be 10 ml l1 at a contact time of 30 min (Table 2) to achieve a DOC removal of 54%. Similar dose and contact time were used for NOM removal in drinking water for a final DOC removal of 60% (Fearing et al., 2004). The optimum conditions for coagulation were always observed at pH 4.5 and a dose of 24 mg l1 (0.89 mM) alum and 44 mg l1 (0.79 mM) ferric when used alone and a reduced dose of only 5 mg l1 for both coagulants when used in conjunction with MIEXÒ. With these optimum doses of both ferric and alum, a general removal of 64% in terms of COD was then achieved. Comparable studies presented similar results. Lin et al. (2005) showed that 25 mg l1 of alum

b

4.5

250

7

6

4.5

200

BOD (mg.l-1)

BOD (mg.l-1)

200

pH

150 100 50

150 100 50 0

0 0

10

20

30

0

40

10

20

30

40

Fe dose (mg.l-1)

Al dose (mg.l-1)

Fig. 3. Influence of coagulant dose and pH on biochemical oxygen demand (BOD) removal in the shower greywater after treatment with MIEXÒ (10 ml l1, 30 min) and coagulation ((a) alum and (b) ferric).

Table 2 Shower greywater characteristics after treatment with the different systems at optimum conditions Optimum

Raw

MIEXÒ 10 ml l1, 30 min

Alum 24 mg l1, pH 4.5

Ferric 44 mg l1, pH 4.5

MIEXÒ + Al 5 mg l1, pH 4.5

MIEXÒ + Fe 5 mg l1, pH 4.5

Turbidity (NTU) COD (mg l1) BOD (mg l1) DOC (mg l1) TN (mg l1) 1 NHþ 4 (mg l ) 1 (mg l ) NO 3 1 PO3 4 (mg l ) Total coliforms (MPN 100 ml1) Escherichia coli (MPN 100 ml1) Faecal Enterococci (MPN 100 ml1)

46.60 791 205 171.4 18 1.2 6.7 1.66 56 500 6490 2790

8.14 272 33 78.2 15.3 1.1 4.7 0.91 59 8 <1

4.28 287 23 93.4 15.7 1.2 5.7 0.09 <1 <1 <1

5.20 288 30 87.4 17.9 1.2 6.1 0.06 <1 1 <1

3.01 247 27 78.8 15.3 1.2 4.4 0.11 <1 <1 <1

3.30 254 29 80.7 17.4 1.2 4.8 0.13 <1 <1 <1

M. Pidou et al. / Chemosphere 71 (2008) 147–155

was needed to achieve a COD removal of 60% in a greywater treated by electro-coagulation. And a dose of only 5 mg l1 alum was needed to achieve a 36% removal in a laundry wastewater with an initial COD of 280 mg l1 (Sostar-Turk et al., 2005). Effluent characteristics in the current study were similar for all five systems tested with a slight improvement in COD and DOC removal observed for the combined MIEXÒ and coagulant systems (Table 2). Residual turbidity was measured as 8.1, 4.2 and 5.2 for MIEXÒ, alum and ferric, respectively. In comparison the levels decreased to 3.3 and 3 NTU for the combined systems with ferric and alum, respectively. This is once again similar to results found in the literature. El Samrani et al., 2004 reported a reduction of the turbidity from 40 to 5 NTU in sewage treated with a dose of 43 mg l1 (0.77 mM) of ferric. The removal of the total coliforms was excellent at 3 log (99.8%) with MIEXÒ corresponding to a residual of 59 MPN 100 ml1. In comparison all the other systems recorded a non-detectable level in the effluent. 4. Discussion Comparison with water recycling standards required for urban reuse indicates that coagulation and MIEXÒ are not always able to meet the required levels of treatment for all situations. In the case of the shower water the treatment systems failed to comply with both the turbidity and organic concentration requirements. Comparison of the influent water strengths indicates that the maximum strength of the mixed systems was 72 mg l1 BOD and the minimum of the shower water was 110 mg l1. Consequently a threshold organic concentration value would appear to exist between these two limits beyond which coagulant is unlikely to be able to meet effluent standards. In fact the systems appear unable to remove sufficient organics even at high doses of coagulant or MIEXÒ indicating that it is likely to be a recalcitrant proportion of the greywater to chemical solutions. MIEXÒ is an ion exchange process specifically designed to remove NOM from potable water and is believed to be effective at removing mid and higher MW compounds. The results in the current study support this suggestion as the greatest removal was achieved with MIEXÒ alone (54%). In comparison in potable water treatment DOC removals with MIEXÒ alone are commonly 10–20% lower than with coagulants although in combined systems the overall removal is slightly better than either MIEXÒ or coagulation alone (Fearing et al., 2004). The main coagulation mechanisms at the optimum pH are charge neutralisation for colloidal material and charge complexation for soluble material (Sharp et al., 2006). In both cases the process is driven by charge interactions such that preferentially removal of charged materials is likely to occur. Consequently low charge or neutral materials are likely to be poorly removed although some removal is possible due to adsorption mechanisms on to the pre-formed

153

flocs. Such suggestions are supported as the zeta potential was always within the ranged previously reported for charge dominated processes (Sharp et al., 2006). Examination of standard speciation diagrams suggests that at the high doses applied sweep flocculation from the precipitated hydroxide should be the dominate removal mechanisms. Whilst this is possible the complexation reaction between organics and the metal coagulants are known to be fast such that it is likely that the complex precipitates rather than the straight hydroxide. Typical dose requirements for organic dominated systems are around 1:1 on a mass DOC per mass coagulant basis (Jarvis et al., 2005). In the current study on the shower water optimum doses were observed at ratios of 7 for alum and 3.9 for ferric. The much lower levels required here relate to the nature of the feed water and its relatively low charge density. Conversion of the data to a mass ratio based just on the hydrophobic content of the greywater converts the dose ratio to around 1.2:1 which is around the level reported for coagulation of hydrophobic rich waters and indicates the differences between the dose ratios used for NOM removal and greywater are based on there respective hydrophilic contents. Comparison to coagulation of organics from other fields reveals a low percentage removal in the case of greywater. To illustrate, typical DOC removals during the coagulation of NOM are between 60% and 80% (Fearing et al., 2004) and this compares to 40–50% in the current case. The difference can be attributed to the make up of the organic molecules in both cases. NOM is commonly rich in anionic hydrophobic humic and fulvic acids which are easily removed by coagulants. The residual are generally the smaller, hydrophilic neutrals that are difficult to remove by most treatment methods. Typical MW sizes for NOM are between 2 and 5 kDa for the hydrophobics and <2 kDa for the hydrophilics based on UV absorption (Fearing et al., 2004). Greywater in contrast is mainly made up of <3 kDa material (Jefferson et al., 2004b) and appears to be mainly hydrophilic in nature (Table 1). Indeed, a relationship is known to exist between the hydrophilic concentration of NOM in the raw water and the residual DOC that can be achieved under optimum coagulation conditions (Sharp et al., 2006). The current data suggest that greywater coagulation fits into a similar correlation whereby the residual DOC after coagulation is approximately 80–90% of the raw water hydrophilic content. Further support is provided by analysis of the SUVA which is between 2 and 3 in the raw water but always 1 or lower in the treated waters suggesting that all the tested processes are removing the hydrophobic components form the water. Overall, the findings suggests that whilst chemicals processes such as coagulation and MIEXÒ can achieve sufficient levels of organic removal to meet some standards for reuse, they are not capable of meeting the most stringent of reuse standards reported around the world. This limit appears to be a fundamental one based on the character of the raw water and suggest that such processes are

154

M. Pidou et al. / Chemosphere 71 (2008) 147–155

never going to be suitable if very tight standards are required without the use of downstream processes such as adsorption to achieve the remaining removal. However, reuse standards are not always so strict especially for applications not based around dwelling such as landscape and garden irrigation, fire protection, vehicle washing etc. (Lu and Leung, 2003). Consequently, chemical processes such as those discussed above can have a role where less stringent standards are required. In particular, they provide a possible alternative to biological systems where such processes are not preferred or technically difficult to operate. Ultimately, the economical feasibility of the systems will determine their potential application. However, based on coagulant costs of £0.05–0.20 kg1 and £0.09–0.42 kg1 for alum and ferric, respectively (Akhtar et al., 1997; Gebbie, 2005; Ahmed, 2007), the costs of coagulants at optimum dose would be £0.0012–0.0048 m3 and £0.0040– 0.0185 m3 of treated greywater. Although better treatment was found for acidic pH, it is not advised to run a full scale system in this condition. Indeed, pH adjustment before treatment and readjustment after, depending on the reuse application, would be required and would then increase the costs of the system. Considering an average drinking water price of £0.9 m3 (European Environment Agency, 2003) and a greywater production of about 100 l d1 capita1 (Kujawa-Roeleveld and Zeeman, 2006), annual savings of £32.2, £321.7 and £1608.7 per year are possible for 1, 10 and 50 persons respectively which would need to meet the capital cost of the equipment and associated running costs. Whilst exact capital costs are difficult to estimate it appears unlikely that adoption of chemical solution will be driven from a purely economic standpoint at small scale. 5. Conclusion Although good organic removal, comparable to the data seen in the literature for NOM removal in potable water or organics in other wastewater, has been observed for the different systems tested, they showed limitations to meet the standards for reuse. The chemical solutions tested in the current study have revealed that both MIEXÒ and coagulation are suitable treatment solutions for low strength greywater sources. However, all the systems tested were unable to achieve the required level of treatment for the reuse of medium to high strength greywaters. Chemical solutions appear to be limited due to the recalcitrant nature of a proportion of the greywater which prevents the necessary level of treatment being achieved. Ultimately chemical treatment solutions with coagulants and ion exchange resin appear to be limited for the reuse of greywater within urban environments. Acknowledgements This work forms part of the ‘Water Cycle Management for New Developments’ (WaND) project funded under the Engineering & Physical Science Research Council’s ‘Sus-

tainable Urban Environment’ Programme by EPSRC, UK government and industrial collaborators (www.wand. uk.net). References Ahmed, S.A.R., 2007. Fast-track evaluation of a compact chemically enhanced-trickling filter system. Braz. J. Chem. Eng. 24, 171–184. Akhtar, W., Rauf, M., Ali, I., Nayyeemuddin, 1997. Optimum design of sedimentation tanks based on settling characteristics of Karachi tannery wastes. Water Air Soil Pollut. 98, 199–211. APHA, 1992. Standard Methods for the Examination of Water and Wastewater, 18th ed. American Public Health Association, Washington, DC. Boyer, T.H., Singer, P.C., 2005. Bench-scale testing of a magnetic ion exchange resin for removal of disinfection by-product precursors. Water Res. 39, 1265–1276. Edzwald, J.K., 1993. Coagulation in drinking water treatment: Particles, organics and coagulants. Water Sci. Technol. 27 (11), 21–35. Edzwald, J.K., Tobiason, J.E., 1999. Enhanced coagulation: US requirements and a broader view. Water Sci. Technol. 40 (9), 63–70. El Samrani, A.G., Lartiges, B.S., Montarges-Pelletier, E., Kazpard, V., Barres, O., Ghanbaja, J., 2004. Clarification of municipal sewage with ferric chloride: the nature of coagulant species. Water Res. 38, 756– 768. Eriksson, E., Auffarth, K., Henze, M., Ledin, A., 2002. Characteristics of grey wastewater. Urban Water 4, 85–104. European Environment Agency, 2003. Indicator fact sheet WQ05, water prices. (accessed on 08/2007). Fearing, D.A., Banks, J., Guyetand, S., Monfort Eroles, C., Jefferson, B., Wilson, D., Hillis, P., Campbell, A.T., Parsons, S.A., 2004. Combination of ferric and MIEX for the treatment of humic rich water. Water Res. 38, 2551–2558. Friedler, E., Kovalio, R., Galil, N.I., 2004. On-site greywater treatment and reuse in multi-storey buildings. In: Proc. of the 1st International Conference on Onsite Wastewater Treatment & Recycling, 11–13th February, Fremantle, Western Australia. Gebbie, P., 2005. A dummy’s guide to coagulants. In: Proc. of the 68th Annual Water Industry Engineers and Operators’ Conference, 7th and 8th September, Bendigo, Australia. Gross, A., Shmueli, O., Ronen, Z., Raveh, E., 2007. Recycled vertical flow constructed wetland (RVFCW) – a novel method of recycling greywater for irrigation in small communities and households. Chemosphere 66, 916–923. Henderson, R., Parsons, S.A., Jefferson, B., 2006. Optimisation of coagulation for algae treatment. In: Proc. Water Quality Technology Conference and Exposition (WQTC), 5–9th November, Denver, USA. Hu, J.Y., Ong, J.H., Shan, J.H., Kang, J.B., Ng, W.J., 2003. Treatability of organic fractions derived from secondary effluent by reverse osmosis membrane. Water Res. 37, 4801–4809. Ikehata, K., El-Din, M.G., 2007. Degradation of recalcitrant surfactants in wastewater by ozonation and advanced oxidation processes: A review. Ozone-Sci. Eng. 26, 327–343. Jarusutthirak, C., Amy, G., Croue´, J.-P., 2002. Fouling characteristics of wastewater effluent organic matter (EfOM) isolates on NF and UF membranes. Desalination 145, 247–255. Jarvis, P., Jefferson, B., Parsons, S.A., 2005. How the natural organic matter to coagulant ratio impacts on floc operational properties. Environ. Sci. Technol. 39, 8919–8924. Jefferson, B., Laine, A., Diaper, C., Parsons, S., Stephenson, T., Judd, S.J., 2000. Water recycling technologies in the UK. In: Proc. Technologies for Urban Water Recycling Conference, Cranfield University, 19th January. Jefferson, B., Jarvis, P., Parsons, S.A., 2004a. The effect of coagulant type on natural organic matter floc structure and strength. In: Proc. of the 11th Intl. Gothenburg Symp. on Chemical Treatment of Water and Wastewater, 8–10th November, Orlando, Florida.

M. Pidou et al. / Chemosphere 71 (2008) 147–155 Jefferson, B., Palmer, A., Jeffrey, P., Stuetz, R., Judd, S.J., 2004b. Greywater characterisation and its impact on the selection and operation of technologies for urban reuse. Water Sci. Technol. 50 (2), 157–164. Jiang, J.-Q., Graham, N.J.D., 1998. Pre-polymerised inorganic coagulants and phosphorus removal by coagulation – a review. Water SA 24, 237– 244. Kam, S.-K., Gregory, J., 2001. The interaction of humic substances with cationic polyelectrolytes. Water Res. 35, 3557–3566. Karanfil, T., Schlautman, M.A., Erdogan, I., 2002. Survey of DOC and UV measurement practices with implications for SUVA determination. J. Am. Water Works Ass. 94 (12), 68–80. Karpiscak, M.M., Foster, K.E., Schmidt, N., 1990. Residential water conservation. Water Res. 26, 939–948. Kujawa-Roeleveld, K., Zeeman, G., 2006. Anaerobic treatment in decentralised and source-separation-based sanitation concepts. Rev. Environ. Sci. Bio-Technol. 5, 115–139. Lin, C.-J., Lo, S.-L., Kuo, C.-Y., Wu, C.-H., 2005. Pilot-scale electrocoagulation with bipolar aluminium electrodes for on-site domestic greywater reuse. J. Environ. Eng. 131, 491–495. Lu, W., Leung, A.Y.T., 2003. A preliminary study on potential of developing shower/laundry wastewater reclamation and reuse system. Chemosphere 52, 1451–1459. Malcolm, R.L., McCarthy, P., 1992. Quantitative evaluation of XAD-8 and XAD-4 resins used in tandem for removing organic solutes from water. Environ Int. 18, 597–607. Metcalf and Eddy, Inc., 2003. In: Tchobanoglous, G., Burton, F.L., Stensel, H.D. (Eds.), Wastewater Engineering – Treatment, Disposal and Reuse, fourth ed. McGraw-Hill, New York. Nolde, E., 1999. Greywater reuse systems for toilet flushing in multi-storey buildings – over ten years experience in Berlin. Urban Water 1, 275– 284. Parsons, S.A., Jefferson, B., 2006. Introduction to Potable Water Treatment Processes. Blackwell Publishing, Oxford, UK.

155

Ratnaweera, H., Hiller, N., Bunse, U., 1999. Comparison of the coagulation behavior of different Norwegian aquatic NOM sources. Environ. Int. 25 (2–3), 347–355. Santala, E., Uotila, J., Zaitsev, G., Alasiurua, R., Tikka, R., Tengvall J., 1998. Microbiological grey water treatment and recycling in an apartment building. In: Proc. of the 2nd International Advanced Wastewater Treatment, Recycling and Reuse, 14–16th September, Milan, Italy. Sharp, E.L., Parsons, S.A., Jefferson, B., 2004. The effects of changing NOM composition and characteristics on coagulation performance, optimisation and control.. Water Sci. Technol. Water Supply 4 (4), 95– 102. Sharp, E.L., Parsons, S.A., Jefferson, B., 2006. The impact of seasonal variations in DOC arising from a moorland peat catchment on coagulation with iron and aluminium salts. Environ. Pollut. 140, 436– 443. Singer, P.C., Bilyk, K., 2002. Enhanced coagulation using a magnetic ion exchange resin. Water Res. 36, 4009–4022. Sostar-Turk, S., Petrinic, I., Simonic, M., 2005. Laundry wastewater treatment using coagulation and membrane filtration. Resour. Conserv. Recy. 44, 185–196. Surendran, S., Wheatley, A.D., 1998. Greywater reclamation for nonpotable reuse. Water Environ. J. 12, 406–413. Tajima, A., 2005. The behaviour of the pathogenic microbes in the treated wastewater reuse system and the establishment of the new technical standard for the reuse of treated wastewater. In: Proc. of the IWA Specialty Conference on Wastewater Reclamation and Reuse for Sustainability, 8–11th November, Jeju, Korea. USEPA (US Environmental Protection Agency), 2004. Guidelines for Water Reuse. Report EPA/625/R-04/108, USEPA, Washington, DC, USA. Wiel-Shafran, A., Ronen, Z., Weisbrod, N., Adar, E., Gross, A., 2006. Potential changes in soil properties following irrigation with surfactant-rich greywater. Ecol. Eng. 26, 348–354.