- Email: [email protected]
H2O2 with and without coagulation pre-treatment
Chemical Engineering Journal 260 (2015) 649–656
Contents lists available at ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Treatment of municipal wastewater reverse osmosis concentrate using UVC-LED/H2O2 with and without coagulation pre-treatment M. Umar a, F.A. Roddick a,⇑, L. Fan a, O. Autin b, B. Jefferson b a b
School of Civil, Environmental and Chemical Engineering, RMIT University, GPO Box 2476, Melbourne, 3001 Victoria, Australia Cranfield Water Sciences Institute, Cranfield University, Cranfield, Bedfordshire MK43 0AL, United Kingdom
h i g h l i g h t s UVC/H2O2 treatment with UV-LEDs can degrade organic matter in RO concentrate. UVC/H2O2 treatment increased the biodegradability of RO concentrate. UV-LEDs were markedly more energy efficient than a conventional mercury UV lamp. Alum coagulation plus UVC/H2O2 treatment led to complementary effect on DOC removal.
a r t i c l e
i n f o
Article history: Received 8 July 2014 Received in revised form 2 September 2014 Accepted 7 September 2014 Available online 16 September 2014 Keywords: UVC/H2O2 UV-LED Reverse osmosis concentrate Alum Coagulation Biodegradability
a b s t r a c t The potential of a prototype batch reactor using ultraviolet light emitting diodes (UVC-LEDs) which emit at 255 nm in conjunction with H2O2 for the treatment of a highly saline (electrical conductivity 22 mS/cm; DOC 32–37.5 mg/L) municipal wastewater reverse osmosis concentrate was investigated. Mineralization of organic content (measured as DOC) was low (22%) due to the low fluence rate (0.14 mW/cm2), however, a large reduction in colour (94%) and A254 (75%) occurred after delivering a UV fluence of 48 103 mJ/cm2 at the original pH of 8.3. Fairly similar results were obtained at pH 7, but the reduction of DOC increased at lower pH with 38% and 36% achieved at pH 4 and 5, respectively. Similar trends were observed for colour and A254 reduction. These results, in conjunction with excitation– emission matrix spectra, biological dissolved organic carbon (BDOC) assay and apparent molecular size distribution, demonstrated that the prototype system led to the breakdown of the chromophore bonds and thus changes in the molecular structure, and degradation of high molecular weight (MW) compounds to low MW compounds. Coagulation (1.5 and 3 mmol L1 Al3+ at pH 5) led to a significant reduction of DOC (34–38%), colour (50–66%) and A254 (47–54%), and subsequent UVC/H2O2 treatment led to further reduction in these parameters. For a target DOC reduction of 15 mg/L, the EE/O was 15 kWh/m3 when coagulation was used as pre-treatment to the UVC/H2O2 treatment (UV fluence 36 103 mJ/cm2) and it reduced to less than half after biological treatment (as BDOC assay). This study demonstrated the potential of UV-LEDs as an alternative UV source for degrading the organic matter in ROC using advanced oxidation. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Reverse osmosis (RO) is widely used as a polishing treatment for secondary effluent in wastewater reclamation schemes. However the resultant RO concentrate (ROC), which comprises 15–20% of the volume of the feed stream, contains almost all the contaminants present in the original wastewater at elevated levels. Depending on the wastewater source, these contaminants may be ⇑ Corresponding author. Tel.: +61 3 9925 2080; fax: +61 3 9639 0138. E-mail address: [email protected] (F.A. Roddick). http://dx.doi.org/10.1016/j.cej.2014.09.028 1385-8947/Ó 2014 Elsevier B.V. All rights reserved.
toxic and/or bio-accumulative, and so disposal of untreated ROC presents a potential environmental risk. The addition of chemicals such as antiscalants, biocides and acids during the treatment process further complicates the situation as they can change the character of the organic and inorganic pollutants and can influence the chemical equilibrium of the dissolved constituents [1]. Tang et al. [2] demonstrated that the genotoxicity of a ROC was 500–559 lg 4-NQO (4-nitroquinoline-1-oxide)/L, which was much higher than for the RO influent (105–160 lg 4-NQO/L). As such, it is mandatory for some facilities to treat ROC prior to its discharge. An example is the Bundamba advanced wastewater treatment plant (Brisbane,
650
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
Australia), which is required treat and monitor the ROC prior to its discharge to the Brisbane River [3]. In addition to minimizing its environmental impacts, beneficial reuse of the treated ROC can help to offset the treatment costs [4]. The application of ultraviolet (UV) irradiation-based advanced oxidation processes (AOPs) has been extensively investigated for the degradation of a wide range of contaminants present in water and wastewater. An AOP involves the in situ generation of hydroxyl radicals (HO) which react non-selectively with organic contaminants at very high rates (rate constant of 108–1010 M1 s1) [5]. UV plus hydrogen peroxide (UV/H2O2) is one of the most promising AOPs which generates HO through photolysis of H2O2. Since the quantum yield of the HO generation due to the direct UV photolysis of H2O2 is near unity and wavelength independent, the process is considered highly efficient and is leading the way in commercial applications [6]. Conventional UV lamps continue to be the main source of radiation, however, alternative UV sources are being sought due to the inherent disadvantages of UV arc lamps such as the use of mercury, low mechanical stability, large size and low energy efficiency. UV light emitting diodes (UV-LEDs) which are a mercury-free source of monochromatic UV radiation are potential alternatives and are regarded as the most promising UV light sources [7]. UV-LEDs are small, compact, robust and durable [8] and offer virtually no limitations on the potential geometry of the emission sources [7]. Furthermore, UV-LEDs are less energy intensive than traditional UV lamps as they convert a greater amount of energy into light (high quantum yield) because the light emission occurs due to the recombination of electrons and holes [9]. Like conventional diodes, UV-LEDs are comprised of a chip of semi-conducting material impregnated or doped with impurities to create a p-n junction capable of emitting light in a narrow wavelength range in the form of electroluminescence [10]. Semiconductor crystals of compounds containing aluminium, nitrogen and gallium are the main materials used in the construction of UV-LEDs [10]. The wavelength of light emitted depends on the band gap energy of the type of material used to construct the UV-LEDs [11] and LEDs with emission wavelengths as low as 210 nm have been developed [10]. Applications of UV-LEDs include water disinfection [8] and degradation of organic compounds such as phenol [12], formaldehyde [13], methylene blue [14], metaldehyde [15] and reactive red 22 dye [16]. Although successfully used for disinfection of water and degradation of individual organic compounds, the potential of UV-LEDs for treating real wastewater is yet to be reported. Furthermore, most of the foregoing studies used high wavelength UV-LEDs (>360 nm) except some [11,15] that used deep UVC-LEDs (<290 nm). Although low output power and high cost are the major limitations at this stage, continuing developments in this technology have been projected to significantly overcome these by 2020 [17]. The trend of development between 2007 and 2012 has followed Haitz’s law which forecasts an increase in output per bulb by a factor of 20 and a decrease in cost by a factor of 10 per decade [18]. The aim of this study was to determine the potential of deep UVC-LEDs in the presence of H2O2 for degrading the organic content of municipal wastewater ROC and its removal after subsequent biological treatment. The removal of organic content by this sequence was compared with that for coagulation (alum), and for when coagulation was used as a pre-treatment to the UVC-LEDs/ H2O2 process to produce a target residual DOC concentration of 15 mg/L. The change in biodegradability was investigated by determining the biological dissolved organic carbon (BDOC) content, and fluorescence excitation-emission matrix (EEM) spectra and liquid chromatography – organic carbon detection (LC-OCD) were used to track the changes in the organic components of the ROC.
2. Materials and methods 2.1. Collection and characterization of ROC The composite ROC sample was collected from a wastewater reclamation facility at a local municipal wastewater treatment plant (WWTP), and stored at 4 °C. In the treatment process at the WWTP, raw sewage is screened and de-gritted and sent to intermittently decanted extended aeration (IDEA) bioreactors where it is treated in a cycle of aeration, settling, and decant. The IDEA effluent is then treated using a combination of ultrafiltration and RO. The sample was characterized (Table 1) according to Standard Methods [19]. The sample was high in alkalinity, salinity and total dissolved solids (TDS) concentration. 2.2. UV-LED reactor UV-LEDs with output wavelength of 255 nm were purchased from Sensor Electronic Technology (Columbia, South Carolina, USA). According to the manufacturer’s specifications, each LED had an emission power of approximately 0.33 mW at 20 mA. The LEDs were soldered to a plate and connected in two series, each containing 5 units. The distance between each LED unit was 17 mm and between the sample surface and LEDs was 7 mm, based on the findings of Autin et al. [15]. A DC power supply (INSTEK GPR-6030 D, 0–60 V, 0–3 A) was utilised. A schematic representation of the experimental set up is shown in Fig. 1. Tests were performed in batch mode using 50 mL of ROC sample placed in a rectangular Perspex vessel which was continuously stirred using a magnetic stirrer; the liquid depth was 7 mm. The average fluence rate of the system was 0.14 mW/cm2, as determined by uridine actinometry [20]. As the treatment time was extended, some evaporation of the sample was observed, consequently the results were adjusted by taking into account the difference in volume after various irradiation times. All experiments were performed at room temperature (22 °C). The temperature was regularly monitored and only a small increase (2–3 °C) in the temperature of the ROC was noted during operation. Preliminary tests using a range of 2–6 mmol L1 H2O2 demonstrated that 3 mmol L1 was a good compromise taking into account the chemical consumption and DOC reduction and therefore was chosen to be used in this study. All experiments were performed in duplicate, and in some cases in triplicate when the difference was more than 5%. Average values of duplicate or triplicate runs are reported, and error bars represent the standard deviation for these results. 2.3. Alum coagulation A stock solution of alum was prepared using Al2(SO4)318H2O (Chem-Supply, Pty Ltd, Australia). Coagulation was conducted with
Table 1 Characteristics of ROC. Parameter
Value
DOC (mg/L) COD (mg/L) pH Colour (Pt.Co) Chloride (mg/L) TDS (mg/L) A254 (1/cm) SUVA (L/mg-m) Alkalinity (as CaCO3, mg/L) Conductivity (mS/cm)
32–37 105 8.3 158 8060 16,140 0.68 1.8 410 22.3
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
651
Fig. 1. Schematic representation of the experimental set-up.
a laboratory jar test apparatus (Phipps and Bird, PB-700) using 2 L ROC sample. The samples were rapidly mixed for 2 min at 250 rpm followed by slow mixing for 30 min at 30 rpm and subsequent settling for 2 h before taking supernatant for analyses. The pH was adjusted using 1 mol L1 H2SO4 or 1 mol L1 NaOH. 2.4. Analytical methods DOC was determined using a total organic carbon analyser (Sievers model 5310C) in in-line mode to purge inorganic carbon, acid/ oxidiser ratio 0.8 and after dilution of the ROC. The DOC of some samples was confirmed using a Shimadzu TOC analyser (TOC-L series) without any dilution and the difference was found to be less than 3–5%. Absorbance was determined using a double beam scanning UV/vis spectrophotometer (Unicam UV2). True colour of the samples was measured with a Hach DR 4000 spectrophotometer at 455 nm in Platinum Cobalt (Pt.Co) units. Fluorescence EEM spectra were determined with a Perkin Elmer LS-50B luminescence spectrometer. The changes in fluorescence were quantitatively analysed using the fluorescence regional integration (FRI) technique which integrates the volumes under the EEM regions [21]. The biodegradability of the organics was evaluated as BDOC using the assay introduced by Joret and Levi [22] and modified by Volk et al. [23]. The DOC was measured daily and the BDOC was determined as the difference between the initial and the lowest DOC recorded over a 5 day period. Removal of residual H2O2 to reduce the impact on biological activity prior to the BDOC assay was found unnecessary. All analyses were conducted in duplicate and average results are reported. The concentration of residual H2O2 was estimated using MerckoquantÒ peroxide test strips. Apparent molecular size distribution was determined using liquid chromatography with organic carbon detection (LC-OCD) at the Water Research Centre of the University of New South Wales (Sydney, Australia) with a DOCLabor LC-OCD Model 8, with a Toyopearl TSK HW-50S column, using a phosphate buffer of pH 6.4 as the mobile phase. The samples were diluted to reduce the DOC and TDS to levels suitable for the analysis. 3. Results 3.1. Treatment efficiency of UVC/H2O2 and the effect of pH Treatment of ROC was carried out at the original pH of 8.3 and at pH 7, 5 and 4 to compare the treatment efficiency in terms of DOC, colour, A254 and specific ultraviolet absorbance (SUVA). For UV fluence of 48 103 mJ/cm2 (96 h exposure) at the original pH, the reduction of DOC was 22%, and reductions in colour and A254
were 94% and 75%, respectively. Little difference in the reduction of DOC was observed at pH 7 and 8.3 (Fig. 2a). Given the low intensity of UV irradiation of the UV-LEDs, the production of hydroxyl radicals was expected to be low. One way of improving the system is to minimize the concentration of the 2 HO scavengers (HCO 3 /CO3 ) by lowering the pH. Reducing the pH to 5 resulted in a greater reduction of DOC, i.e., 36% for UV fluence of 48 103 mJ/cm2; a marginally greater reduction of DOC was obtained at pH 4 (Fig. 2a). The reduction of colour was greater and faster at pH 4 than all other pH values, however the final reductions were comparable (Fig. 2b). This comparable reduction of colour (93–97%) at all pH values demonstrates that the disintegration of chromophore bonds was little influenced by alkalinity. Consistent with the trend for colour, the reduction of A254 was greater under acidic conditions and although more rapid at pH 4 than pH 5, final reductions were fairly similar (80–84%) (Fig. 2c). Little difference in the reduction of colour and A254 was noted for pH 7 and pH 8.3 at all UV fluence values. The trends for SUVA (decreased from 1.8 to 0.5–0.8 after UV fluence of 48 103 mJ/cm2) were expected given those for DOC and A254, indicating the loss of aromatic structures but incomplete disintegration of the molecules and hence low mineralization. The high reductions of colour and A254 and corresponding low reduction of DOC were attributed to the partial breakdown of large molecules (humic-like compounds) to low MW and/or non-coloured compounds and were consistent with our previous findings using a conventional UV lamp system [24]. It should be noted that adequate H2O2 was present for UV fluence of 48 103 mJ/cm2 (residual H2O2 > 25 mg/L, i.e., 0.73 mmol L1) to ensure continuous generation of HO, sufficient to partially oxidize the organic content but not enough to bring about major molecular changes or mineralize the humic-like compounds, leaving major fragments or bulk of the original molecules intact. As there was little difference in the results for pH 4 and pH 5 and the acid and subsequent alkali addition required if the lower pH were used, pH 5 was selected for use in further experiments. A control test was performed comprising the addition of H2O2 to ROC at pH 5 in the absence of irradiation. Some reduction of colour (5%) but no reduction of DOC or A254 was observed over 96 h. Similarly, little reduction of colour (4–5%) and no change in DOC and A254 value was noted in a control test performed at pH 5 without the addition of H2O2. Chloride ions are major scavengers of hydroxyl radicals, particularly in the absence of bicarbonate ions. As noted by Liao et al. [25] pH is more important than chloride concentration in the presence of bicarbonate ions with regard to minimising the scavenging of hydroxyl radicals. Consequently, when both chloride and bicarbonate are present (particularly at elevated levels as in the present study) an appropriate pH must be selected to optimise performance. In this study, of the pH values investigated, pH 5 proved
652
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
(a)
(b)
(c)
Fig. 2. (a) DOC, (b) colour and (c) A254, reduction at various pH values and UV fluences.
the most appropriate. The impact of the addition of NaCl on the degradation of the organic content of ROC using UVC/H2O2 treatment was investigated in an earlier study by Liu et al. [26]. The addition of NaCl (up to 4-fold the initial value) did not significantly impact the reduction of organic content. 3.2. Effect of alum coagulation on the UVC/H2O2 treatment The effect of alum pre-treatment was investigated with a view to improving the overall reduction of organic matter by comparing the treatment efficiency with and without coagulation at various irradiation times. Based on the results obtained in our previous work [27], coagulation was carried out at pH 5 using 1.5 and 3 mmol L1 Al3+. Coagulation removed a significant proportion of the DOC, i.e., 34% and 38% at 1.5 mmol L1 and 3 mmol L1 Al3+, respectively, (Fig. 3a). The reduction of colour and A254 was 50% and 47% at 1.5 mmol L1 Al3+, increasing to 66% and 54%, respectively, at 3 mmol L1 Al3+. Reduction of DOC after coagulation at both Al3+ dosages increased gradually with increasing UV irradiation with an overall reduction of 61–62% for UV fluence of 48 103 mJ/cm2 , i.e., 25– 28% greater than after UVC/H2O2 only treatment. The reduction of DOC between 72 h (36 103 mJ/cm2) and 96 h (48 103 mJ/cm2) irradiation of the 1.5 mmol L1 Al3+ pre-treated sample was the greatest for any 24 h irradiation period for two experimental runs (Fig. 3a). Hence the effect of prolonged irradiation of the sample pre-treated with 1.5 mmol L1 Al3+ was investigated. However, increasing the irradiation time to 182 h (91 103 mJ/cm2) led to only an additional 5% reduction of DOC (results not shown), showing little benefit of doubling the irradiation time; this decreased impact was attributed to the recalcitrant nature of the remaining organic content. A large reduction of colour (Fig. 3b) and A254 (Fig. 3c) was noted during all treatments. Coagulation improved the reduction of
colour and A254, and the pre-treatment with the higher alum concentration showed greater reductions for UV fluence of 12 103 mJ/cm2, but this difference gradually reduced with increasing UV fluence to give comparable final reductions after UV fluence of 48 103 mJ/cm2. Similarly, a small difference (11%) in the SUVA was observed after the two alum doses but it reduced on irradiation of the sample and comparable values (0.51–0.53) were obtained after UV fluence of 36 103 mJ/cm2. 3.3. Kinetics The reduction of DOC, colour and A254 were modelled for UV fluence of 36 103 mJ/cm2 as first order kinetics according to Eq. (1):
ln ½C t =C 0 ¼ kt
ð1Þ 1
where k is the reaction rate constant (min ) and C0 and Ct are the concentrations at irradiation time 0 and t, respectively. By plotting ln (Ct)/(C0) versus time (t), the values of k were obtained and are given in Table 2. The reduction of DOC, COD, colour and A254 followed a pseudo first-order reaction for UV fluence of 36 103 mJ/cm2. The data correlated well with high R2 values (P0.95). The rates of DOC, colour and A254 reduction were lower after pre-treatment due to the removal of a considerable fraction of the large molecular weight compounds (humics) that were preferentially targeted by the UV-LED treatment alone. A fairly similar rate of colour and A254 reduction was noted for both 1.5 and 3 mmol L1 Al3+ pre-treated samples. 3.4. Fluorescence excitation–emission matrix spectra Fluorescence EEM spectra provide a ‘‘fingerprint’’ of the types of organics in water and wastewater [21]. The impact of UVC/H2O2 treatment with and without coagulation was investigated using
653
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
(a)
(b)
(c)
Fig. 3. Comparison of (a) DOC, (b) colour and (c) A254 reduction on irradiation with and without coagulation.
Table 2 Constants to model the loss of DOC, colour and A254 for UV fluence of 36 103 mJ/cm2 treatment with and without coagulation. Treatment
Parameter
k (min1)
R2
UV-LED only
DOC Colour A254
0.0092 0.055 0.032
0.99 0.95 0.98
1.5 mM Al + UV/H2O2
DOC Colour A254
0.0054 0.036 0.017
0.98 0.98 0.99
3 mM Al + UV/H2O2
DOC Colour A254
0.0051 0.038 0.015
0.99 0.96 0.99
the FRI technique [21]. The DOC of each sample was adjusted to 10 mg/L to avoid the inherent quenching effect. The EEM volumes obtained for the raw samples showed humic matter as two large peaks: humic acid (HA)-like matter and fulvic acid (FA)-like matter, soluble microbial products (SMPs) and aromatic proteins (AP I and AP II). A comparison of EEM volumes of ROC before and after UVC/H2O2 only treatment is shown in Fig. 4a. A gradual decrease in all the regions was observed. The reduction of aromatic proteins was fairly similar at both pH values for UV fluence of 48 103 mJ/ cm2 whereas the reduction of HA-like (74%), FA-like (68%) and SMPs (88%) was marginally greater at pH 5 than at pH 8.3, i.e., 67%, 61% and 83%, respectively, which was consistent with the reductions in colour, A254 and SUVA. Although large reductions of colour (50%) and A254 (47%) occurred after coagulation, an increase in the EEM volumes in most of the fractions was observed (Fig. 4b). These were attributed to complexes formed between anionic humic and cationic coagulant species at low pH as shown in our previous work [24]. As the
coagulation was carried out at pH 5, the resulting impact (in this case an increase) on fluorescence was expected. Hence there was a small increase in most fractions after coagulation. However, faster and greater reductions in all the regions were observed when the pre-treated sample was subjected to UVC/H2O2 treatment which is consistent with the reductions in colour, A254 and SUVA. For example, the reduction in the fluorescence of the HA-like and FA-like compounds was markedly greater after coagulation, i.e., 35–44% without pre-treatment compared with 88–85% after pre-treatment for UV fluence of 36 103 mJ/cm2, respectively. Although this difference reduced with increasing fluence to 48 103 mJ/cm2, it was still larger, i.e., 22–28% than for standalone UVC/H2O2 treatment.
3.5. Changes in molecular weight The impact on the molecular size, and thus apparent molecular weight, of the organic components of the ROC was investigated using LC-OCD which separates DOC into five different chromatographic fractions: biopolymers (>20,000 Da), high MW humic substances (1000–20,000 Da), building blocks (350–500 Da), low MW (LMW) acids and humic substances (<350 Da) and LMW neutrals (<350 Da) [28]. Coagulation removed a large proportion (55%) of the humic substances, which tend to be hydrophobic high molecular weight compounds, followed by building blocks (34%). The concentration of biopolymers was very low compared with the other fractions with some reduction (23%) after coagulation, but no impact on LMW neutrals was observed (Fig. 5). Increasing irradiation without pre-treatment led to increasing reduction of humics with corresponding increase in the concentration of building blocks and LMW neutrals. For example, the reduction of humics was 26% for
654
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
(b)
(a)
Fig. 4. EEM volumes of (a) raw and UVC/H2O2 treated ROC without and (b) with pre-treatment (1.5 mM Al3+).
Fig. 5. DOC concentration of various fractions according to LC-OCD after various treatments (the data were derived from the LC-OCD chromatograms).
UV fluence of 12 103 mJ/cm2 and increased to 58% when UV fluence was increased to 36 103 mJ/cm2. The reduction of biopolymers was low and little influenced by increasing UV fluence, i.e., 11% after 12 103 mJ/cm2 cf. 15% after 36 103 mJ/cm2. The reduction of biopolymers (42%) and humics (78%) was greater for the coagulated ROC treated by 72 h UVC/H2O2 than stand-alone coagulation or UVC/H2O2 treatment. Coagulation followed by UVC/H2O2 treatment exhibited a complementary effect for the reduction of humics, however, a net increase in building blocks and effectively no net change in LMW neutrals was observed.
removed by this process. A greater than 4-fold increase in biodegradability was observed when the coagulated ROC was treated by UVC/H2O2 (36 103 mJ/cm2) leading to a residual DOC of 11 mg/L, corresponding to an overall reduction of 67%. The improvement in biodegradability after UVC/H2O2 treatment with and without pre-treatment demonstrates the potential for DOC removal by a subsequent biological process contributing to overall DOC reduction.
3.6. Effect on biodegradability
UVC/H2O2 treatment utilises highly oxidising hydroxyl radicals (HO) which attack organic molecules such as humic-like compounds in a non-specific manner, breaking them down to lower MW intermediates and compounds such as carboxylic acids and aldehydes and eventually converting some of them to carbon dioxide. These intermediates are more biodegradable than the humic-like parent compounds and act as a major food source for microbes [29]. Thus there was a reduction in DOC of 27% after UVC/H2O2 treatment which was increased to 40% after removal of the biodegradable compounds by agency of the BDOC assay (Table 4). Unlike our previous findings with a conventional UVC lamp [24], reduction of pH was required to effectively degrade the organic matter and so increase the biodegradability. However, the potential of UVC-LEDs as alternative sources of UV radiation for this process was demonstrated. Coagulation is an effective pre-treatment for the high salinity ROC as it removed a large proportion of the humics which corresponded with the decrease in DOC, colour, A254 and fluorescence. The humic-like compounds contain acidic phenolic and carboxylic
The generation of LMW compounds during UVC/H2O2 treatment has been reported in our earlier work using a conventional UVC lamp [24]. These LMW compounds are amenable to removal by downstream biological treatment. The biodegradability of the ROC was more than doubled after exposure to a UV fluence of 36 103 mJ/cm2 (Table 3). After coagulation the biodegradability was decreased indicating that some biodegradable matter was
Table 3 Biodegradability after various treatments.*
*
Treatment
Biodegradability (%)
BDOC Al + BDOC UV/H2O2 + BDOC Al + UV/H2O2 + BDOC
7 4 16 33
Al refers to coagulation by alum; UV fluence was 36 103 mJ/cm2.
4. Discussion
655
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656 Table 4 Overall reductions in DOC of ROC after various treatments.*
*
Table 5 EE/O for residual DOC of 15 mg/L.
Treatment
% reduction
Treatment
EE/O (kWh/m3)
BDOC UVC/H2O2 Al UVC/H2O2 + BDOC Al + BDOC Al + UVC/H2O2 Al + UVC/H2O2 + BDOC
7 27 34 40 39 52 67
UVC/H2O2 only UVC/H2O2 + BAC Coagulation + UVC/H2O2 Coagulation + UVC/H2O2 + BDOC
48 37 20 6
Al refers to coagulation by alum; UV fluence was 36 103 mJ/cm2.
groups which provide most of the negative charge carried by the molecule [30]. The mechanism for the removal of humics at pH 5 includes: (a) binding of metal species to anionic sites leading to charge neutralization and reduced solubility and precipitation of the metal–humic complex and (b) adsorption of humics on amorphous metal hydroxide precipitate [31]. DOC removal by coagulation at 3 mmol L1 Al3+ was only slightly better than at 1.5 mmol L1 Al3+ (33% cf. 38%), hence the latter was considered the most effective in terms of DOC removed for alum dosed and so used for the subsequent experiments. Coagulation with alum decreased the humics to a level similar to that after 72 h UVC/H2O2 treatment, and decreased the level of building blocks, but had no impact on the LMW neutrals. Although the reduction of humics was comparable after the stand-alone coagulation and UVC/H2O2 treatments, the greater reduction of colour and A254 after UVC/H2O2 treatment was attributed to the loss of chromophores (some of which would also be fluorophores) through breakdown of the humic-like compounds compared with their removal by coagulation. There was little further reduction of DOC (5%) after biological treatment of the coagulated ROC (Table 3). Coagulation with 1.5 mmol L1 Al3+ followed by UVC/H2O2 treatment led to greater organic matter removal (52%) than by coagulation alone with 1.5 mmol L1 Al3+ (33%). There was a complementary effect for the sequential process as shown by the markedly greater reduction of humic substances (Fig. 5) which corresponded with colour, A254, fluorescence and DOC reductions, however, a net increase in building blocks and no change in LMW neutrals was observed (Fig. 5). Biological treatment (as BDOC assay) removed many of those smaller MW compounds as indicated by the improved total organic content removal of 67%. Coagulation has been reported to improve UV transmittance [27], leading to improved UV penetration and enhanced breakdown of the remaining organic content and therefore greater generation of the biodegradable lower MW compounds. Thus when coagulation is used as a pre-treatment, enhanced reduction of the organic content can be achieved using UV/H2O2 followed by biological treatment as demonstrated in this study, the enhancement was due to the complementary effect of the treatments which led to the removal of a wide range of organic compounds.
study, the energy requirement would be 1 kWh/m3 which is insignificant compared with that needed for generation of UV radiation. The EE/O was calculated for a target DOC residual of 15 mg/L (Table 5). The EE/O needed to achieve the desired residual was high (48 kWh/m3) using UVC/H2O2 only treatment but was reduced by 23% after biological treatment (as BDOC assay). Pre-treatment by coagulation led to a markedly lower EE/O which was further reduced after biological treatment, an overall decrease of 87% compared with stand-alone UVC/H2O2 treatment. The EE/O was 86% lower than for UVC/H2O2 treatment using a conventional UV lamp during treatment of ROC of comparable initial DOC concentration [27]. These EE/O estimates are preliminary as coagulation leads to the production of sludge which requires appropriate processing, increasing the energy requirement and cost of the treatment. However, taking into account that coagulation is generally used as a pre-treatment in municipal WWTPs, the sludge produced during ROC pre-treatment can be handled collectively where appropriate. 6. Future of UV-LED technology Considering that the current UV-LEDs have low quantum efficiency (up to 8%) and low output power (0.3 mW), ongoing developments are expected to significantly improve these parameters which would increase the energy efficiency of the UV-LEDs. UV-LEDs (255 nm) up to 50 mW have recently been developed but are very expensive. The improvements in the efficiency of UV-LEDs are projected to follow the visible LED trajectories which can operate at 75% efficiency [37,38] and their efficiency is projected to match that of visible LEDs by 2020 [13]. These improvements are happening along with an increase in the lifespan of the UV-LEDs. In 2007, they had a lifetime of 200 h which increased to 10,000 h in 2012 and is projected to increase to 100,000 h (11 years) by 2020 [13]. The UV-LEDs used in this study have been operated for more than 5000 h and are expected to be operational up to 10,000 h. Autin et al. [11] calculated that the whole of life cost of UV-LEDs was 150 times greater than the traditional low pressure UV system. However, it was projected to reduce to 40% by 2020 provided the UV-LED technology follows the projected trends. Although the widespread commercial use of UV-LEDs is years away, the developments over the last 5 years are promising and the continuing improvements in their output power and energy efficiency are expected to make this technology a competitive alternative to traditional UV systems.
5. Estimation of electrical energy per order (EE/O) 7. Conclusions A useful tool for measuring the electrical efficiency of the UV-based processes is electrical energy per order (EE/O) developed by Bolton et al. [32]. The EE/O is defined as the electrical energy (in kilowatt hours) required to reduce the concentration of a pollutant by one order of magnitude in 1000 L of water. As coagulation is considered to be a low energy process [33,34] and the energy required for biological treatment is negligible [35] compared with the AOPs, only the energy required for UVC/H2O2 treatment was considered. The energy needed for the production of 1 kg H2O2 is 10 kWh [36], hence for the concentration (3 mmol L1) used in this
The potential of the UVC/H2O2 process using UVC-LEDs (255 nm) for the degradation of the organic content of ROC, a concentrated complex wastewater of very high salinity, was demonstrated. Greater reductions of DOC, colour, A254 and SUVA were obtained by reducing the pH to 5 during UVC/H2O2 only treatment. Coagulation using alum at pH 5 removed a significant proportion of the organic matter (34–38%) and led to enhanced overall reduction after UVC/H2O2 treatment. The UVC/H2O2 treatment increased the biodegradability, and using biological treatment (as BDOC assay)
656
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
after the sequence of coagulation and UVC/H2O2 treatment (UV fluence 36 103 mJ/cm2) led to an overall DOC reduction of 67%, demonstrating the effectiveness of the proposed treatment scheme. Although the low power output of the UV-LEDs and high cost are the main limitations for their application in water and wastewater treatment, the ongoing developments in this technology, combined with improved reactor design, are expected to improve the viability of their use. Acknowledgements We would like to acknowledge the partial financial support provided by Water Research Australia and the Smart Water Fund, Victoria, Australia. References [1] B. Van der Bruggen, L. Lejon, C. Vandecasteele, Reuse, treatment, and discharge of the concentrate of pressure-driven membrane processes, Environ. Sci. Technol. 37 (2003) 3733–3738. [2] F. Tang, H.Y. Hu, Q.Y. Wu, X. Tang, X.Y. Sun, X.L. Shi, J.J. Huang, Effects of chemical agent injections on genotoxicity of wastewater in a microfiltrationreverse osmosis membrane process for wastewater reuse, J. Hazard. Mater. 260 (2013) 231–237. [3] C. Vargas, A. Buchanan, Monitoring ecotoxicity and nutrients load in the reverse osmosis concentrate from Bundamba Advanced Water Treatment Plant, Queensland, Australia, Water Pract. Technol. 6 (2011). [4] S.J. Khan, D. Murchland, M. Rhodes, T.D. Waite, Management of concentrated waste streams from high-pressure membrane water treatment systems, Crit. Rev. Environ. Sci. Technol. 39 (2009) 367–415. [5] W.R. Haag, C.C.D. Yao, Rate constants for reaction of hydroxyl radicals with several drinking water contaminants, Environ. Sci. Technol. 26 (1992) 1005– 1013. [6] S. Sarathy, M. Mohseni, Effects of UV/H2O2 advanced oxidation on chemical characteristics and chlorine reactivity of surface water natural organic matter, Water Res. 44 (2010) 4087–4096. [7] J. Malley, UV in Water Treatment – Issues from the Next Decade, International Ultraviolet Association News 12 (2010) 18–25. [8] M.A. Würtele, T. Kolbe, M. Lipsz, A. Külberg, M. Weyers, M. Kneissl, M. Jekel, Application of GaN-based ultraviolet-C light emitting diodes – UV LEDs – for water disinfection, Water Res. 45 (2011) 1481–1489. [9] H.W. Chen, Y. Ku, A. Irawan, Photodecomposition of o-cresol by UV-LED/TiO2 process with controlled periodic illumination, Chemosphere 69 (2007) 184– 190. [10] Y. Taniyasu, M. Kasu, T. Makimoto, An aluminium nitride light-emitting diode with a wavelength of 210 nanometres, Nature 441 (2006) 325–328. [11] S. Vilhunen, J. Puton, J. Virkutyte, M. Sillanpää, Efficiency of hydroxyl radical formation and phenol decomposition using UV light emitting diodes and H2O2, Environ. Technol. 32 (2011) 865–872. [12] A. Jamali, R. Vanraes, P. Hanselaer, T. Van Gerven, A batch LED reactor for the photocatalytic degradation of phenol, Chem. Eng. Process. 71 (2013) 43–50. [13] J.L. Shie, C.H. Lee, C.S. Chiou, C.T. Chang, C.C. Chang, C.Y. Chang, Photodegradation kinetics of formaldehyde using light sources of UVA, UVC and UV LED in the presence of composed silver titanium oxide photocatalyst, J. Hazard. Mater. 155 (2008) 164–172. [14] R.J. Tayade, T.S. Natarajan, H.C. Bajaj, Photocatalytic degradation of methylene blue dye using ultraviolet light emitting diodes, Ind. Eng. Chem. Res. 48 (2009) 10262–10267.
[15] O. Autin, C. Romelot, L. Rust, J. Hart, P. Jarvis, J. MacAdam, S.A. Parsons, B. Jefferson, Evaluation of a UV-light emitting diodes unit for the removal of micropollutants in water for low energy advanced oxidation processes, Chemosphere 92 (2013) 745–751. [16] W.Y. Wang, Y. Ku, Photocatalytic degradation of Reactive Red 22 in aqueous solution by UV-LED radiation, Water Res. 40 (2006) 2249–2258. [17] M.A.S. Ibrahim, Commercial evaluation of UV-LED in water treatment applications (MSc. thesis), Cranfield University, UK, 2012. [18] R.V. Steele, The story of a new light source, Nat. Photonics 1 (2007) 25–26. [19] APHA, Standard Methods for the Examination of Water & Wastewater, American Public Health Association, Washington DC, USA, 2005. [20] S. Jin, A.A. Mofidi, K.G. Linden, Polychromatic UV fluence measurement using chemical actinometry, biodosimetry, and mathematical techniques, J. Environ. Eng. 132 (2006) 831–841. [21] W. Chen, P. Westerhoff, J.A. Leenheer, K. Booksh, Fluorescence excitation– emission matrix regional integration to quantify spectra for dissolved organic matter, Environ. Sci. Technol. 37 (2003) 5701–5710. [22] J.C. Joret, Y. Levi, Methode rapide d’evaluation du carbone eliminable des eaux voie biologique, Tribune du Cebedeau 510 (1986) 3–9. [23] C. Volk, C. Renner, P. Roche, H. Paillard, J.C. Joret, Effects of ozone on the production of biodegradable dissolved organic carbon (BDOC) during water treatment, Ozone Sci. Eng. 15 (1993) 389–404. [24] M. Umar, F. Roddick, L. Fan, Assessing the potential of a UV-based AOP in treating high-salinity municipal wastewater reverse osmosis concentrate, Water Sci. Technol. 68 (2013) 1994–1999. [25] C.H. Liao, S.F. Kang, F.A. Wu, Hydroxyl radical scavenging role of chloride and bicarbonate ions in the H2O2/UV process, Chemosphere 44 (2001) 1193–1200. [26] K. Liu, F.A. Roddick, L. Fan, Impact of salinity and pH on the UVC/H2O2 treatment of reverse osmosis concentrate produced from municipal wastewater reclamation, Water Res. 46 (2012) 3229–3239. [27] M. Umar, F. Roddick, L. Fan, Effect of coagulation on treatment of municipal wastewater reverse osmosis concentrate by UVC/H2O2, J. Hazard. Mater. 266 (2014) 10–18. [28] S.A. Huber, F.H. Frimmel, Size-exclusion chromatography with organic carbon detection (LC-OCD): a fast and reliable method for the characterization of hydrophilic organic carbon, Vom Wasser 86 (1996) 277–290. [29] E.H. Goslan, F. Gurses, J. Banks, S.A. Parsons, An investigation into reservoir NOM reduction by UV photolysis and advanced oxidation processes, Chemosphere 65 (2006) 1113–1119. [30] J. Bratby, Coagulation and Flocculation in Water and Wastewater Treatment, IWA Publishing, London, UK, 2006. [31] J. Duan, J. Gregory, Coagulation by hydrolysing metal salts, Adv. Colloid Interface Sci. 100–102 (2003) 475–502. [32] J.R. Bolton, K.G. Bircher, W. Tumas, C.A. Tolman, Figures-of-merit for the technical development and application of advanced oxidation technologies for both electric-and solar-driven systems (IUPAC Technical Report), Pure Appl. Chem. 73 (2001) 627–638. [33] D. Bromley, M. Gamal El-Din, D.W. Smith, A low cost treatment process to reduce phosphorus and suspended solids in liquid waste from animal farm operations, Penang, Malaysia, 2002. [34] C. Huang, M. Cao, J. Luo, C. Zhang, Study of chemically enhanced primary treatment of domestic sewage. International Conference on Agriculture and Natural Resources Engineering, Adv. Biomed. Eng. 3–5 (2011) 270–273. [35] A. Marco, S. Esplugas, G. Saum, How and why combine chemical and biological processes for wastewater treatment, Water Sci. Technol. 35 (1997) 321–327. [36] E.J. Rosenfeldt, K.G. Linden, S. Canonica, U. von Gunten, Comparison of the efficiency of OH radical formation during ozonation and the advanced oxidation processes O3/H2O2 and UV/H2O2, Water Res. 40 (2006) 3695–3704. [37] T. Bettles, S. Schujman, J.A. Smart, W. Liu, L. Schowalter, UV light emitting diodes – their applications and benefits, International Ultraviolet Association Conference, Los Angeles, CA, 2007, USA. [38] C. Chatterley, K. Linden, Demonstration and evaluation of germicidal UV-LEDs for point-of-use water disinfection, J. Water Health 8 (2010) 479–486.
Contents lists available at ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Treatment of municipal wastewater reverse osmosis concentrate using UVC-LED/H2O2 with and without coagulation pre-treatment M. Umar a, F.A. Roddick a,⇑, L. Fan a, O. Autin b, B. Jefferson b a b
School of Civil, Environmental and Chemical Engineering, RMIT University, GPO Box 2476, Melbourne, 3001 Victoria, Australia Cranfield Water Sciences Institute, Cranfield University, Cranfield, Bedfordshire MK43 0AL, United Kingdom
h i g h l i g h t s UVC/H2O2 treatment with UV-LEDs can degrade organic matter in RO concentrate. UVC/H2O2 treatment increased the biodegradability of RO concentrate. UV-LEDs were markedly more energy efficient than a conventional mercury UV lamp. Alum coagulation plus UVC/H2O2 treatment led to complementary effect on DOC removal.
a r t i c l e
i n f o
Article history: Received 8 July 2014 Received in revised form 2 September 2014 Accepted 7 September 2014 Available online 16 September 2014 Keywords: UVC/H2O2 UV-LED Reverse osmosis concentrate Alum Coagulation Biodegradability
a b s t r a c t The potential of a prototype batch reactor using ultraviolet light emitting diodes (UVC-LEDs) which emit at 255 nm in conjunction with H2O2 for the treatment of a highly saline (electrical conductivity 22 mS/cm; DOC 32–37.5 mg/L) municipal wastewater reverse osmosis concentrate was investigated. Mineralization of organic content (measured as DOC) was low (22%) due to the low fluence rate (0.14 mW/cm2), however, a large reduction in colour (94%) and A254 (75%) occurred after delivering a UV fluence of 48 103 mJ/cm2 at the original pH of 8.3. Fairly similar results were obtained at pH 7, but the reduction of DOC increased at lower pH with 38% and 36% achieved at pH 4 and 5, respectively. Similar trends were observed for colour and A254 reduction. These results, in conjunction with excitation– emission matrix spectra, biological dissolved organic carbon (BDOC) assay and apparent molecular size distribution, demonstrated that the prototype system led to the breakdown of the chromophore bonds and thus changes in the molecular structure, and degradation of high molecular weight (MW) compounds to low MW compounds. Coagulation (1.5 and 3 mmol L1 Al3+ at pH 5) led to a significant reduction of DOC (34–38%), colour (50–66%) and A254 (47–54%), and subsequent UVC/H2O2 treatment led to further reduction in these parameters. For a target DOC reduction of 15 mg/L, the EE/O was 15 kWh/m3 when coagulation was used as pre-treatment to the UVC/H2O2 treatment (UV fluence 36 103 mJ/cm2) and it reduced to less than half after biological treatment (as BDOC assay). This study demonstrated the potential of UV-LEDs as an alternative UV source for degrading the organic matter in ROC using advanced oxidation. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Reverse osmosis (RO) is widely used as a polishing treatment for secondary effluent in wastewater reclamation schemes. However the resultant RO concentrate (ROC), which comprises 15–20% of the volume of the feed stream, contains almost all the contaminants present in the original wastewater at elevated levels. Depending on the wastewater source, these contaminants may be ⇑ Corresponding author. Tel.: +61 3 9925 2080; fax: +61 3 9639 0138. E-mail address: [email protected] (F.A. Roddick). http://dx.doi.org/10.1016/j.cej.2014.09.028 1385-8947/Ó 2014 Elsevier B.V. All rights reserved.
toxic and/or bio-accumulative, and so disposal of untreated ROC presents a potential environmental risk. The addition of chemicals such as antiscalants, biocides and acids during the treatment process further complicates the situation as they can change the character of the organic and inorganic pollutants and can influence the chemical equilibrium of the dissolved constituents [1]. Tang et al. [2] demonstrated that the genotoxicity of a ROC was 500–559 lg 4-NQO (4-nitroquinoline-1-oxide)/L, which was much higher than for the RO influent (105–160 lg 4-NQO/L). As such, it is mandatory for some facilities to treat ROC prior to its discharge. An example is the Bundamba advanced wastewater treatment plant (Brisbane,
650
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
Australia), which is required treat and monitor the ROC prior to its discharge to the Brisbane River [3]. In addition to minimizing its environmental impacts, beneficial reuse of the treated ROC can help to offset the treatment costs [4]. The application of ultraviolet (UV) irradiation-based advanced oxidation processes (AOPs) has been extensively investigated for the degradation of a wide range of contaminants present in water and wastewater. An AOP involves the in situ generation of hydroxyl radicals (HO) which react non-selectively with organic contaminants at very high rates (rate constant of 108–1010 M1 s1) [5]. UV plus hydrogen peroxide (UV/H2O2) is one of the most promising AOPs which generates HO through photolysis of H2O2. Since the quantum yield of the HO generation due to the direct UV photolysis of H2O2 is near unity and wavelength independent, the process is considered highly efficient and is leading the way in commercial applications [6]. Conventional UV lamps continue to be the main source of radiation, however, alternative UV sources are being sought due to the inherent disadvantages of UV arc lamps such as the use of mercury, low mechanical stability, large size and low energy efficiency. UV light emitting diodes (UV-LEDs) which are a mercury-free source of monochromatic UV radiation are potential alternatives and are regarded as the most promising UV light sources [7]. UV-LEDs are small, compact, robust and durable [8] and offer virtually no limitations on the potential geometry of the emission sources [7]. Furthermore, UV-LEDs are less energy intensive than traditional UV lamps as they convert a greater amount of energy into light (high quantum yield) because the light emission occurs due to the recombination of electrons and holes [9]. Like conventional diodes, UV-LEDs are comprised of a chip of semi-conducting material impregnated or doped with impurities to create a p-n junction capable of emitting light in a narrow wavelength range in the form of electroluminescence [10]. Semiconductor crystals of compounds containing aluminium, nitrogen and gallium are the main materials used in the construction of UV-LEDs [10]. The wavelength of light emitted depends on the band gap energy of the type of material used to construct the UV-LEDs [11] and LEDs with emission wavelengths as low as 210 nm have been developed [10]. Applications of UV-LEDs include water disinfection [8] and degradation of organic compounds such as phenol [12], formaldehyde [13], methylene blue [14], metaldehyde [15] and reactive red 22 dye [16]. Although successfully used for disinfection of water and degradation of individual organic compounds, the potential of UV-LEDs for treating real wastewater is yet to be reported. Furthermore, most of the foregoing studies used high wavelength UV-LEDs (>360 nm) except some [11,15] that used deep UVC-LEDs (<290 nm). Although low output power and high cost are the major limitations at this stage, continuing developments in this technology have been projected to significantly overcome these by 2020 [17]. The trend of development between 2007 and 2012 has followed Haitz’s law which forecasts an increase in output per bulb by a factor of 20 and a decrease in cost by a factor of 10 per decade [18]. The aim of this study was to determine the potential of deep UVC-LEDs in the presence of H2O2 for degrading the organic content of municipal wastewater ROC and its removal after subsequent biological treatment. The removal of organic content by this sequence was compared with that for coagulation (alum), and for when coagulation was used as a pre-treatment to the UVC-LEDs/ H2O2 process to produce a target residual DOC concentration of 15 mg/L. The change in biodegradability was investigated by determining the biological dissolved organic carbon (BDOC) content, and fluorescence excitation-emission matrix (EEM) spectra and liquid chromatography – organic carbon detection (LC-OCD) were used to track the changes in the organic components of the ROC.
2. Materials and methods 2.1. Collection and characterization of ROC The composite ROC sample was collected from a wastewater reclamation facility at a local municipal wastewater treatment plant (WWTP), and stored at 4 °C. In the treatment process at the WWTP, raw sewage is screened and de-gritted and sent to intermittently decanted extended aeration (IDEA) bioreactors where it is treated in a cycle of aeration, settling, and decant. The IDEA effluent is then treated using a combination of ultrafiltration and RO. The sample was characterized (Table 1) according to Standard Methods [19]. The sample was high in alkalinity, salinity and total dissolved solids (TDS) concentration. 2.2. UV-LED reactor UV-LEDs with output wavelength of 255 nm were purchased from Sensor Electronic Technology (Columbia, South Carolina, USA). According to the manufacturer’s specifications, each LED had an emission power of approximately 0.33 mW at 20 mA. The LEDs were soldered to a plate and connected in two series, each containing 5 units. The distance between each LED unit was 17 mm and between the sample surface and LEDs was 7 mm, based on the findings of Autin et al. [15]. A DC power supply (INSTEK GPR-6030 D, 0–60 V, 0–3 A) was utilised. A schematic representation of the experimental set up is shown in Fig. 1. Tests were performed in batch mode using 50 mL of ROC sample placed in a rectangular Perspex vessel which was continuously stirred using a magnetic stirrer; the liquid depth was 7 mm. The average fluence rate of the system was 0.14 mW/cm2, as determined by uridine actinometry [20]. As the treatment time was extended, some evaporation of the sample was observed, consequently the results were adjusted by taking into account the difference in volume after various irradiation times. All experiments were performed at room temperature (22 °C). The temperature was regularly monitored and only a small increase (2–3 °C) in the temperature of the ROC was noted during operation. Preliminary tests using a range of 2–6 mmol L1 H2O2 demonstrated that 3 mmol L1 was a good compromise taking into account the chemical consumption and DOC reduction and therefore was chosen to be used in this study. All experiments were performed in duplicate, and in some cases in triplicate when the difference was more than 5%. Average values of duplicate or triplicate runs are reported, and error bars represent the standard deviation for these results. 2.3. Alum coagulation A stock solution of alum was prepared using Al2(SO4)318H2O (Chem-Supply, Pty Ltd, Australia). Coagulation was conducted with
Table 1 Characteristics of ROC. Parameter
Value
DOC (mg/L) COD (mg/L) pH Colour (Pt.Co) Chloride (mg/L) TDS (mg/L) A254 (1/cm) SUVA (L/mg-m) Alkalinity (as CaCO3, mg/L) Conductivity (mS/cm)
32–37 105 8.3 158 8060 16,140 0.68 1.8 410 22.3
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
651
Fig. 1. Schematic representation of the experimental set-up.
a laboratory jar test apparatus (Phipps and Bird, PB-700) using 2 L ROC sample. The samples were rapidly mixed for 2 min at 250 rpm followed by slow mixing for 30 min at 30 rpm and subsequent settling for 2 h before taking supernatant for analyses. The pH was adjusted using 1 mol L1 H2SO4 or 1 mol L1 NaOH. 2.4. Analytical methods DOC was determined using a total organic carbon analyser (Sievers model 5310C) in in-line mode to purge inorganic carbon, acid/ oxidiser ratio 0.8 and after dilution of the ROC. The DOC of some samples was confirmed using a Shimadzu TOC analyser (TOC-L series) without any dilution and the difference was found to be less than 3–5%. Absorbance was determined using a double beam scanning UV/vis spectrophotometer (Unicam UV2). True colour of the samples was measured with a Hach DR 4000 spectrophotometer at 455 nm in Platinum Cobalt (Pt.Co) units. Fluorescence EEM spectra were determined with a Perkin Elmer LS-50B luminescence spectrometer. The changes in fluorescence were quantitatively analysed using the fluorescence regional integration (FRI) technique which integrates the volumes under the EEM regions [21]. The biodegradability of the organics was evaluated as BDOC using the assay introduced by Joret and Levi [22] and modified by Volk et al. [23]. The DOC was measured daily and the BDOC was determined as the difference between the initial and the lowest DOC recorded over a 5 day period. Removal of residual H2O2 to reduce the impact on biological activity prior to the BDOC assay was found unnecessary. All analyses were conducted in duplicate and average results are reported. The concentration of residual H2O2 was estimated using MerckoquantÒ peroxide test strips. Apparent molecular size distribution was determined using liquid chromatography with organic carbon detection (LC-OCD) at the Water Research Centre of the University of New South Wales (Sydney, Australia) with a DOCLabor LC-OCD Model 8, with a Toyopearl TSK HW-50S column, using a phosphate buffer of pH 6.4 as the mobile phase. The samples were diluted to reduce the DOC and TDS to levels suitable for the analysis. 3. Results 3.1. Treatment efficiency of UVC/H2O2 and the effect of pH Treatment of ROC was carried out at the original pH of 8.3 and at pH 7, 5 and 4 to compare the treatment efficiency in terms of DOC, colour, A254 and specific ultraviolet absorbance (SUVA). For UV fluence of 48 103 mJ/cm2 (96 h exposure) at the original pH, the reduction of DOC was 22%, and reductions in colour and A254
were 94% and 75%, respectively. Little difference in the reduction of DOC was observed at pH 7 and 8.3 (Fig. 2a). Given the low intensity of UV irradiation of the UV-LEDs, the production of hydroxyl radicals was expected to be low. One way of improving the system is to minimize the concentration of the 2 HO scavengers (HCO 3 /CO3 ) by lowering the pH. Reducing the pH to 5 resulted in a greater reduction of DOC, i.e., 36% for UV fluence of 48 103 mJ/cm2; a marginally greater reduction of DOC was obtained at pH 4 (Fig. 2a). The reduction of colour was greater and faster at pH 4 than all other pH values, however the final reductions were comparable (Fig. 2b). This comparable reduction of colour (93–97%) at all pH values demonstrates that the disintegration of chromophore bonds was little influenced by alkalinity. Consistent with the trend for colour, the reduction of A254 was greater under acidic conditions and although more rapid at pH 4 than pH 5, final reductions were fairly similar (80–84%) (Fig. 2c). Little difference in the reduction of colour and A254 was noted for pH 7 and pH 8.3 at all UV fluence values. The trends for SUVA (decreased from 1.8 to 0.5–0.8 after UV fluence of 48 103 mJ/cm2) were expected given those for DOC and A254, indicating the loss of aromatic structures but incomplete disintegration of the molecules and hence low mineralization. The high reductions of colour and A254 and corresponding low reduction of DOC were attributed to the partial breakdown of large molecules (humic-like compounds) to low MW and/or non-coloured compounds and were consistent with our previous findings using a conventional UV lamp system [24]. It should be noted that adequate H2O2 was present for UV fluence of 48 103 mJ/cm2 (residual H2O2 > 25 mg/L, i.e., 0.73 mmol L1) to ensure continuous generation of HO, sufficient to partially oxidize the organic content but not enough to bring about major molecular changes or mineralize the humic-like compounds, leaving major fragments or bulk of the original molecules intact. As there was little difference in the results for pH 4 and pH 5 and the acid and subsequent alkali addition required if the lower pH were used, pH 5 was selected for use in further experiments. A control test was performed comprising the addition of H2O2 to ROC at pH 5 in the absence of irradiation. Some reduction of colour (5%) but no reduction of DOC or A254 was observed over 96 h. Similarly, little reduction of colour (4–5%) and no change in DOC and A254 value was noted in a control test performed at pH 5 without the addition of H2O2. Chloride ions are major scavengers of hydroxyl radicals, particularly in the absence of bicarbonate ions. As noted by Liao et al. [25] pH is more important than chloride concentration in the presence of bicarbonate ions with regard to minimising the scavenging of hydroxyl radicals. Consequently, when both chloride and bicarbonate are present (particularly at elevated levels as in the present study) an appropriate pH must be selected to optimise performance. In this study, of the pH values investigated, pH 5 proved
652
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
(a)
(b)
(c)
Fig. 2. (a) DOC, (b) colour and (c) A254, reduction at various pH values and UV fluences.
the most appropriate. The impact of the addition of NaCl on the degradation of the organic content of ROC using UVC/H2O2 treatment was investigated in an earlier study by Liu et al. [26]. The addition of NaCl (up to 4-fold the initial value) did not significantly impact the reduction of organic content. 3.2. Effect of alum coagulation on the UVC/H2O2 treatment The effect of alum pre-treatment was investigated with a view to improving the overall reduction of organic matter by comparing the treatment efficiency with and without coagulation at various irradiation times. Based on the results obtained in our previous work [27], coagulation was carried out at pH 5 using 1.5 and 3 mmol L1 Al3+. Coagulation removed a significant proportion of the DOC, i.e., 34% and 38% at 1.5 mmol L1 and 3 mmol L1 Al3+, respectively, (Fig. 3a). The reduction of colour and A254 was 50% and 47% at 1.5 mmol L1 Al3+, increasing to 66% and 54%, respectively, at 3 mmol L1 Al3+. Reduction of DOC after coagulation at both Al3+ dosages increased gradually with increasing UV irradiation with an overall reduction of 61–62% for UV fluence of 48 103 mJ/cm2 , i.e., 25– 28% greater than after UVC/H2O2 only treatment. The reduction of DOC between 72 h (36 103 mJ/cm2) and 96 h (48 103 mJ/cm2) irradiation of the 1.5 mmol L1 Al3+ pre-treated sample was the greatest for any 24 h irradiation period for two experimental runs (Fig. 3a). Hence the effect of prolonged irradiation of the sample pre-treated with 1.5 mmol L1 Al3+ was investigated. However, increasing the irradiation time to 182 h (91 103 mJ/cm2) led to only an additional 5% reduction of DOC (results not shown), showing little benefit of doubling the irradiation time; this decreased impact was attributed to the recalcitrant nature of the remaining organic content. A large reduction of colour (Fig. 3b) and A254 (Fig. 3c) was noted during all treatments. Coagulation improved the reduction of
colour and A254, and the pre-treatment with the higher alum concentration showed greater reductions for UV fluence of 12 103 mJ/cm2, but this difference gradually reduced with increasing UV fluence to give comparable final reductions after UV fluence of 48 103 mJ/cm2. Similarly, a small difference (11%) in the SUVA was observed after the two alum doses but it reduced on irradiation of the sample and comparable values (0.51–0.53) were obtained after UV fluence of 36 103 mJ/cm2. 3.3. Kinetics The reduction of DOC, colour and A254 were modelled for UV fluence of 36 103 mJ/cm2 as first order kinetics according to Eq. (1):
ln ½C t =C 0 ¼ kt
ð1Þ 1
where k is the reaction rate constant (min ) and C0 and Ct are the concentrations at irradiation time 0 and t, respectively. By plotting ln (Ct)/(C0) versus time (t), the values of k were obtained and are given in Table 2. The reduction of DOC, COD, colour and A254 followed a pseudo first-order reaction for UV fluence of 36 103 mJ/cm2. The data correlated well with high R2 values (P0.95). The rates of DOC, colour and A254 reduction were lower after pre-treatment due to the removal of a considerable fraction of the large molecular weight compounds (humics) that were preferentially targeted by the UV-LED treatment alone. A fairly similar rate of colour and A254 reduction was noted for both 1.5 and 3 mmol L1 Al3+ pre-treated samples. 3.4. Fluorescence excitation–emission matrix spectra Fluorescence EEM spectra provide a ‘‘fingerprint’’ of the types of organics in water and wastewater [21]. The impact of UVC/H2O2 treatment with and without coagulation was investigated using
653
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
(a)
(b)
(c)
Fig. 3. Comparison of (a) DOC, (b) colour and (c) A254 reduction on irradiation with and without coagulation.
Table 2 Constants to model the loss of DOC, colour and A254 for UV fluence of 36 103 mJ/cm2 treatment with and without coagulation. Treatment
Parameter
k (min1)
R2
UV-LED only
DOC Colour A254
0.0092 0.055 0.032
0.99 0.95 0.98
1.5 mM Al + UV/H2O2
DOC Colour A254
0.0054 0.036 0.017
0.98 0.98 0.99
3 mM Al + UV/H2O2
DOC Colour A254
0.0051 0.038 0.015
0.99 0.96 0.99
the FRI technique [21]. The DOC of each sample was adjusted to 10 mg/L to avoid the inherent quenching effect. The EEM volumes obtained for the raw samples showed humic matter as two large peaks: humic acid (HA)-like matter and fulvic acid (FA)-like matter, soluble microbial products (SMPs) and aromatic proteins (AP I and AP II). A comparison of EEM volumes of ROC before and after UVC/H2O2 only treatment is shown in Fig. 4a. A gradual decrease in all the regions was observed. The reduction of aromatic proteins was fairly similar at both pH values for UV fluence of 48 103 mJ/ cm2 whereas the reduction of HA-like (74%), FA-like (68%) and SMPs (88%) was marginally greater at pH 5 than at pH 8.3, i.e., 67%, 61% and 83%, respectively, which was consistent with the reductions in colour, A254 and SUVA. Although large reductions of colour (50%) and A254 (47%) occurred after coagulation, an increase in the EEM volumes in most of the fractions was observed (Fig. 4b). These were attributed to complexes formed between anionic humic and cationic coagulant species at low pH as shown in our previous work [24]. As the
coagulation was carried out at pH 5, the resulting impact (in this case an increase) on fluorescence was expected. Hence there was a small increase in most fractions after coagulation. However, faster and greater reductions in all the regions were observed when the pre-treated sample was subjected to UVC/H2O2 treatment which is consistent with the reductions in colour, A254 and SUVA. For example, the reduction in the fluorescence of the HA-like and FA-like compounds was markedly greater after coagulation, i.e., 35–44% without pre-treatment compared with 88–85% after pre-treatment for UV fluence of 36 103 mJ/cm2, respectively. Although this difference reduced with increasing fluence to 48 103 mJ/cm2, it was still larger, i.e., 22–28% than for standalone UVC/H2O2 treatment.
3.5. Changes in molecular weight The impact on the molecular size, and thus apparent molecular weight, of the organic components of the ROC was investigated using LC-OCD which separates DOC into five different chromatographic fractions: biopolymers (>20,000 Da), high MW humic substances (1000–20,000 Da), building blocks (350–500 Da), low MW (LMW) acids and humic substances (<350 Da) and LMW neutrals (<350 Da) [28]. Coagulation removed a large proportion (55%) of the humic substances, which tend to be hydrophobic high molecular weight compounds, followed by building blocks (34%). The concentration of biopolymers was very low compared with the other fractions with some reduction (23%) after coagulation, but no impact on LMW neutrals was observed (Fig. 5). Increasing irradiation without pre-treatment led to increasing reduction of humics with corresponding increase in the concentration of building blocks and LMW neutrals. For example, the reduction of humics was 26% for
654
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
(b)
(a)
Fig. 4. EEM volumes of (a) raw and UVC/H2O2 treated ROC without and (b) with pre-treatment (1.5 mM Al3+).
Fig. 5. DOC concentration of various fractions according to LC-OCD after various treatments (the data were derived from the LC-OCD chromatograms).
UV fluence of 12 103 mJ/cm2 and increased to 58% when UV fluence was increased to 36 103 mJ/cm2. The reduction of biopolymers was low and little influenced by increasing UV fluence, i.e., 11% after 12 103 mJ/cm2 cf. 15% after 36 103 mJ/cm2. The reduction of biopolymers (42%) and humics (78%) was greater for the coagulated ROC treated by 72 h UVC/H2O2 than stand-alone coagulation or UVC/H2O2 treatment. Coagulation followed by UVC/H2O2 treatment exhibited a complementary effect for the reduction of humics, however, a net increase in building blocks and effectively no net change in LMW neutrals was observed.
removed by this process. A greater than 4-fold increase in biodegradability was observed when the coagulated ROC was treated by UVC/H2O2 (36 103 mJ/cm2) leading to a residual DOC of 11 mg/L, corresponding to an overall reduction of 67%. The improvement in biodegradability after UVC/H2O2 treatment with and without pre-treatment demonstrates the potential for DOC removal by a subsequent biological process contributing to overall DOC reduction.
3.6. Effect on biodegradability
UVC/H2O2 treatment utilises highly oxidising hydroxyl radicals (HO) which attack organic molecules such as humic-like compounds in a non-specific manner, breaking them down to lower MW intermediates and compounds such as carboxylic acids and aldehydes and eventually converting some of them to carbon dioxide. These intermediates are more biodegradable than the humic-like parent compounds and act as a major food source for microbes [29]. Thus there was a reduction in DOC of 27% after UVC/H2O2 treatment which was increased to 40% after removal of the biodegradable compounds by agency of the BDOC assay (Table 4). Unlike our previous findings with a conventional UVC lamp [24], reduction of pH was required to effectively degrade the organic matter and so increase the biodegradability. However, the potential of UVC-LEDs as alternative sources of UV radiation for this process was demonstrated. Coagulation is an effective pre-treatment for the high salinity ROC as it removed a large proportion of the humics which corresponded with the decrease in DOC, colour, A254 and fluorescence. The humic-like compounds contain acidic phenolic and carboxylic
The generation of LMW compounds during UVC/H2O2 treatment has been reported in our earlier work using a conventional UVC lamp [24]. These LMW compounds are amenable to removal by downstream biological treatment. The biodegradability of the ROC was more than doubled after exposure to a UV fluence of 36 103 mJ/cm2 (Table 3). After coagulation the biodegradability was decreased indicating that some biodegradable matter was
Table 3 Biodegradability after various treatments.*
*
Treatment
Biodegradability (%)
BDOC Al + BDOC UV/H2O2 + BDOC Al + UV/H2O2 + BDOC
7 4 16 33
Al refers to coagulation by alum; UV fluence was 36 103 mJ/cm2.
4. Discussion
655
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656 Table 4 Overall reductions in DOC of ROC after various treatments.*
*
Table 5 EE/O for residual DOC of 15 mg/L.
Treatment
% reduction
Treatment
EE/O (kWh/m3)
BDOC UVC/H2O2 Al UVC/H2O2 + BDOC Al + BDOC Al + UVC/H2O2 Al + UVC/H2O2 + BDOC
7 27 34 40 39 52 67
UVC/H2O2 only UVC/H2O2 + BAC Coagulation + UVC/H2O2 Coagulation + UVC/H2O2 + BDOC
48 37 20 6
Al refers to coagulation by alum; UV fluence was 36 103 mJ/cm2.
groups which provide most of the negative charge carried by the molecule [30]. The mechanism for the removal of humics at pH 5 includes: (a) binding of metal species to anionic sites leading to charge neutralization and reduced solubility and precipitation of the metal–humic complex and (b) adsorption of humics on amorphous metal hydroxide precipitate [31]. DOC removal by coagulation at 3 mmol L1 Al3+ was only slightly better than at 1.5 mmol L1 Al3+ (33% cf. 38%), hence the latter was considered the most effective in terms of DOC removed for alum dosed and so used for the subsequent experiments. Coagulation with alum decreased the humics to a level similar to that after 72 h UVC/H2O2 treatment, and decreased the level of building blocks, but had no impact on the LMW neutrals. Although the reduction of humics was comparable after the stand-alone coagulation and UVC/H2O2 treatments, the greater reduction of colour and A254 after UVC/H2O2 treatment was attributed to the loss of chromophores (some of which would also be fluorophores) through breakdown of the humic-like compounds compared with their removal by coagulation. There was little further reduction of DOC (5%) after biological treatment of the coagulated ROC (Table 3). Coagulation with 1.5 mmol L1 Al3+ followed by UVC/H2O2 treatment led to greater organic matter removal (52%) than by coagulation alone with 1.5 mmol L1 Al3+ (33%). There was a complementary effect for the sequential process as shown by the markedly greater reduction of humic substances (Fig. 5) which corresponded with colour, A254, fluorescence and DOC reductions, however, a net increase in building blocks and no change in LMW neutrals was observed (Fig. 5). Biological treatment (as BDOC assay) removed many of those smaller MW compounds as indicated by the improved total organic content removal of 67%. Coagulation has been reported to improve UV transmittance [27], leading to improved UV penetration and enhanced breakdown of the remaining organic content and therefore greater generation of the biodegradable lower MW compounds. Thus when coagulation is used as a pre-treatment, enhanced reduction of the organic content can be achieved using UV/H2O2 followed by biological treatment as demonstrated in this study, the enhancement was due to the complementary effect of the treatments which led to the removal of a wide range of organic compounds.
study, the energy requirement would be 1 kWh/m3 which is insignificant compared with that needed for generation of UV radiation. The EE/O was calculated for a target DOC residual of 15 mg/L (Table 5). The EE/O needed to achieve the desired residual was high (48 kWh/m3) using UVC/H2O2 only treatment but was reduced by 23% after biological treatment (as BDOC assay). Pre-treatment by coagulation led to a markedly lower EE/O which was further reduced after biological treatment, an overall decrease of 87% compared with stand-alone UVC/H2O2 treatment. The EE/O was 86% lower than for UVC/H2O2 treatment using a conventional UV lamp during treatment of ROC of comparable initial DOC concentration [27]. These EE/O estimates are preliminary as coagulation leads to the production of sludge which requires appropriate processing, increasing the energy requirement and cost of the treatment. However, taking into account that coagulation is generally used as a pre-treatment in municipal WWTPs, the sludge produced during ROC pre-treatment can be handled collectively where appropriate. 6. Future of UV-LED technology Considering that the current UV-LEDs have low quantum efficiency (up to 8%) and low output power (0.3 mW), ongoing developments are expected to significantly improve these parameters which would increase the energy efficiency of the UV-LEDs. UV-LEDs (255 nm) up to 50 mW have recently been developed but are very expensive. The improvements in the efficiency of UV-LEDs are projected to follow the visible LED trajectories which can operate at 75% efficiency [37,38] and their efficiency is projected to match that of visible LEDs by 2020 [13]. These improvements are happening along with an increase in the lifespan of the UV-LEDs. In 2007, they had a lifetime of 200 h which increased to 10,000 h in 2012 and is projected to increase to 100,000 h (11 years) by 2020 [13]. The UV-LEDs used in this study have been operated for more than 5000 h and are expected to be operational up to 10,000 h. Autin et al. [11] calculated that the whole of life cost of UV-LEDs was 150 times greater than the traditional low pressure UV system. However, it was projected to reduce to 40% by 2020 provided the UV-LED technology follows the projected trends. Although the widespread commercial use of UV-LEDs is years away, the developments over the last 5 years are promising and the continuing improvements in their output power and energy efficiency are expected to make this technology a competitive alternative to traditional UV systems.
5. Estimation of electrical energy per order (EE/O) 7. Conclusions A useful tool for measuring the electrical efficiency of the UV-based processes is electrical energy per order (EE/O) developed by Bolton et al. [32]. The EE/O is defined as the electrical energy (in kilowatt hours) required to reduce the concentration of a pollutant by one order of magnitude in 1000 L of water. As coagulation is considered to be a low energy process [33,34] and the energy required for biological treatment is negligible [35] compared with the AOPs, only the energy required for UVC/H2O2 treatment was considered. The energy needed for the production of 1 kg H2O2 is 10 kWh [36], hence for the concentration (3 mmol L1) used in this
The potential of the UVC/H2O2 process using UVC-LEDs (255 nm) for the degradation of the organic content of ROC, a concentrated complex wastewater of very high salinity, was demonstrated. Greater reductions of DOC, colour, A254 and SUVA were obtained by reducing the pH to 5 during UVC/H2O2 only treatment. Coagulation using alum at pH 5 removed a significant proportion of the organic matter (34–38%) and led to enhanced overall reduction after UVC/H2O2 treatment. The UVC/H2O2 treatment increased the biodegradability, and using biological treatment (as BDOC assay)
656
M. Umar et al. / Chemical Engineering Journal 260 (2015) 649–656
after the sequence of coagulation and UVC/H2O2 treatment (UV fluence 36 103 mJ/cm2) led to an overall DOC reduction of 67%, demonstrating the effectiveness of the proposed treatment scheme. Although the low power output of the UV-LEDs and high cost are the main limitations for their application in water and wastewater treatment, the ongoing developments in this technology, combined with improved reactor design, are expected to improve the viability of their use. Acknowledgements We would like to acknowledge the partial financial support provided by Water Research Australia and the Smart Water Fund, Victoria, Australia. References [1] B. Van der Bruggen, L. Lejon, C. Vandecasteele, Reuse, treatment, and discharge of the concentrate of pressure-driven membrane processes, Environ. Sci. Technol. 37 (2003) 3733–3738. [2] F. Tang, H.Y. Hu, Q.Y. Wu, X. Tang, X.Y. Sun, X.L. Shi, J.J. Huang, Effects of chemical agent injections on genotoxicity of wastewater in a microfiltrationreverse osmosis membrane process for wastewater reuse, J. Hazard. Mater. 260 (2013) 231–237. [3] C. Vargas, A. Buchanan, Monitoring ecotoxicity and nutrients load in the reverse osmosis concentrate from Bundamba Advanced Water Treatment Plant, Queensland, Australia, Water Pract. Technol. 6 (2011). [4] S.J. Khan, D. Murchland, M. Rhodes, T.D. Waite, Management of concentrated waste streams from high-pressure membrane water treatment systems, Crit. Rev. Environ. Sci. Technol. 39 (2009) 367–415. [5] W.R. Haag, C.C.D. Yao, Rate constants for reaction of hydroxyl radicals with several drinking water contaminants, Environ. Sci. Technol. 26 (1992) 1005– 1013. [6] S. Sarathy, M. Mohseni, Effects of UV/H2O2 advanced oxidation on chemical characteristics and chlorine reactivity of surface water natural organic matter, Water Res. 44 (2010) 4087–4096. [7] J. Malley, UV in Water Treatment – Issues from the Next Decade, International Ultraviolet Association News 12 (2010) 18–25. [8] M.A. Würtele, T. Kolbe, M. Lipsz, A. Külberg, M. Weyers, M. Kneissl, M. Jekel, Application of GaN-based ultraviolet-C light emitting diodes – UV LEDs – for water disinfection, Water Res. 45 (2011) 1481–1489. [9] H.W. Chen, Y. Ku, A. Irawan, Photodecomposition of o-cresol by UV-LED/TiO2 process with controlled periodic illumination, Chemosphere 69 (2007) 184– 190. [10] Y. Taniyasu, M. Kasu, T. Makimoto, An aluminium nitride light-emitting diode with a wavelength of 210 nanometres, Nature 441 (2006) 325–328. [11] S. Vilhunen, J. Puton, J. Virkutyte, M. Sillanpää, Efficiency of hydroxyl radical formation and phenol decomposition using UV light emitting diodes and H2O2, Environ. Technol. 32 (2011) 865–872. [12] A. Jamali, R. Vanraes, P. Hanselaer, T. Van Gerven, A batch LED reactor for the photocatalytic degradation of phenol, Chem. Eng. Process. 71 (2013) 43–50. [13] J.L. Shie, C.H. Lee, C.S. Chiou, C.T. Chang, C.C. Chang, C.Y. Chang, Photodegradation kinetics of formaldehyde using light sources of UVA, UVC and UV LED in the presence of composed silver titanium oxide photocatalyst, J. Hazard. Mater. 155 (2008) 164–172. [14] R.J. Tayade, T.S. Natarajan, H.C. Bajaj, Photocatalytic degradation of methylene blue dye using ultraviolet light emitting diodes, Ind. Eng. Chem. Res. 48 (2009) 10262–10267.
[15] O. Autin, C. Romelot, L. Rust, J. Hart, P. Jarvis, J. MacAdam, S.A. Parsons, B. Jefferson, Evaluation of a UV-light emitting diodes unit for the removal of micropollutants in water for low energy advanced oxidation processes, Chemosphere 92 (2013) 745–751. [16] W.Y. Wang, Y. Ku, Photocatalytic degradation of Reactive Red 22 in aqueous solution by UV-LED radiation, Water Res. 40 (2006) 2249–2258. [17] M.A.S. Ibrahim, Commercial evaluation of UV-LED in water treatment applications (MSc. thesis), Cranfield University, UK, 2012. [18] R.V. Steele, The story of a new light source, Nat. Photonics 1 (2007) 25–26. [19] APHA, Standard Methods for the Examination of Water & Wastewater, American Public Health Association, Washington DC, USA, 2005. [20] S. Jin, A.A. Mofidi, K.G. Linden, Polychromatic UV fluence measurement using chemical actinometry, biodosimetry, and mathematical techniques, J. Environ. Eng. 132 (2006) 831–841. [21] W. Chen, P. Westerhoff, J.A. Leenheer, K. Booksh, Fluorescence excitation– emission matrix regional integration to quantify spectra for dissolved organic matter, Environ. Sci. Technol. 37 (2003) 5701–5710. [22] J.C. Joret, Y. Levi, Methode rapide d’evaluation du carbone eliminable des eaux voie biologique, Tribune du Cebedeau 510 (1986) 3–9. [23] C. Volk, C. Renner, P. Roche, H. Paillard, J.C. Joret, Effects of ozone on the production of biodegradable dissolved organic carbon (BDOC) during water treatment, Ozone Sci. Eng. 15 (1993) 389–404. [24] M. Umar, F. Roddick, L. Fan, Assessing the potential of a UV-based AOP in treating high-salinity municipal wastewater reverse osmosis concentrate, Water Sci. Technol. 68 (2013) 1994–1999. [25] C.H. Liao, S.F. Kang, F.A. Wu, Hydroxyl radical scavenging role of chloride and bicarbonate ions in the H2O2/UV process, Chemosphere 44 (2001) 1193–1200. [26] K. Liu, F.A. Roddick, L. Fan, Impact of salinity and pH on the UVC/H2O2 treatment of reverse osmosis concentrate produced from municipal wastewater reclamation, Water Res. 46 (2012) 3229–3239. [27] M. Umar, F. Roddick, L. Fan, Effect of coagulation on treatment of municipal wastewater reverse osmosis concentrate by UVC/H2O2, J. Hazard. Mater. 266 (2014) 10–18. [28] S.A. Huber, F.H. Frimmel, Size-exclusion chromatography with organic carbon detection (LC-OCD): a fast and reliable method for the characterization of hydrophilic organic carbon, Vom Wasser 86 (1996) 277–290. [29] E.H. Goslan, F. Gurses, J. Banks, S.A. Parsons, An investigation into reservoir NOM reduction by UV photolysis and advanced oxidation processes, Chemosphere 65 (2006) 1113–1119. [30] J. Bratby, Coagulation and Flocculation in Water and Wastewater Treatment, IWA Publishing, London, UK, 2006. [31] J. Duan, J. Gregory, Coagulation by hydrolysing metal salts, Adv. Colloid Interface Sci. 100–102 (2003) 475–502. [32] J.R. Bolton, K.G. Bircher, W. Tumas, C.A. Tolman, Figures-of-merit for the technical development and application of advanced oxidation technologies for both electric-and solar-driven systems (IUPAC Technical Report), Pure Appl. Chem. 73 (2001) 627–638. [33] D. Bromley, M. Gamal El-Din, D.W. Smith, A low cost treatment process to reduce phosphorus and suspended solids in liquid waste from animal farm operations, Penang, Malaysia, 2002. [34] C. Huang, M. Cao, J. Luo, C. Zhang, Study of chemically enhanced primary treatment of domestic sewage. International Conference on Agriculture and Natural Resources Engineering, Adv. Biomed. Eng. 3–5 (2011) 270–273. [35] A. Marco, S. Esplugas, G. Saum, How and why combine chemical and biological processes for wastewater treatment, Water Sci. Technol. 35 (1997) 321–327. [36] E.J. Rosenfeldt, K.G. Linden, S. Canonica, U. von Gunten, Comparison of the efficiency of OH radical formation during ozonation and the advanced oxidation processes O3/H2O2 and UV/H2O2, Water Res. 40 (2006) 3695–3704. [37] T. Bettles, S. Schujman, J.A. Smart, W. Liu, L. Schowalter, UV light emitting diodes – their applications and benefits, International Ultraviolet Association Conference, Los Angeles, CA, 2007, USA. [38] C. Chatterley, K. Linden, Demonstration and evaluation of germicidal UV-LEDs for point-of-use water disinfection, J. Water Health 8 (2010) 479–486.