Environ Sci Pollut Res DOI 10.1007/s11356-015-6034-x

RESEARCH ARTICLE

Integration of biofiltration and advanced oxidation processes for tertiary treatment of an oil refinery wastewater aiming at water reuse A. A. Nogueira 1 & J. P. Bassin 1 & A. C. Cerqueira 2 & M. Dezotti 1

Received: 28 October 2015 / Accepted: 29 December 2015 # Springer-Verlag Berlin Heidelberg 2016

Abstract The combination of biological and chemical oxidation processes is an interesting approach to remove ready, poor, and non-biodegradable compounds from complex industrial wastewaters. In this study, biofiltration followed by H2O2/UV oxidation (or microfiltration) and final reverse osmosis (RO) step was employed for tertiary treatment of an oil refinery wastewater. Biofiltration alone allowed obtaining total organic carbon (TOC), chemical oxygen demand (COD), UVabsorbance at 254 nm (UV254), ammonium, and turbidity removal of around 46, 46, 23, 50, and 61 %, respectively. After the combined biological-chemical oxidation treatment, TOC and UV254 removal amounted to 88 and 79 %, respectively. Whereas, the treatment performance achieved with different UV lamp powers (55 and 95 W) and therefore distinct irradiance levels (26.8 and 46.3 mW/cm2, respectively) were very similar and TOC and UV254 removal rates were highly affected by the applied C/H2O2 ratio. Silt density index (SDI) was effectively reduced by H2O2/UV oxidation, favoring further RO application. C/H2O2 ratio of 1:4, 55 W UV lamp, and 20-min oxidation reaction corresponded to the experimental condition which provided the best cost/benefit ratio for TOC, Responsible editor: Bingcai Pan Electronic supplementary material The online version of this article (doi:10.1007/s11356-015-6034-x) contains supplementary material, which is available to authorized users. * J. P. Bassin [email protected]

1

COPPE—Chemical Engineering Program, Federal University of Rio de Janeiro, P.O. Box 68502, 21941-972 Rio de Janeiro, Brazil

2

CENPES—Centro de Pesquisa e Desenvolvimento Leopoldo Américo Miguez de Mello—Petrobras, Rio de Janeiro, Brazil

UV254, and SDI reduction from the biofilter effluent. The array of treatment processes proposed in this study has shown to be adequate for tertiary treatment of the oil refinery wastewater, ensuring the mitigation of membrane fouling problems and producing a final effluent which is suitable for reuse applications. Keywords Oil industry wastewater . Reuse . Residual organic matter . Biofilter . H2O2/UV . SDI

Introduction The environmental impact of oil production and processing industries is often negative, since they consume large quantities of water and generate a considerable amount of wastewater. As oil demand is expected to increase in the next two decades (Kjärstad and Johnsson 2014), there is an increasing awareness of the need to avoid the pollution caused by oil industry waste streams. Oil refinery wastewaters commonly show complex and time-variable composition and contain some hardly biodegradable and refractory chemical compounds which may present serious toxic hazards to humans and environment and tend to be persistent (Chavan and Mukherji 2008). In order to minimize the negative impact on the natural resources, a wide range of methods have been investigated in earlier studies for treatment of oil refinery wastewaters: biological processes (Ma et al. 2009; Schneider et al. 2011; Souza et al. 2011), coagulation (El-Naas et al. 2009), electrochemical processes (Santos et al. 2006), catalytic wet air oxidation (Sun et al. 2008), and chemical oxidation (Abdelwahab 2009). From the aforementioned techniques, environmentally friendly biological processes have sustainable and economic characteristics and are widely employed in the treatment of oil processing

Environ Sci Pollut Res

industry wastewaters to remove their biodegradable constituents (Mazzeo et al. 2010). Nevertheless, given the complexity of the effluents generated by this industrial sector, some constituents are difficult to be removed biologically (Saien and Nejati 2007) and often persist after conventional secondary biological treatment. Biofiltration seems to be an interesting option for tertiary treatment of oil refinery wastewaters, as it was reported to efficiently remove both residual organic matter and nutrients from different types of waste streams (Zhidong et al. 2010; Reungoat et al. 2011). Biofilter systems are of simple construction and present robust operation and low energy requirements (Reungoat et al. 2012). Besides the role of microorganisms attached to the biofilter support in the biodegradation of pollutants, the retention of suspended material by the filtering media considerably improves the quality of the final effluent. However, recalcitrant and inhibitory organic compounds may still persist after tertiary biofiltration and the treated effluent might not meet the standard required for direct water reuse activities (Ried et al. 2006). In such circumstances, further treatment processes are necessary. Advanced oxidation processes (AOP) are increasingly drawing attention in recent years and appear as an option to break down residual hardly biodegradable compounds from many industrial wastewaters previously subjected to biological treatment (Oller et al. 2011). Based on the generation of powerful non-selective hydroxyl radicals (HO·), AOP present high reaction rates and can decompose simultaneously a large number of organic pollutants, including the most refractory ones, into mineralized products, such as carbon dioxide, inorganic ions, and water (Esplugas et al. 2002; Antonopoulou et al. 2014). The generation of HO· radicals can be achieved by means of homogenous chemical reactions which employ a specific oxidant or a combination oxidants such as ozone (O3), iron (Fe2+) hydrogen peroxide (H2O2), and ultraviolet (UV) light (Comninellis et al. 2008). Hydroxyl radicals may also be generated in heterogeneous oxidation processes by employing semiconductive materials, such as titanium dioxide (TiO2), iron oxide (Fe2O3), and zinc oxide (ZnO) (Muruganandham et al. 2014). From all these chemical treatments, the use of hydrogen peroxide is very cost-effective, since it consists of a relatively cheap and readily available chemical (Munter 2001). Hydrogen peroxide can be decomposed in radical species but can also directly react with chemical substances present in a given wastewater. However, its oxidation potential can be strongly enhanced by combining it with UV irradiation (Andreozzi et al. 2003). Many studies reported that peroxide-based oxidation processes such as H2O2/UV show promising results in the removal of organic matter from wastewaters generated by oil processing facilities (Stepnowski et al. 2002; Souza et al. 2011; Zoschke et al. 2012).

In general, due to their high associated costs (Oller et al. 2011), AOP are more indicated for the treatment of wastewaters with low total organic carbon (TOC) content (lower than 5 g/L) (Andreozzi et al. 1999). Therefore, a combination of AOP with biological techniques is often used to minimize costs, increase the overall treatment performance, and open up the possibility of reuse of the treated wastewater (Andreozzi et al. 1999; Oller et al. 2011). Up to date, the secondary/tertiary treatment of oil refinery wastewaters aiming at reuse is scarcely reported in literature (Dos Santos et al. 2005; Schneider et al. 2011; Souza et al. 2011), while the combination of biological, oxidation, and membrane separation processes has not yet been investigated for this particular purpose. From our previous investigations (Schneider et al. 2011; Souza et al. 2011), it was shown that the utilization of AOP to oxidize and enhance the biodegradability of recalcitrant organic matter from secondary biological treatment effluent did not lead to an improvement of TOC removal in further tertiary biological activated carbon (BAC) treatment. On the contrary, the TOC removal in the BAC systems was even lower when the secondary effluent has undergone AOP pretreatment (Schneider et al. 2011; Souza et al. 2011). In this context, this study addressed the treatment of an oil refinery wastewater previously subjected to a conventional activated sludge process by coupling a biofiltration system (without prior AOP treatment step) with an H2O2/UV oxidation process. Microfiltration (MF) was also evaluated as a replacement for the AOP process, whereas reverse osmosis (RO) was applied as a final step to allow subsequent reuse of the final effluent at the industrial facility. The most appropriate experimental conditions to achieve such a goal are evaluated and discussed.

Material and methods Wastewater source The wastewater used in this study was previously subjected to physicochemical and biological treatment at the wastewater treatment plant of an oil refinery (REGAP, Petrobras, Brazil). The wastewater was collected after sand filtration and immediately stored under refrigeration (4 °C) after arrival in the laboratory. A schematic representation of the main processes taking place in the industrial wastewater treatment facility is shown in Fig. S1 (Supplementary Material). Table 1 shows the composition of the industrial wastewater under study. As the transport of the wastewater from the oil refinery to the laboratory was dependent on local services, the amount of wastewater available to feed the biofilter was not enough in some occasions. In such circumstances, the real wastewater had to be mixed with a synthetic medium in a proportion 1:1 (v/v) to ensure normal operation of the biofilter. The

Environ Sci Pollut Res Table 1 Characteristics of the oil refinery wastewater used in this study. The range of values corresponds to the samples collected in 49 campaigns Parameter

Concentration (mg/L)

Chemical oxygen demand (COD) (mg/L)

18–61

Total organic carbon (TOC) (mg/L)

6.3–17.51

Inorganic carbon (IC) (mg/L) Ammonium (mgN/L)

1.2–22 0.8–7.97

Nitrite (mgN/L) Nitrate (mgN/L)

0.1–1.3 1.1–4.4

Abs 254

0.11–0.32

Chloride (mg/L)

220–356

Conductivity (μS/cm)

418–793

pH

6.3–7.6

Turbidity (NTU)

3

Dissolved oxygen (mg/L)

3.6–4.7

Advanced oxidation experiments

composition of this medium is detailed in Table S1. The TOC and the composition of the medium prepared in laboratory were chosen to be similar to that of the real wastewater. The tertiary treatment of the oil refinery wastewater involved a combination of different processes (biofiltration, H2O2/UV oxidation or microfiltration, and finally reverse osmosis), as displayed in Fig. S2. The detailed description of each step can be found as follows. Biofilter operation A lab-scale biofilter (10 cm height, 9.6 cm diameter, and 0.72 L of useful volume) was filled with expanded clay Biolite® media, whose characteristics are shown in Table S2. The media filling ratio (volume of media occupied per volume of reactor) was 60 %. The biofilter setup is shown in Fig. 1a. The reactor was continuously fed in upflow mode with the

(a)

biologically treated oil refinery for 170 days. The influent wastewater was fed to the biofilter through five inlets located at the bottom of the system to have a homogenous flow through the filter media, avoiding preferential flow paths. The surface flow and the hydraulic retention time were set at 0.3 m3/(m2 h) and 0.6 h, respectively. No external aeration was provided and dissolved oxygen concentration varied from 3.6–4.7 mg/L by normal circulation of liquid through the biofilter media. Sludge detached from the plastic carriers of a lab-scale moving bed bioreactor system fed with synthetic medium was used as inoculum for the biofiltration system. pH was kept mostly at around 7.0–8.0 and the running temperature was 25.7 ± 1.2 °C.

A combined H2O2/UV process was used for post-treatment of the biofilter effluent aiming at the removal of residual organic matter. The oxidation experiments were carried out in 4-L cooling jacket-equipped glass cylindrical reactors (60 cm height, 12.5 cm diameter, and 4 L of total volume) containing quartz tubes to accommodate the low pressure mercury lamps used to emit UV-C light at 254 nm (Fig. 1b). A porous diffuser was placed at the bottom of the reactors for homogenization of the bulk liquid during the oxidation reaction. Different tests were carried out in which both the ratio between the total organic carbon (assessed from TOC measurements of the biofilter effluent) and the hydrogen peroxide concentrations (hereafter referred to as C/H2O2 ratio) and the power of the UV lamps were varied. For comparison, experiments were also conducted with either UVor H2O2 only. The experimental conditions are summarized in Table 2. The UV irradiance (mW/cm2) was calculated by considering the power of the UV-C lamps (W) and the total area of the surface (cm2) subjected to the light source. Given that the light was emitted symmetrically in all directions, the irradiance was considered

(b)

Fig. 1 Simplified scheme of the experimental set-up: a upflow biofilter filled with Biolite® media (real picture of the biofilter shown on the left side of the scheme); b H2O2/UV reactor (photograph of the cylindrical glass reactor displayed on the right side of the illustration)

Environ Sci Pollut Res Table 2

Experimental conditions of the H2O2/UV oxidation tests

Molar proportions (C/H2O2)

Power of the UV-C lamps

UV irradiance (mW/cm2)

Reaction time (min)

55 W

26.8

20 30

95 W

46.3

40

1:1

10

1:2 1:4

50 60

to have the same value at all points of the working surface of the cylindrical reactor. The number of photons supplied by each UV-C lamp, calculated from Eq. S1 (Supplementary Material), was estimated to be 7.01E19 and 1.21E20 photons/s for 55 and 95 W lamps. The oxidation reaction time varied from 10 to 60 min. Samples were collected every 10 min and the performance of the H2O2/UV treatment was evaluated in terms of removal of TOC and UV absorbance at 254 nm wavelength (hereafter referred to as UV254). The latter parameter is associated with wavelength in which the most complex forms of organic matter (e.g., aromatic compounds with double or triple bonds) have maximum absorption (Huang et al. 2004; Lamsal et al. 2011) and therefore can provide an indication of their contents. Membrane separation processes Microfiltration Microfiltration tests were also conducted as a post-treatment of the biofilter effluent, and the results were compared to those obtained in the H2O2/UV oxidation experiments. A 0.45-μm pore diameter polymeric membrane consisting of a mixture of cellulose acetate and cellulose nitrate was employed in the microfiltration experiments. The work pressure was kept at 2 bar. The performance of the microfiltration system was evaluated in terms of TOC, conductivity, UV254, silt density index (SDI), and turbidity removal. Determination of the silt density index The SDI indicates the fouling capacity of a wastewater in a reverse osmosis (RO) unit. Essentially, this measurement provides the percentage decrease, per minute, in the flow rate of an aqueous matrix through a membrane over predefined intervals of the time due to silt built up. As fouling increases with the SDI value, increased cleaning frequency or decreased permeate flux are expected for wastewaters presenting high SDI values (Raffin et al. 2013). The SDI can be determined by Eq. (2.1), which takes into account the difference between the time necessary to collect a determined amount of permeate

(500 mL) (ti) and the time necessary to collect the same volume of permeate (tf) after a defined time interval (tt = 15 min). The experiment was conducted according to ASTM (ASTM D4189-07 2014). To estimate the SDI, a pressurized (2 bar) filtration system was used. The apparatus was composed by a pump, a flow regulator, a pressure regulator valve, a manometer, and a permeation cell containing a 0.45-μm pore size microfiltration membrane. The SDI tests were conducted at different stages of the treatment: (1) before biofiltration with and without subsequent microfiltration post-treatment and (2) after biofiltration followed by microfiltration, UV, or combined H2O2/UV treatment. For the last condition, SDI of the effluent was determined after it was subjected different oxidation reaction times (from 10 to 60 min).   ti 100 1− tf SDI ¼ ð2:1Þ tt

Reverse osmosis A bench scale reverse osmosis (RO) system composed by a polyamide membrane (FILMTEC BW30-2540, Dow Chemical Company) was used in this work. In one experiment, the RO system was fed with 5 L of the biofilter effluent. In a second test, the RO unit was fed with 5 L of the biofilter effluent subjected to H2O2/UV oxidation under the following conditions: C/H2O2 ratio of 1:4, UV lamp of 55 W, and 20min reaction). The permeate flux obtained in each experiment was monitored every hour over a long-term period (25 h) for comparison. All experiments were conducted at a pressure of 2 bar. Analytical methods TOC, chemical oxygen demand (COD), ammonium, turbidity, and conductivity were measured according to standard procedures (APHA et al. 2005). H2O2 concentrations were determined by a colorimetric method with ammonium metavanadate at 450 nm (Nogueira et al. 2005). Nitrite and nitrate were determined by means of an ion chromatography system (Dionex ICS90). For determination of TOC, COD, and nitrogen compounds, samples were filtered in 0.45-μm cellulose ester membranes (Millipore). UV254, used to provide an indication of the amount of complex organic substances present in the wastewater, was measured in a Shimadzu UV spectrophotometer (UV mini 1240). pH and temperature were regularly monitored by means of a Digimed DM-23 electrode. The removal of the several parameters evaluated in this work was calculated based on their concentrations before and after the wastewater has undergone a particular treatment process. Microscopic observations of

Environ Sci Pollut Res

the biomass attached to the biofilter media were carried out in an optical microscope (Hund H500) to follow the characteristics of the microfauna over time. Images were acquired using a Nikon Coolpix camera connected to the microscope.

Results and discussion Biofilter operation TOC and COD profiles over the entire operation of the biofilter (170 days) are displayed in Fig. 2. During the first 2 months of operation, TOC removal varied from 30 to 50 %. Further operation allowed a slight improvement in TOC removal efficiency, which remained between 40 and 65 % in the subsequent 4 months (Fig. 2a). Average influent and effluent TOC corresponded to 9.1 and 4.9 mg/L, respectively, yielding an average TOC removal of around 46 %. Similar TOC removal efficiency was obtained by Melin and Odegaard (1999), who also employed an upflow biofilter filled with expanded clay media. These authors reported a TOC removal of 18–37 % in the treatment of humic water previously subjected to ozonation to enhance its biodegradability. Comparable results were also reported in other previous investigations in which several types of media were used as filling material in biofiltration systems (Hozalski et al. 1995; Melin et al. 2000). The data representing the COD profile over the course of the biofilter operation is shown in Fig. 2b. The mean values for influent and effluent COD amounted to 37 and 20 mg/L, respectively. This implies a removal percentage comparable to that obtained for TOC. COD/TOC ratio throughout the biofilter operation was observed to be around 4. To ensure that the reduction of COD was not only attributed to physical filtration through the media but also to biodegradation, soluble COD measurements were also conducted. Similar values were found for both total and soluble COD, indicating that the microorganisms colonizing the biofilter played an important role in the removal of organic matter.

(a)

The results show that the removal percentages of TOC and COD are relatively high taking into consideration that the amount of carbonaceous matter fed to the biofilter is significantly low. Besides that, it should be noted that this organic matter comes from a biological treatment, where the easily biodegradable compounds were already removed. Therefore, the biofilter influent wastewater comprised mainly the nonbiodegradable or hardly biodegradable substances. Commonly, this organic matter fraction is considered to be non-utilizable by bacteria. Nevertheless, the outcomes of this study show the contrary, i.e., it is still possible to remove a considerable amount of this carbonaceous material by biological means by employing a biofiltration system downstream of the conventional secondary activated sludge process. Although the biofilter was effective in terms of TOC and COD removal, UV254 reduction was observed to be rather low (Fig. 3a). Average UV254 removal was close to 23 %, meaning that the biofiltration system could not assimilate some complex organic compounds containing double and triple bonds. This implies that further treatment steps were still required to remove the persistent organic pollutants remaining in the wastewater after biofiltration. The influent ammonium concentrations were low and amounted up to 10 mg N/L (Fig. 3b). Ammonium removal was quite unstable, ranging from 30 to 80 %. During the operation of the biofilter, a rapid growth of the biofilm thickness was observed in the expanded clay media. This may have influenced the competition between fast-growing heterotrophs and slow-growing nitrifiers for space and oxygen, affecting nitrification performance which significantly oscillated over time (Morgenroth and Wilderer 2000). Indeed, the lowest ammonium removal was observed in the last 3 months of the biofilter operation, period when the influent COD was higher (Fig. 2b). When biodegradable organic compounds are available, heterotrophic organisms will be preferentially located in the outer layers of the biofilm, where they have easier access to nutrients and oxygen (Harremoës 1982). On the other hand, nitrifying organisms are often placed in the inner part of the biofilm, where oxygen and substrate limitation might occur as

(b)

Fig. 2 TOC (a) and COD (b) profiles in the influent and effluent of the biofilter (right y-axis) and TOC and COD removal over time (left y-axis). (Black square) influent and (white square) effluent TOC and COD; (diamond) TOC and COD removal efficiency

Environ Sci Pollut Res

(a)

(b)

Fig. 3 UV254 absorbance (a) and NH4 (b) profiles in the influent and effluent of the biofilter (right y-axis) and UV254 and NH4 removal over time (left y-axis). (Black square) influent and (white square) effluent UV254 and NH4; (diamond) UV254 and NH4 removal efficiency

a consequence of heterotrophic activity (Harremoës 1982). Actually the competition for space between heterotrophic and nitrifying bacteria has been reported as a limiting factor to obtain stable nitrification in fixed-film reactor systems (Wijeyekoona et al. 2004). Another reason for the variable ammonium removal performance might be related to the fact that alkalinity was not monitored over time. As such, the amount of alkalinity supplied by the incoming wastewater could not be enough to ensure proper nitrification in some days of the biofilter operation. Taking into account the influent and effluent concentration of ammonium, nitrite and nitrate corresponding to several days of operation, a nitrogen mass balance was conducted. A nitrogen loss of about 65 % was observed in the biofilter, which is possibly attributed to denitrification occurring in the anoxic zones established inside the expanded clay media. The denitrification potential was more relevant at higher COD input, as the thickness of the biofilm was more pronounced. Since the amount of solids produced in the system was very low, assimilation of nitrogen for cell synthesis only accounted for a small part of the total nitrogen removed. An advantage shown by the biofiltration technology is the capacity to retain suspended particles which contribute to increase effluent turbidity and may potentially cause deterioration of the effluent quality. Hence, the turbidity of the influent and effluent of the biofilter was continuously monitored (data not shown). The average turbidity of the influent and effluent flows corresponded to 3 and 1 NTU, respectively, which implies a removal of higher than 60 %. The presence of a highly diverse community of ciliated and stalked protozoa (e.g., Euplotes and Vorticella, respectively) and micrometazoans (mainly rotifers) in the biofilter microfauna (Fig. S3) might also have contributed to obtain low amount of effluent suspended particles. These organisms are often regarded as flocs predators, helping to keep low amount of dispersed microorganisms and particulate matter in biological treatment systems (Madoni 1994; Metcalf and Eddy 2003). The ciliated protozoa belonging to the Euplotes genera were abundantly observed over the entire operation. In some samples, rotifers

of the class Litonotus (Fig. S3) were also detected. These organisms have been related to aeration deficiency, low hydraulic retention times and overloading (Madoni 1994). However, its presence was of minor importance compared to other higher organisms. Post-treatment of biofilter effluent by the combined H2O2/UV oxidation process A combined H2O2/UV oxidation process was used as a posttreatment of the biofilter effluent to remove the remaining organic matter fraction which could impede the implementation of subsequent treatment processes (e.g., reverse osmosis) aiming at water reuse. The use of an AOP downstream the biological process minimizes the operating costs of the chemical treatment and is also convenient for disinfection of the treated effluent (Wert et al. 2007; Zhao et al. 2008). Several oxidation tests were conducted by varying the carbon/hydrogen peroxide (C/H2O2) ratio (1:1, 1:2, and 1:4) and the power of the UV lamps (55 and 95 W). The results in terms of TOC removal obtained in this set of experiments are displayed in Fig. 4a. To track the degradation of complex compounds with double or triple bonds in the biologically treated effluent, UV254 reduction was also monitored over the course of the oxidation tests in which different C/H2O2 ratios were employed (Fig. 4b). Although the power of the two lamps was quite different, the difference in the TOC removal obtained with the 95 and the 55-W UV lamps was observed to be marginal (less than 6 %) for all C/H2O2 ratios applied (Fig. 4a). For the experiments conducted with 95-W UV lamps and C/H2O2 ratio of 1:1, the cumulative TOC removal achieved after 60 min of reaction was only 0.2 mg/L higher than that obtained with the 55-W UV lamp. Similar results were obtained for the experiment applying a C/H2O2 ratio of 1:4, in which the cumulative removal of TOC obtained with the UV lamp with the higher power was only 0.4 mg/L higher than that obtained with the 55-W lamp. Interestingly, when a C/H2O2 ratio of 1:2 was used, the TOC removal observed in the test with 95-W UV

Environ Sci Pollut Res

Fig. 4 Cumulative TOC (a) and UV254 (b) removed (bar points) and energy supplied (triangle data points) to the 55 W (white triangle) and 95 W (black triangle) UV lamps over time in the experiments in which the C/H2O2 ratio was varied: 1:1–55 W (pink square) and 95 W (green

square), 1:2–55 W (blue square) and 95 W (orange square), 1:4–55 W (red square) and 95 W (yellow square). The irradiance levels obtained from the 55 and 95-W lamps corresponded to 26.8 and 46.3 mW/cm2, respectively

lamp was lower but still very comparable with that obtained with the 55-W lamp. More than 50 % of UV254 abatement was observed within the first 10 min of reaction (Fig. 4b). Similar to the observed for TOC degradation, the reduction of UV254 obtained with the two different UV lamps was comparable and no major difference was found. On the other hand, the influence of hydrogen peroxide concentration was very noticeable. By increasing the amount of this oxidant, both TOC and UV254 removal efficiencies were greatly enhanced. The best results were obtained with the C/H2O2 ratio of 1:4, UV lamp of 95 W, and 60 min of reaction, for which the TOC and UV254 removal amounted to 78 and 73 %. The significant removal of UV254 implies that, unlike the biofiltration system, the oxidation process was able to greatly reduce the complexity of the compounds present in the aqueous matrix. The reason why only slight differences in the TOC and UV254 removal were observed in the experiments with 55 and 95-W lamps is attributed to the amount of peroxide available during the oxidation reaction. As displayed in Fig. S4, the peroxide consumption profile was similar for both 55 and 95-W tests, and the amount of residual peroxide at the end of the experiment (60 min) was found to be very low for both cases. Furthermore, the rate constant (k) for the peroxide consumption obtained with different lamps was found to be similar for each C/H2O2 ratio (0.0161 and 0.0208 min−1 for 1:2–55 W and 1:2– 95 W, respectively; and 0.0285 and 0.0325 min−1 for 1:4–55 W and 1:4–95 W, respectively). Hence, the UV dose applied in the test using 55-W lamps was sufficient for the initial peroxide concentrations employed (12– 22 mg/L). Table 3 summarizes all the results obtained in the whole set of experiments varying the C/H2O2 ratio and the power of the UV lamp. It should be mentioned that in the tests using H2O2, there was no significant changes in the TOC and UV254. Moreover, in the

experiments conducted only with UV (photolysis), a negligible decline in TOC values was observed, while a reduction of 23 % was noticed for UV254. Some works addressing the treatment of wastewaters generated by petroleum/petrochemical industrial facilities by AOPs (UV/H2O2, photocatalysis, Fenton and photoFenton) are summarized in Table 4. The surveyed works employed AOP for the secondary treatment of such waste streams, which contained relatively high TOC/DOC concentrations (>100 mg/L). However, in this study, TOC concentrations downstream of the biofiltration process were much lower, mostly around 20 mg/L. Under such conditions, the degradation of the remaining organic compounds becomes more difficult. Indeed, this could be noticed in Fig. 4a, which shows the decreasing TOC removal rate (cumulative TOC removed) as the oxidation reaction proceeded. Nevertheless, the overall TOC removal at the C/H2O2 ratio of 4 ranged from 64 % (55 W) to 78 % (95 W), which is very comparable to those obtained in former studies employing higher (Parilti 2010; Diya’uddeen et al. 2015) or similar (Bustillo-Lecompte et al. 2015) initial TOC concentrations at secondary level treatment. As already reported in previous investigations evaluating the H2O2/UV process (Li et al. 2008), the results of this study sugg est th at, to obtain the best cost/efficiency ratio, it is crucial to make use of appropriate UV light source and apply preliminary treatment steps to remove the organic matter as much as possible. The findings indicate that there is no need to use 95-W lamps for TOC and UV254 abatement, and the amount of H2O2 is in fact determining the maximum removal efficiency. Other literature studies also pointed out that the removal of organic matter in H2O2-based AOPs is strongly influenced by the concentration of this oxidant (Hu et al. 2011). As the running costs of the UV lamps

Environ Sci Pollut Res Table 3 Summary of the results obtained in the different oxidation tests in which both the C/H2O2 ratio and the power of the UV lamps were varied

Initiala

Finalb

TOC

TOC

(mg/L)

(mg/L)

1:1–55 W

4.2 ± 1.2

1:1–95 W 1:2–55 W

4.4 ± 0.8 5.4 ± 0.6

1:2–95 W 1:4–55 W 1:4–95 W

C:H2O2 ratio and UV lamp power

TOC removal (%)

Initiala

Finalb

UV254

UV254

3.8 ± 0.9

9

0.107 ± 0.061

0.064 ± 0.045

40

3.9 ± 0.4 2.3 ± 0.5

11 57

0.112 ± 0.063 0.155 ± 0.028

0.069 ± 0.048 0.050 ± 0.028

38 68

5.6 ± 0.7 5.7 ± 0.9

2.7 ± 0.4 2.0 ± 0.6

52 65

0.155 ± 0.021 0.220 ± 0.027

0.052 ± 0.012 0.067 ± 0.017

66 70

7.1 ± 2.1

1.6 ± 0.2

78

0.205 ± 0.033

0.056 ± 0.017

73

a

At the beginning of the experiment (t = 0)

b

After 60 min of oxidation reaction

are very significant (Esplugas et al. 2002) and much higher than that of hydrogen peroxide (Munter 2001), it is convenient to perform tests where the amount of

UV254 reduction (%)

peroxide is increased while the UV lamp power is kept low to minimize the energy-related costs, making the oxidation process more economically feasible.

Table 4 Summary of some works addressing the treatment of wastewater generated by petroleum/petrochemical industries by means of advanced oxidation processes Wastewater

Process

Petroleum refinery wastewater

UV/H2O2

Main findings

H2O2 concentrations tested: 1.17, 3.52, 5.88, and 11.76 mM. 1,2-dichloroethane and t-butyl methyl ether degradation were favored at higher H2O2 levels Synthetic wastewater with raw Solar photo-Fenton The photodegradation experiments showed that, on gasoline average, 60 % of the initial organic content was mineralized in the first 3 h. After 4.5 h, the mineralization ranged from 66 to 91 %. Inicial DOC: 300–440 mg/L. DOC removal of 25 % Petroleum refinery sourwater H2O2, H2O2/UV, UV, (H2O2/UV), 35 % (O3), 21 % (photocatalysis), 55 % photocatalysis, O3, Fenton and photo-Fenton (Fenton), 83 % (photo-Fenton) Wastewater contaminated Photo-Fenton Photo-Fenton removed up to 99 % of TOC. UV with diesel oil photolysis alone and the thermal Fenton reaction resulted in TOC reduction of 28 and 26 %, respectively. UV/H2O2 removed 71 % of TOC, but required a significant irradiation time. Petroleum refinery wastewater TiO2/UV A maximum COD reduction of more than 90 % was achieved after about 4 h. The analysis of the contained materials showed that the efficiency of the degradation process is high for all the identified organic pollutants. Produced water from petroleum O3/UV/TiO2 and biological After 5 min of oxidation treatment, the removal of phenol, refineries remediation by sulfide, COD, oils and grease, and ammonium reached macroalgae 99.9, 53.0, 37.7, 5.2, and 1.9 %, respectively. After 60 min of reaction, their respective removal was observed to be 99.9, 97.2, 89.2, 98.2, and 15 %. Petrochemical refinery Fe(III)/ H2O2/Solar-UV TOC: 820 mg/L to 1385 mg/L, low biodegradability. wastewater Maximum TOC reduction was 49 % with the addition of 2677 mg/L H2O2 and 0.5 mM Fe(III) at a 10 L/h flowrate after 8 h of exposure to solar irradiation. AOP removed 31–79 % of UV254, 10–18 % of TOC. Oil refinery Wastewater H2O2/UV or O3/UV coupled with biological BAC filters showed to be effective, reaching average activated carbon (BAC) efficiencies of 65 % in a sufficiently long period of operation. Effluent TOC in the range of 4 to 8.5 mg/L were achieved by the combined H2O2/UV + BAC treatment Petroleum refinery wastewater UV/H2O2 Initial TOC = 42 mg/L and H2O2/TOC molar ratios = 2.9, 5.8, and 8.6 Toc removal of 78.38 % Petroleum refinery wastewater

Hybrid Fenton and sequencing batch reactor

45 % of the TOC was removed by Fenton (186.1 to 102.4 mg/L) and 37.8 % in the subsequent SBR (102.4 to 63.7 mg/L)

Reference Stepnowski et al. (2002)

Moraes et al. (2004)

Coelho et al. (2006)

Galvão et al. (2006)

Saien and Nejati (2007)

Corrêa et al. (2010)

Parilti et al. (2010)

Souza et al. (2011)

Bustillo-Lecompte et al. (2015) Diya’uddeen et al. (2015)

Environ Sci Pollut Res

Post-treatment of biofilter effluent by microfiltration Microfiltration was employed as an alternative for the H2O/ UVAOP (Fig. S2). However, the results were not satisfactory. The removal of both TOC and UV254 was marginal, since the majority of the organic matter was present in the dissolved form, being therefore retrieved in the permeate of the microfiltration unit. Furthermore, conductivity remained constant before and after microfiltration, as soluble salts are not removed in this process. The results obtained in this set of experiments are summarized in Table S3. Silt density index tests Aiming at further treatment by RO to produce a final effluent suitable for reuse, the determination of SDI was conducted. Wastewaters showing high SDI may cause irreversible fouling in RO membranes (Baker 2004). Therefore, they should undergo a pretreatment step before being fed to an RO system. An SDI lower than 1 means that the RO plant can operate for several years without fouling problems. When this index assumes values between 1 and 3, the RO unit can be run for many months without the need for constant cleaning. On the other hand, SDI between 3 and 5 implies that the RO system is very susceptible to particle incrustation and cleaning procedures are regularly required. SDIs higher than 5 are not satisfactory and a pretreatment must be carried out before the RO system (Baker 2004). In this study, the SDI was evaluated at different stages of the treatment process: before biofiltration with and without microfiltration post-treatment, after biofiltration, after biofiltration + microfiltration, after biofiltration + UV radiation, and finally after biofiltration + H2O2/UV oxidation. SDI values obtained with the industrial wastewater before biofiltration (both with and without microfiltration posttreatment) were higher than 5 and therefore not suitable for RO application. The results obtained with microfiltrated effluent before biofiltration raises an important question which should be discussed. The removal of suspended solids was not enough to prevent membrane clogging and permeate flux reduction. In fact, the residual soluble components present in the effluent matrix were triggering the bad functioning of the membrane permeation system. It was also observed that, despite the considerable removal of residual organic compounds achieved by means of biofiltration, SDI values obtained in the tests carried out with the biofilter effluent amounted to around 4. Such result is probably related to the fact that the soluble microbial products may have arisen from the biofilm established in the biofilter media, affecting the membrane permeation characteristics and therefore the SDI obtained. Thus, a further treatment step was necessary to reduce the SDI to an acceptable value and therefore enable the effluent to be used as feed for a subsequent RO

step. Conversely, the combined biofiltration/microfiltration process ensured SDI values lower than 2 (data not shown). Nevertheless, the significant drop in the permeate flux observed during the microfiltration of biofilter effluent led to a periodic membrane change and would potentially limit the applicability of this treatment strategy at larger scales. Furthermore, post-treatment of the biofilter effluent by simply UV irradiation did not cause a reduction of SDI, which remained within the range of 4–5. It is known that the degradation rates with UV alone are generally lower when compared to other UV processes, such as the combined UV/ H2O2 (Esplugas et al. 2002). Indeed, when the biofilter effluent was subjected to UV/H2O2 treatment, efficient SDI reduction was observed. This is particularly valid for the tests conducted at C/H2O2 ratios of 1:2 and 1:4, for which the SDI varied between 3.6 and 2.4. The unsatisfactory SDI values obtained in the test employing a C/H2O2 ratio of 1:1 are possibly linked to the relatively low TOC removal (11 %) achieved under these conditions (Fig. 5). In this study, the best performance in terms of TOC and UV254 reduction was attained when the C/H2O2 ratio was set at 1:4, for which the highest SDI reduction was also observed. Nevertheless, all the aforementioned results of SDI were obtained after exposing the wastewater to 60min reaction with H2O2/UV. In order to observe the reduction of SDI as a function of the reaction time and its relation to the TOC removal, additional experiments were carried out and the following conditions were chosen: C/H2O2 of 1:4–55-W UV lamp and C/H2O2 of 1:4–95-W UV lamp. The comparison between the two tests is displayed in Fig. 6. A significant SDI reduction is observed within 20 min of oxidation reaction, after which an SDI of around 1 (i.e., appropriate for reverse osmosis application) was obtained. Subsequently, no noticeable decrease in SDI was observed. Similarly, more than 50 % of the TOC was removed in the first 20 min of reaction. This means that a period of 20 min of oxidation is long enough to obtain desirable effluent quality (low TOC and SDI), and

Fig. 5 SDI values obtained after H2O2/UV oxidation applying C/H2O2 ratios of 1:2 e 1:4 and UV lamps of 55 and 95 W

Environ Sci Pollut Res 60

6

24

4

40

2

20

0

0 0

20

40

16

H2O2 (mg/L)

(a)

UV dose (kJ/L)

Cumulative TOC removed (mg/L); SDI

Fig. 6 SDI values (white triangle) and cumulative TOC removal (black squares connected by lines) obtained over the course of the H2O2/UV oxidation experiment in which the C/H2O2 ratio was fixed at 1:4 and the power of the lamp and UV irradiance was either 55 W/ 26.8 mW/cm2 (a) or 95 W/ 46.3 mW/cm2 (b). The UV dose (black square) and H2O2 concentration (white circle) are displayed on the right y-axis

8

0

60

6

100

24

(b)

4 60

40

2

16

8

H2O2 (mg/L)

80

UV dose (kJ/L)

Cumulative TOC removed (mg/ L); SDI

Time (min)

20

0

0

0

20

40

60

0

Time (min)

the extra energy spent from 20 to 60 min of reaction can be avoided. Moreover, similar results were observed for both UV lamps (55 and 95 W). Therefore, for economic reasons, the 55-W lamp is preferable.

the non-oxidized effluent. This finding emphasizes the importance of applying an advanced oxidation process to reduce

Reverse osmosis experiments After choosing the most appropriate conditions for the H2O2/ UV oxidation of the biofilter effluent to achieve satisfactory SDI values (i.e., C/H2O2 ratio of 1:4, 55-W UV lamp, and 20min reaction), an RO system was used for final polishing of the effluent, aiming at its reuse in the oil industry facility. The biofilter effluent without oxidation treatment was also fed to the RO system for comparison purposes. The permeation test conducted in the RO unit lasted for 25 h, and the results are displayed in Fig. 7. At the end of the experiment, the permeate flux obtained with the biofilter effluent post-treated by H2O2/ UVoxidation was around 35 % higher than that obtained with

Fig. 7 Permeate flux over time obtained in the reverse osmosis experiments carried out with the biofilter effluent with and without POA post-treatment

Environ Sci Pollut Res

SDI values of a biologically pretreated wastewater and therefore minimize fouling problems in RO applications. However, before implementation at larger scales, further tests should be conducted on a longer time span, given that the H2O2/UV oxidation may lead to an increased amount of easily assimilable organic carbon compounds, thereby potentially inducing the development of biofilm on the RO membrane in the longterm. The characteristics shown by the final effluent downstream of the RO step (NH4 < 10 mgN/L; COD < 50; negligible conductivity/salinity) (EPA 2004, 2012; Lahnsteiner 2007) made it suitable to be employed in reuse applications such as highpressure boilers and cooling towers.

Conclusions The application of biofiltration coupled to H2O2/UVoxidation allowed for significant reduction of the residual organic carbon from an oil refinery wastewater previously treated by biological processes. TOC abatement and UV254 reduction achieved at UV irradiance levels of 26.8 mW/cm2 (for 55-W lamps) and 46.3 mW/cm2 (for 95-W lamps) were very comparable, suggesting that there is no need for the use of higherpower lamps. However, the concentration of hydrogen peroxide greatly influenced the amount of TOC removal and UV254 reduction. Oxidation by H2O2/UV was effective for SDI reduction, particularly when the C/H2O2 ratio was set at 1:2 and 1:4. Taking into account the best cost/efficiency choice, the most appropriate conditions for TOC, UV254, and SDI reduction were determined to be the C/H2O2 ratio of 1:4, UV lamp of 55 W, and 20-min oxidation reaction. The sequence of treatment proposed in this study is suitable for the non-stop operation of a reverse osmosis system in the long-term by preventing the occurrence of membrane fouling and contributes to produce an effluent adequate for reuse in the industry. Acknowledgments The authors would like to express their gratitude to Capes and CNPq for providing the financial support necessary to develop this research. Great thanks also go to Petrobras which provided the wastewater source, object of this study.

References Abdelwahab O, Amin NK, El-Ashtoukhy ESZ (2009) Electrochemical removal of phenol from oil refinery wastewater. J Hazard Mater 163: 711–716 Andreozzi R, Caprio V, Insola A, Marotta R (1999) Advanced oxidation processes (AOP) for water purification and recovery. Catal Today 53:51–59 Andreozzi R, Caprio V, Marotta R, Vogna D (2003) Paracetamol oxidation from aqueous solutions by means of ozonization and UV/H2O2 system. Water Res 37:993–1004

Antonopoulou M, Evgenidou E, Lambropoulou D, Konstantinou I (2014) A review on advanced oxidation processes for the removal of taste and odor compounds from aqueous media. Water Res 53:215–234 APHA, AWWA, WEF (2005) Standard methods for the examination of water and wastewater, 21st edn. American Public Health Association, American Water Works Association, and Water Environment, Washington DC ASTM D4189-07 (2014) Standard test method for silt density index (SDI) of water. ASTM International, West Conshohocken. doi:10. 1520/D4189 Baker RW (2004) Reversis osmosis—membrane fouling control. In: Membrane technology and applications, 2nd edn. John Wiley & Sons Ltd, California, pp 217–219 Bustillo-Lecompte CF, Knight M, Mehrvar M (2015) Assessing the performance of UV/H2O2 as a pretreatment process in TOC removal of an actual petroleum refinery wastewater and its inhibitory effects on activated sludge. Can J Chem Eng 93:798–807 Chavan A, Mukherji S (2008) Treatment of hydrocarbon-rich wastewater using oil degrading bacteria and phototrophic microorganisms in rotating biological contactor: effect of N:P ratio. J Hazard Mater 154:63–72 Coelho A, Castro AV, Dezotti M, Sant’Anna GL Jr (2006) Treatment of petroleum refinery sourwater by advanced oxidation processes. J Hazard Mater B137:178–184 Comninellis C, Kapalka A, Malato S, Parsons SA, Poulios I, Mantzavinos D (2008) Advanced oxidation processes for water treatment: advances and trends for R&D. J Chem Technol Biotechnol 83:769–76 Corrêa AXR, Tiepo EN, Somensi CA, Sperb RM, Radetski CM (2010) Use of ozone-photocatalytic oxidation (O3 /UV/TiO2) and biological remediation for treatment of produced water from petroleum refineries. J Environ Eng 136(1):40–45 Diya’uddeen BH, Pouran SR, Aziz ARA, Nashwan SM, Daud WMAW, Shaaban MG (2015) Hybrid of Fenton and sequencing batch reactor for petroleum refinery wastewater treatment. J Ind Eng Chem 25: 186–191 Dos Santos VL, Veiga AA, Mendonça RS, Alves AL, Pagnin S, Santiago VM (2005) Reuse of refinery’s tertiary-treated wastewater in cooling towers: microbiological monitoring. Environ Sci Pollut Res 22: 2945–2955 El-Naas MH, Al-Zuhair S, Al-Lobaney A, Makhlouf S (2009) Assessment of electrocoagulation for the treatment of petroleum refinery wastewater. J Environ Manag 91:180–185 EPA—Environmental Protection Agency (2004) Guidelines for Water Reuse. EPA/625/R-04/108, Washington, DC EPA—Environmental Protection Agency (2012) Guidelines for Water Reuse. EPA/600/R-14/618, Washington, DC Esplugas S, Gimenez J, Contreras S, Pascual E, Rodríguez M (2002) Comparison of different advanced oxidation processes for phenol degradation. Water Res 36:1034–1042 Galvão SAO, Mota ALN, Silva DN, Moraes JEF, Nascimento CAO, Chiavone-Filho O (2006) Application of the photo-Fenton process to the treatment of wastewaters contaminated with diesel. Sci Total Environ 367:42–49 Harremoës P (1982) Criteria for nitrification in fixed film reactors. Water Sci Technol 14:167–187 Hozalski RM, Goel S, Bouwer EJ (1995) TOC removal in biological filters. J AWWA 87(12):40–54 Hu X, Wang X, Ban Y, Ren B (2011) A comparative study of UV-Fenton, UV-H2O2 and Fenton reaction treatment of landfill leachate. Environ Technol 32:945–951 Huang WJ, Chen LY, Peng HS (2004) Effect of NOM characteristics on brominated organics formation by ozonation. Environ Int 29:1049– 1055 Kjärstad J, Johnsson F (2014) Fossil fuels: climate change and security of supply. Int J Sustainable Water Environ Syst 4:79–87

Environ Sci Pollut Res Lahnsteiner J, Klegraf F, Mittal R, Andrade P (2007) Reclamation of wastewater for industrial purposes—advanced treatment of secondary effluents for reuse as boiler and cooling make-up water. 6th IWA Specialist Conference on Wastewater Reclamation and Reuse for Sustainability, 1–8 Lamsal R, Walsh ME, Gagnon GA (2011) Comparison of advanced oxidation processes for the removal of natural organic matter. Water Res 45:3263–3269 Li K, Hokanson DR, Crittenden JC, Trussell RR, Minakata D (2008) Evaluating UV/H2O2 processes for methyl tert-butyl ether and tertiary butyl alcohon removal: effect of pretreatment options and light sources. Water Res 42:5045–5053 Ma F, Guo JB, Zhao LJ, Chang CC, Cui D (2009) Application of bioaugmentation to improve the activate sludge system into the contact oxidation system treatment petrochemical wastewater. Bioresource Technol 100:597–602 Madoni P (1994) Quantitative importance of ciliated protozoa in activated sludge and biofilm. Bioresource Technol 48:245–249 Mazzeo DEC, Levy CE, Angelis DF, Marin-Morales MAM (2010) BTEX biodegradation by bacteria from effluents of petroleum refinery. Sci Total Environ 20:4334–4340 Melin ES, Odegaard H (1999) Biofiltration of ozonated humic water in expanded clay aggregate filters. Water Sci Technol 40:165–172 Melin ES, Bohne RA, Sjovold F, Odegaard H (2000) Treatment of ozonated water in biofilters containing different media. Water Sci Technol 41:57–60 Metcalf X, Eddy X (2003) Wastewater engineering, treatment, disposal and reuse. In: Techobanoglous G, Burton FL, Stensel HD (ed), 4th edn. McGraw-Hill Book, New York Moraes JEF, Silva DN, Quina FH, Chiavone-Filho O, Nascimento CAO (2004) Utilization of solar energy in the photodegradation of gasoline in water and of oil-field-produced water. Environ Sci Technol 38:3746–3751 Morgenroth E, Wilderer PA (2000) Influence of detachment mechanisms on competition in biofilms. Water Res 34:417–426 Munter R (2001) Advanced oxidation processes—current status and prospects. Proc Est Acad Sci Chem 50(2):59–80 Muruganandham M, Suri RP, Sillanpää M, Wu JJ, Ahmmad B, Balachandran S, Swaminathan M (2014) Recent developments in heterogeneous catalyzed environmental remediation processes. J Nanosci Nanothecnol 14:1898–1910 Nogueira RFP, Oliveira MC, Paterlini WC (2005) Simple and fast spectrophotometric determination of H2O2 in photo-Fenton reactions using metavanadate. Talanta 66(1):86–91 Oller I, Malato S, Sánchez-Pérez JÁ (2011) Combination of advanced oxidation processes and biological treatment for wastewater decontamination—a review. Sci Total Environ 20:4141–4166 Parilti NB (2010) Treatment of a petrochemical industry wastewater by a solar oxidation process using the Box-Wilson experimental design method. Ekoloji 19(77):9–15

Raffin M, Germain E, Judd S (2013) Wastewater polishing using membrane technology: a review of existing installations. Environ Technol 34:617–627 Reungoat J, Escher BI, Macova M, Keller J (2011) Biofiltration of wastewater treatment plant effluent: effective removal of pharmaceuticals and personal care products and reduction of toxicity. Water Res 45: 2751–2762 Reungoat J, Escher B, Macova M, Farré MJ, Argaud FX, Rattier M, Dennis PG, Gernjak W, Keller J (2012) Biofiltration for advanced treatment of wastewater. Urban Water Security Research Alliance Technical Report 73:1–45 Ried A, Mielche J, Kampmann M (2006) The right treatment step ozone and ozone/H2O2 for the degradation of none-biodegradable COD. Proceeding, Wasser Berlin 2006: International Conference Ozone and UV, April 3rd pp 25–33 Saien J, Nejati H (2007) Enhanced photocatalytic degradation of pollutants in petroleum refinery wastewater under mild conditions. J Hazard Mater 148:491–495 Santos MRG, Goulart MOF, Tonholo J, Zanta CLPS (2006) The application of electrochemical technology to the remediation of oily wastewater. Chemosphere 64:393–399 Schneider EE, Cerqueira ACFP, Dezotti M (2011) MBBR evaluation for oil refinery wastewater treatment, with post-ozonation and BAC, for wastewater reuse. Water Sci Technol 63:143–148 Souza BM, Cerqueira AC, Sant’Anna GL Jr, Dezotti M (2011) Oilrefinery wastewater treatment aiming reuse by advanced oxidation processes (AOPs) combined with biological activated carbon (BAC). Ozone Sci Eng 33:403–409 Stepnowski P, Siedleckaa EM, Behrendb P, Jastorff B (2002) Enhanced photo-degradation of contaminants in petroleum refinery wastewater. Water Res 36:2167–2172 Sun Y, Zhang YB, Quan X (2008) Treatment of petroleum refinery wastewater by microwave-assisted catalytic wet air oxidation under low temperature and low pressure. Sep Purif Technol 62:565–570 Wert EC, Rosario-Ortiza FL, Druryb DD, Snyder SA (2007) Formation of oxidation by products from ozonation of wastewater. Water Res 41:1481–1490 Wijeyekoona S, Minob T, Satohb H, Matsuo T (2004) Effects of substrate loading rate on biofilm structure. Water Res 38:2479–2488 Zhao Y, Boyd JM, Woodbeck M, Andrews RC, Qin F, Hrudey SE, Li X (2008) Formation of N-nitrosamines from eleven disinfection treatments of seven different surface waters. Environ Sci Technol 42: 4857–4862 Zhidong L, Yebin D, Xincheng X (2010) An aerated biofilter for treating petrochemical wastewater. Pet Sci Technol 28:1147– 1157 Zoschke K, Dietrich N, Börnick H, Worch E (2012) UV-based advanced oxidation processes for the treatment of odour compounds: efficiency and by-product formation. Water Res 46:5365–5373