Factors affecting on hythane bio-generation via anaerobic digestion of mono-ethylene glycol contaminated wastewater: Inoculum-to-substrate ratio, nitrogen-to-phosphorus ratio and pH

Accepted Manuscript Factors affecting on hythane bio-generation via anaerobic digestion of monoethylene glycol contaminated wastewater: inoculum-to-substrate ratio, nitrogen-to-phosphorus ratio and pH Ahmed Elreedy, Manabu Fujii, Ahmed Tawfik PII: DOI: Reference:

S0960-8524(16)31432-8 http://dx.doi.org/10.1016/j.biortech.2016.10.026 BITE 17181

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

4 August 2016 6 October 2016 9 October 2016

Please cite this article as: Elreedy, A., Fujii, M., Tawfik, A., Factors affecting on hythane bio-generation via anaerobic digestion of mono-ethylene glycol contaminated wastewater: inoculum-to-substrate ratio, nitrogen-tophosphorus ratio and pH, Bioresource Technology (2016), doi: http://dx.doi.org/10.1016/j.biortech.2016.10.026

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Factors affecting on hythane bio-generation via anaerobic digestion of mono-ethylene glycol contaminated wastewater: inoculum-to-substrate ratio, nitrogen-to-phosphorus ratio and pH Ahmed Elreedy a,b,*, Manabu Fujii a, Ahmed Tawfik b a

Department of Civil and Environmental Engineering, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8552, Japan b

Environmental Engineering Department, Egypt-Japan University of Science and

Technology (E-JUST), P.O. Box 179, New Borg Al-Arab City, Postal Code 21934, Alexandria, Egypt Abstract This study aims to assess the effect of inoculum-to-substrate ratio (ISR) and nitrogento-phosphorus balance on hythane production from thermophilic anaerobic decomposition of mono-ethylene glycol (MEG) contaminated wastewater. ISRs ranging from 2.65 to 13.23 gVSS/gCOD were employed, whereas the tested N/P ratios varied from 4.6 to 8.5. Maximum methane and hydrogen yields (MY and HY) of 151.86±10.8 and 22.27±1.1 mL/gCODinitial were achieved at ISRs of 5.29 and 3.78 gVSS/gCOD, respectively. HY increased 1.45-fold by decreasing N/P from 8.5 to 4.6, while MY improved 1.6-fold by increasing N/P from 4.6 to 5.5. Methane production was strongly influenced by initial NH4N, compared to initial PO4-P. Optimal HY of 47.55 mL/gCODinitial was achieved at pH 5.0 and ISR of 3.78 gVSS/gCOD using thermal-treated sludge. Three-dimensional regression model was applied for the combined effect of initial MEG, NH4-N and PO4-P on hythane

*

Corresponding author: (A. Elreedy) [email protected], and [email protected]; (A. Tawfik) [email protected]

1

production. Potential economic benefits of hythane production from MEG contaminated wastewater were assessed. Keywords: Mono-ethylene glycol; Hydrogen and methane; Nitrogen-to-phosphorus balance; Regression model; Energy recovery 1. Introduction Mono-ethylene glycol (MEG) is one of the important contaminants produced in the manufacturing processes of several petrochemical industries, such as polyethylene terephthalate (PET), engine coolants, aircraft deicing fluid, natural gas extraction and polyester fibers (Elreedy et al., 2016). Discharge of this pollutant without proper treatment is of serious environmental concern given its persistency in the aqueous environment. Treatment of MEG-contaminated wastewater have been previously investigated by aerobic treatment (Hassani et al., 2013), direct and catalytic oxidations (Kim and Hoffmann, 2008; Zerva et al., 2003), nano-filtration (Orecki et al., 2006) and vacuum membrane distillation (Mohammadi and Akbarabadi, 2005). Although, these technologies are effective for treatment of such wastewater, they require substantial energy consumption. Utilization of the organic matters present in wastewater, including MEG, for the generation of bioenergy (i.e., H2, CH4 and bio-ethanol) via anaerobic digestion (AD) process is a promising alternative renewable energy source due to the depletion of fossil fuels. Indeed, the lower mass heating values for H2 and CH4 have been reported to be 120 and 50 MJ/kg, respectively (Elreedy and Tawfik, 2015). So far, few studies have been conducted on the AD of petrochemical industry wastewater. Formerly, 75% COD removal and an average methane yield (MY) of 300 mL/gCOD removed were achieved for treatment of diluted MEG-based aircraft deicing fluid 2

(ADF) (Marin et al., 2010). Another study by Elreedy et al. (2015) used stepped anaerobic baffled reactor, inoculated with mixed culture bacteria, for hythane production from MEGcontained wastewater. They found that maximum H2 and CH4 yields of 98 and 118 mL/gCODinitial were achieved at organic loading rates (OLRs) of 1.67 and 0.67 gCOD/L/d, respectively. Higher values of H2 and CH4 yields of 122 and 170 mL/gCODinitial was obtained in an anaerobic packed bed baffled reactor at OLRs of 2.0 and 0.67, respectively (Elreedy et al., 2016). Besides, the hythane production via two-stage AD have been previously investigated, using different substrates (e.g., waste biomass, wheat straw, glucose), considering various operating parameters, reactor configurations and using pure cultures (Willquist et al., 2012; Liu et al., 2014, 2013; Si et al., 2016). Nonetheless, much less attention has been given to the factors influencing hythane production from the MEGrich wastewater. The balance between inoculum and substrate (i.e., ISR) is considered as crucial parameter in stabilizing the AD processes including extracellular hydrolysis and methane production with minimum value of 1 gVS/gCOD being reported (Ali Shah et al., 2014). Darlington and Kennedy (1998) found that 85% COD removal was achieved from ADF at ISR of 3.7 gVSS/gCOD. Maximum H2 and CH4 yields were recorded at ISR of 6.4 and 16 gVS/gCOD, respectively, using MEG as a substrate at ambient temperature (Elreedy et al., 2015). In addition, Ali Shah et al. (2014) indicated that the effect of ISR on the AD performance strongly depends on the inoculum source and substrate composition. Nitrogen and phosphorus are main macro-nutrients affecting AD, especially for thermophilic reactors compared to mesophilic reactors, as the former reactors have higher nutrients demand (Romero-Güiza et al., 2016). Nitrogen is mainly responsible for protein synthesis in the 3

microbial metabolism, whereas phosphorus is used for nucleic acid and membrane synthesis (Mao et al., 2015). The optimum nutrients condition for AD is dependent on substrate and microbial culture compositions (Romero-Güiza et al., 2016). Speece (1983) stated that the required amount of phosphorus is approximately 15% of the nitrogen demand (N/P ratio of 6.67). Minimum values of required nitrogen and phosphorus were determined to be 45 and 8 mg/l, respectively, using petrochemical wastewater containing carboxylic acids (Britz et al., 1988). However, a wide range of COD/N/P (e.g., 400/7/1, 500/5/1 and 200/4/1) ratios has been considered for the AD of different substrates. Furthermore, the initial pH strongly affects the hydrolysis and acidification processes through the maximization of hydrogen production. The optimum initial pH for H2 production mostly ranged from 5.0 to 6.5, depending on the substrate composition as reported in a range of studies (El-Bery et al., 2013; Mao et al., 2015). Since the valorization of wastewater is one of the important aspects in petrochemical industries, optimization of the AD performance and proper understanding of factors affecting on hythane production is essential. Besides, relatively low degradation rate of MEG, under anaerobic conditions, has been previously reported (Elreedy et al., 2016). Inoculum-to-substrate ratio, nitrogen-to-phosphorus (N/P) ratio and initial pH are major factors affecting on hythane production, as aforementioned. However, to our knowledge, optimization of the AD of MEG-contaminated wastewater has not yet been addressed by considering these parameters. Moreover, it is indispensable to properly choose the mixed culture’s origin, which is capable of providing active species responsible for the MEG degradation pathway catalyzed by alcohol dehydrogenase enzyme. In this aspect, the excess sludge from baking yeast industry wastewater treatment has been recognized to have 4

greater activity of alcohol dehydrogenase enzymes, which is responsible for MEG metabolism by baker’s yeast, namely Saccharomyces cerevisiae (Leskovac et al., 2002). Therefore, the main objectives of this study are to: (1) investigate the effect of various ISR and N/P ratios on the simultaneous H2/CH4 production, (2) maximize the H2 production, using thermal-pretreated sludge, at different initial pH values, (3) assess the economic aspects of energy recovery from AD of MEG containing wastewater, and (4) apply a three-dimensional regression model in order to mathematically describe the combined effect of initial MEG, NH4-N and PO4-P on hythane production. 2. Materials and methods 2.1. Inoculum and substrate The seed sludge was collected from the thickener of wastewater treatment plant, in a baking yeast factory. The harvested sludge was settled for 24 h, and the supernatant was withdrawn. The sludge was sieved (2 mm-diameter) to remove any large particles or debris, and then was characterized, as shown in Table S1 (Supplementary). The inoculum was fed by MEG containing wastewater and acclimated for 40 days under anaerobic conditions prior to starting the batch experiments. Synthetic petrochemical wastewater spiked with MEG (C2H6O2) was used in this experiment. The feed stock was supplemented with ammonium chloride (NH4Cl) and potassium dihydrogen phosphate (KH2PO4) as nitrogen and phosphorus sources, respectively, in order to attain a COD:N:P of 400:7:1 (Wahab et al., 2014). Buffer and trace elements solution (Table S1, supplementary) were added to enhance the growth of anaerobic microbes, according to Elreedy et al. (2015). 2.2.Experimental design and procedures

5

Triplicate batch experiments were conducted in serum bottles (working volume of 250 ml and 50 ml headspace). The first experiment was designed to assess the effect of ISR on the hydrogen and methane productions. The synthetic MEG contaminated wastewater compositions and inoculum sludge characteristics are presented in Table S1 (Supplementary). The inoculum sludge was kept constant at a value of 13.23 gVSS/L in all batches. The MEG (C2H6O2) concentrations were 1566, 2349, 3915, 5481, and 7830 mg/L. These concentrations corresponded to COD values of 2000, 3000, 5000, 7000 and 10000 mg/L and ISRs of 13.23, 8.82, 5.29, 3.78 and 2.65 gVSS/gCOD, respectively. The effect of nitrogen-to-phosphorus (N/P) ratio on the sequential hydrogen/methane production at fixed ISR (i.e., 5.29 gVSS/gCOD, which achieved maximum methane production) was investigated in the second experiment where N/P ratios were varied to 4.6, 5.5 and 8.3. These ratios have been selected according to the different standard values (i.e., 5 and 7) used in most of the former studies. The initial pH for all batches in these two experiments was adjusted to 7.0±0.15. The third experiment was carried out to optimize the hydrogen production by using thermal-pretreated sludge (105°C for 30 min (Wimonsong et al., 2014)) and different initial pH values (i.e., 5.0, 5.5, 6.0 and 7.0). The thermal-pretreatment has been proved, using different substrates, to suppress hydrogen consuming and sporeforming bacteria (Farghaly et al., 2015). This experiment was conducted at optimum conditions obtained for hydrogen production; i.e., ISR of 3.78 gVSS/gCOD and N/P ratio of 4.6. Prior to the experiment, pH values of batches were adjusted using sodium hydroxide (NaOH) and hydrochloric acid (HCl). In addition, control batches were run using only inoculum sludge with distilled water. The headspaces of batch cultures were flushed with oxygen-free dinitrogen gas for 3 min and capped with rubber stoppers and aluminum seals. 6

The cultures were then incubated in shakers (Incu-shaker Mini, Benchmark Scientific) operated at rotation rate of 180 rpm and temperature of 55°C. 2.3. Analytical technique Total solids (TS), total suspended solids (TSS), volatile solids (VS), volatile suspended solids (VSS), COD, total Kjeldahl nitrogen (TKN), ammonium-nitrogen (NH4-N) and total volatile fatty acids as acetate (TVFAs) were determined according to APHA (2005). The protein content (mg/L) was calculated according to the following equation: (6.25 x (TKN – NH4-N)). The total and soluble chemical oxygen demand (CODtotal and CODsoluble) and total phosphorus (TP) were measured using HACH procedures and test kits (Hach DR/900). All the soluble samples were determined after the filtration with membrane filter (0.45-µm Whatman, Japan). Concentrations of soluble metabolite products such as acetate, butyrate, propionate and ethanol (EtOH) were measured using high-performance liquid chromatography (HPLC) (LC- 10AD, Shimadzu, Japan) equipped with UV detector and a Shim-pack HPLC column (4.6×250 mm, VP-ODS, Vertical). The temperature of column oven was adjusted to 40 °C. Sulfuric acid (H2SO4, 4 mM) was used as a mobile phase at a flow rate of 0.5 ml/min for 22 min followed by 0.4 ml/min for 8 min. The total biogas volume was measured by water displacement method, and the gas volumes were normalized to standard conditions (temperature of 25°C and pressure of 1 atm). Then, the gas samples were collected for the analysis of composition using a gas chromatograph (GC2014, Shimadzu, Japan) equipped with a thermal conductivity detector (TCD) and Shin carbon column. The operational temperatures of the injection port, the column oven, and

7

the detector were 100, 120, and 150 °C, respectively. Helium gas was used as carrier at a flow rate of 25 ml/min. 2.4. Mathematical model The modified Gompertz equation has been used to separately describe the cumulative hydrogen and methane production in batch experiments (Elbeshbishy and Nakhla, 2012):

 R e  H = P ⋅ exp -exp  m ( λ-t ) +1  ,  P  

(1)

where H (mL) is the cumulative hydrogen/methane production recorded at time t (h), P (mL) is the hydrogen/methane potential, Rm (mL/h) is the maximum hydrogen/methane production rate, e = exp (1) = 2.718, and λ (h) is the lag-phase period. Using 3D Gaussian function, statistical regression of the experimental results for the cumulative hydrogen and methane production (HP and MP) rates was considered with two independent variables; initial MEG concentrations (x) and either initial NH4-N or PO4-P concentrations (y), as follow in Eq. 2.

  x-x 0  2  y-y 0  2   Z = a ⋅ exp -0.5   +   ,  b   c    

(2)

where Z is the dependent variable (cumulative HP and MP rates, mL/Lsubstrate), a is the maximum value of Z (the peak of the equation curve "bell"), b and c are coefficients to control the bell width, x0 and y0 represent the position of the peak value of Z on X and Yaxis, respectively. The fitness of this model was evaluated using analysis of variance (ANOVA) and the significance (p-value) was considered according to a 95% confidence level. The statistical analyses, in addition to the response surface plots were carried out using SigmaPlot 10.0 software. 8

3. Results and discussion 3.1. Effect of inoculum-to-substrate ratio (ISR) on simultaneous H2/CH4 production Figs. 1a, b and c present the time course of cumulative total biogas, hydrogen production (HP) and methane production (MP) at different ISRs, respectively. The results obtained indicated that two successive biodegradation processes (i.e., acidification and methanogenesis) occurred in the batch reactors. These two processes are classified to: (i) conversion of MEG into bio-ethanol (EtOH) and acetate (HAc) (Eq. 3), (ii) degradation of ethanol to acetate and hydrogen (Eq. 4) and, (iii) consumption of hydrogen and acetate yielding methane (i.e., methanogenesis, Eqs. 5-6), as described below: 4HO-CH2-CH2-OH (MEG) → 2CH3CH2OH (Ethanol) + 2CH3COOH (Acetate),

(3)

2CH3CH2OH (Ethanol) + 2H2O → 2CH3COOH (Acetate) + 4H2,

(4)

4H2 + CO2 → CH4 + 2H2O,

(5)

CH3COOH (Acetate) + 2H2O → CH4 + CO2

(6)

Figs. 1 and 3 show that hydrogen was initially generated from day 1 to 12 and methane was subsequently progressed from day 12 to 36. Dwyer and Tiedje (1983) found similar trends where methane production progressed after the complete degradation of MEG into bio-ethanol and acetate. Whereas, cumulative methane contents up to 18-20% were observed on day 11, as revealed in Figs. 1 and 3. This is mainly due to the activity of acetotrophic methanogenesis, which is favorable at 55-60°C and pH 7.0 as compared to hydrogenotrophic methanogenesis (Demirel and Scherer, 2008). Moreover, Si et al., (2016) proved the prevalence of acetotrophic methanogenesis in one-stage AD for hythane production.

9

As can be seen in Figs. 1a, b and c, hythane (H2 and CH4) production and metabolite products were found to be ISR dependent. The cumulative MP significantly increased from 97±7 to 290±18 mL at decreasing ISR from 13.23 to 5.29 gVSS/gCOD, respectively (Fig. 1c). Nevertheless, the cumulative MP substantially dropped to 12±2 mL at ISR exceeding 2.65 gVSS/gCOD. This decrement can be attributed to the substrate over-loading condition that suppresses the methanogenesis; moreover, inadequate ISR created unfavorable conditions for the consortium. Likely, Sreela-Or et al. (2011) indicated that microorganisms’ consortium under the inadequate ISR condition are not capable of competing with indigenous microflora, resulting in the incomplete substrate conversion processes. The high residual of soluble metabolites were coincided as well, to the minimal MP at ISR of 2.65 gVSS/gCOD, which was consistent with the negative effect of substrate over-loading on the methanogenesis. MP decreased at ISRs of 8.82 and 13.23 gVSS/gCOD where excessive ammonia-nitrogen was released. The produced ammonia substantially increased up to 493.95±44.33 and 515.50±47.94 mgNH4-N/L at ISRs of 8.82 and 13.23 gVSS/gCOD, respectively, as depicted in Fig. 2d. High concentrations of ammonia negatively affect the growth of methanogenic bacteria (Chen et al., 2008). Consistently, Elbeshbishy and Nakhla (2012) found that MP declined by up to 32%, as final NH4-N concentration increased from 310 to 720 mg/L. The results in Fig. 1b indicate that the cumulative HP increased from 19±2 to 48±3 mL at decreasing ISR from 13.23 to 3.78 gVSS/gCOD, respectively. The cumulative HP dropped from 48±3 to 27±2 mL at decreasing ISR from 3.78 to 2.65 gVSS/gCOD. Better performance for hydrogenproducing bacteria was achieved at low ISR of 3.78 gVSS/gCOD as compared to methanogenesis (i.e., 5.29 gVSS/gCOD). These results indicated that the hythane 10

production rates are substantially affected by inoculum concentration and substrate availability, which might not be sufficient to achieve optimal microbial activities. Similar trends were achieved by Farghaly et al. (2015). The maximum hydrogen and methane contents were 40.91 and 54.51% at ISR of 2.65 and 5.29 gVSS/gCOD, respectively. Comparable H2 and CH4 contents were reported for biohythane production from wheat straw (Willquist et al., 2012). Whereas, lower hydrogen contents (ranging from 11 to 15%) were attained at ISRs from 3.78 to 13.23 gVSS/gCOD (Figs. 1 and 3). This can be due to the H2/CO2 methane pathway (Eq. 5) by hydrogenotrophic methanogenesis. These results are consistent with Liu et al. (2013) who reported that hydrogen content in hythane ranges from 10 to 25%. As depicted in Table S2 (Supplementary), the obtained data for hydrogen and methane production were successfully modeled using Gompartz equation (Eq. 1) (R2 ranged from 0.9617 to 0.9981). The model indicated that maximum hydrogen and methane production rates (Rm) of 0.42 and 1.33 mL/h can be achieved at ISR of 3.78 and 5.29 gVSS/gCOD, respectively. At the highest ISR of 13.23 gVSS/gCOD, lower production rates (Rm) of 0.21 and 0.39 mL/h were recorded for hydrogen and methane production, respectively, indicating the adverse effect of high ISR on the rate of MEG metabolism. The results in Fig. 2a show that maximum methane yield (MY) of 151.86±10.8 mL/gCODinitial (289.45±17.9 mL/gMEGinitial), was registered at ISR of 5.29 gVSS/gCOD. Then, reducing ISR up to 2.65 gVSS/gCOD caused a large drop in MY to 4.71±0.3 mL/gCODinitial (8.04±0.6 mL/gMEGinitial). The minimal MY at the lowest ISR is likely associated with the high production of TVFAs (mainly acetate) of 1803.94±84.1 mg/L, which resulted in a pH value drop to 5.35±0.24 creating unfavorable conditions for 11

methanogenesis (Fig. 2b). Consistent with Eqs. 5-6, low residual values of acetate (HAc) and ethanol (EtOH) were recorded at the maximum MY obtained in this study (i.e., 46.24 ± 2.7 for HAc and 117.9 ± 7.5 mg/L for EtOH), indicating the prevalence of acetotrophic methanogens that convert acetate into methane gas. The COD removal efficiency increased from 61.46±3.77 to 80.02±5.90% at decreasing ISR from 13.23 to 3.78 gVSS/gCOD, respectively as shown in Fig. 2b, At ISR of 2.65 gVSS/gCOD, the COD removal efficiency largely dropped to 28.98±1.42% due likely to the accumulation of VFAs. The maximum H2 yield (HY) of 22.27±1.1 mL/gCODinitial (40.22±2.5 mL/gMEGinitial) was obtained at ISR of 3.78 gVSS/gCOD, which declined further to 10.82±0.7 mL/gCODinitial (18.46±1.2 mL/gMEGinitial) at the lowest ISR of 2.65 gVSS/gCOD. The maximum H2 content of 40.91% was recorded at the latter ISR as a result of methanogens inhibition. Higher ISR of 6.0 gVS/gCOD was stated for maximum HY of 50.29 mL/gCODinitial, using paperboard mill wastewater (Farghaly et al., 2015). This is also due to the faster ethanol conversion to acetate and H2, resulting in the lowest ethanol concentration of 110.30±7.2 mg/L (Eq. 4). The HY (based on COD conversion) substantially increased from 19.62±1.06 to 37.34±2.51 mL/gCODremoved with decreasing ISR from 13.23 to 2.65 gVSS/gCOD (Fig. 2b), respectively. The maximum HY was achieved at ISR of 2.65 gVSS/gCOD, which is mainly attributed to the decrease in COD removal efficiency caused by the acidification process. The maximum MY of 208.62±13.72 mL/gCODremoved was recorded at ISR of 5.29 gVSS/gCOD. The pH value of 7.09±0.08 was almost kept constant at the experiments where ISR was varied from 3.78 to 13.23 gVSS/gCOD (Fig. 2c). The relatively stable pH is likely due to the rapid consumption of VFAs, in addition to the increase in the buffering capacity during 12

ammonia release (Fig. 2d). These results are comparable to those obtained by Raposo et al. (2009), where no significant pH variation (from 7.1 to 7.6) was observed at ISR ranging from 0.8 to 3.0 gVS/gCOD. At ISR of 2.65 gVSS/gCOD, the pH value dropped to 5.35±0.2 (Fig. 2c), which can be attributed to the VFAs production with insufficient buffering capacity to attain a suitable environment for acetate-consuming bacteria. This pH drop was concomitant with the low ammonification efficiency at ISR of 2.65 gVSS/gCOD (Fig. 2d). Fig. 2d shows the effect of different ISRs on the ammonification process, considering initial protein content of 3.32±0.23 g/L, introduced by the constant initial inoculum concentration of 13.23 gVSS/L. The NH4-N release (after subtracting the initial NH4-N) substantially increased from 295.50±17.33 to 515.50±31.05 mg/L at increasing ISR from 2.65 to 13.23 gVSS/gCOD, respectively. Since the initial organic nitrogen was constant (i.e., 531±36 mgN/L) at all the tested ISRs, the ISR variation (normalized by the initial COD) was the main factor affecting the ammonification process. The highest ammonification efficiency was observed at the lowest COD/N ratio of 11.56. Elbeshbishy and Nakhla (2012) found that ammonia release was highest (720 mg/L) at low C/N ratio resulting in low methane production. In addition, Xu et al. (2012) indicated that protein degradation is higher at higher ISRs (as compared to carbohydrate degradation). Concerning the effect of substrate composition, Moset et al. (2015) compared four substrates (i.e., wheat straw, whole crop maize, cattle manure, grass and cellulose) with respect to the biofuels production in AD, and found that the variation in optimum ISR depends on the substrate type. Table 1 summarizes the simultaneous H2 and CH4 production rates from various substrates in previous studies. In this study, hythane production was comparable to that for carbohydrates, protein, and petrochemical-based 13

substrates. Higher H2 and CH4 yields of 0.064 and 0.464 L/gCOD were achieved, as compared to total biogas yield of 0.062 L/gCOD from ethylene oxide wastewater (note that MEG is the main contaminant) (Schonberg et al., 1997). Higher ISR (i.e., 3.0) achieved higher H2/CH4 yields using cellulose as substrate, as compared to glucose, acetate, starch and protein. This highlights the significance of ISR-optimization for hythane production from different substrates. 3.2.Effect of nitrogen-to-phosphorus (N/P) ratio on simultaneous H2/CH4 production The obtained results in Fig. 3b indicated that maximum cumulative HP was dropped from 58±2 to 37±2 mL with increasing N/P ratio from 4.6 to 8.5, respectively. Likewise, Yossan et al. (2012) found that decreasing N/P ratio caused a slight increase in the HP from 958 to 988 mL/Lsubstrate. However, the cumulative HP was quite high at N/P ratio of 4.6, as compared to N/P ratio of 5.5; moreover, this enhancement was higher than the recorded one with changing the ISR. This finding undoubtedly emphasizes the crucial effect of nitrogento-phosphorus balance on bio-hydrogen production from petrochemical wastewater containing MEG. The maximum total biogas production of 532±33 mL was recorded at N/P ratio of 5.5 (Fig. 3a). The cumulative MP was substantially improved by 1.6-fold with increasing N/P ratio from 4.6 to 5.5, resulting in the maximum MP of 290±18 mL (Fig. 3c). The increase in N/P ratio up to 8.5 slightly decreased the cumulative MP providing 274±21 mL. Based on these results, N/P ratios of 4.6 and 5.5 are optimum values for the sequential hydrogen and methane production, which is also consistent with the favorable N/P ratio (i.e., 5.0) for microbial growth (Leite et al., 2016). However, it should be noted that the maximum methane and hydrogen contents of 56.15 and 15.14% were registered at N/P ratios of 8.5 and 4.6, respectively. The hydrogen and methane production was successfully 14

fitted to Gompartz equation (Eq. 1), and maximum HP and MP rates (Rm) of 0.46 and 1.33 mL/h were registered at N/P ratios of 4.6 and 5.5, respectively, as presented in Table S2 (Supplementary). Fig. 4a depicts the variation of HY and MY at the N/P ratios. The results revealed that maximum MY and HY of 151.86±10.8 and 30.50±2.1 mL/gCODinitial (289.45±17.9 and 58.12±13.1 mL/gMEGinitial) were achieved at N/P ratio of 5.5 and 4.6, respectively. The COD removal efficiency significantly increased from 46.73±3.92 to 77.71±4.25%, as N/P ratio increased from 4.6 to 5.5, respectively. A slight decrease in both MY and COD removal was observed at N/P ratio of 8.5, which was adjusted by decreasing PO4-P concentration from 17.35 to 11.10 mg/L. The results indicated that, as compared to phosphorus, the nitrogen supplement is the key factor for methane production under the conditions examined (Figs. 4a and c), consistent with the previous report by Lei et al. (2010). The residual values of ethanol were relatively high (136.9±9.7 mg/L) as compared to 93.02±6.7 mg/L at N/P ratio of 5.5 (Fig. 4b), indicating insufficient availability of acetate to be converted by methanogens. Moreover, the former study by Demirer et al. (2013) revealed that high N/P ratios have no significant effect on the methane yield. Consistently, in this study, the MY was slightly decreased from 151.86±10.8 to 143.39±10.1 mL/gCODinitial at increasing N/P ratios from 5.5 to 8.5, respectively. The maximum soluble metabolite products (ethanol, acetate and TVFAs) were recorded at N/P ratio of 4.6 (Fig. 4b), which may be ascribed to the insufficient nutrients content to provide high process stability, particularly for methanogenesis (Romero-Güiza et al., 2016). Romero-Güiza et al. (2016) reported that propionate and acetate utilization rates were relatively enhanced under the nutrient-amended conditions. The NH4-N production 15

decreased from 535.25±41.5 to 238.33±17.6 mg/L with increasing N/P ratio from 4.6 to 8.5 (Fig. 4c). This can be attributed to the reason that the release of NH4-N, as well as acetate are produced by acidogenesis (Ali Shah et al., 2014). Higher initial buffering capacity was attained at N/P ratio of 5.5 and 8.5, which is favorable environment for methanogenesis over acidogenesis, resulting in higher acetate consumption pathway for methane production (Fig. 4b and c). Similar trends were observed for the concurrent decrement of NH4-N and H2 productions, as a result of increasing initial NH4-N (Gallert et al., 1998). As well, further alleviation of NH4-N production was recorded at increasing N/P ratio from 5.5 to 8.5. 3.3.Effect of initial pH on H2 production (using pretreated sludge) The results in Fig. 5 show the effects of initial pH and sludge-thermal pretreatment on the hydrogen production at ISR of 3.78 gVSS/gCOD, which achieved maximum hydrogen production (as shown in Fig. 1). Firstly, the sludge thermal pretreatment positively enhanced the hydrogen production (at same pH value of 7.0) from 48±3 mL (untreated) to 55±5 mL (thermal pretreated), and HY was also enhanced by 1.14-fold, as shown in Fig. 5b. Moreover, the contact time to achieve the maximum cumulative H2 production was noticeably reduced from 480 to 168 h, indicating acceleration of acidogenesis activity. The H2 content was substantially increased from 9.80±0.61 to 46.38±4.22% with negligible content of methane, indicating methanogenesis suppression during sludge thermal pretreatment process. Low concentrations of acetate and TVFAs were observed in the untreated sludge samples (Fig. 5c), where methanogenesis converted H2 and acetate into methane gas. Thus, the residual acetate was significantly increased from 31.83±1.79 (untreated sludge) to 1976.85±107.15 mg/L (pretreated sludge), which is consistent with previous report that a majority of methane (e.g., 70%) is produced by acetate reduction 16

process (Ali Shah et al., 2014). As a result, the highest COD removal efficiency of 80.02±5.90% was recorded for the batches with untreated sludge. Similar trends were reported using various substrate compositions (El-Bery et al., 2013; Liu et al., 2014). Secondly, the decrease in initial pH values from 7.0 to 5.0 significantly enhanced the cumulative H2 production by 1.44-fold and the maximum value of 79±6 mL was achieved at pH value of 5.0, as depicted in Fig. 5a. Correspondingly, the maximum HY was increased from 33.11±2.1 to 47.55±4.3 mL/gCODinitial (67.99±5.6 to 97.65±6.1 mL/gMEGinitial) with decreasing initial pH from 7.0 to 5.0. This improvement is attributed to the positive effect of lower pH values (5.0-6.0) on the hydrogenase enzyme activity (ElBery et al., 2013). Nonetheless, decreasing initial pH values less than 5.0 (e.g., 4-4.5) may inhibit the growth of bacteria as well as their hydrogenase activities. The optimum pH value for hydrogen production using heat-treated mixed culture bacteria is dependent on the substrate composition; for example, glucose with optimum pH value of 5.5 (Fang and Liu, 2002), paper mill effluent with pH 5.0 (Farghaly et al., 2015), palm oil mill effluent with pH 6.0 (Yossan et al., 2012)) and rice slurry with pH 4.5 (Fang et al., 2006). The COD removal efficiency also peaked (33.08±2.48 %) at pH value of 5.0, as a result of maximum MEG metabolism into acetate (Eqs. 3 and 4). The generated acetate increased from 1976.85±96.9 to 2581.55±144.6 mg/L with increasing initial pH value from 5.0 to 7.0, respectively. Conversely, the residual EtOH largely dropped from 563.00±40.81 at pH 7.0 to 39.84±2.53 mg/L at pH 5.0, indicating the better conversion of EtOH into H2 and acetate (Eq. 4). This is ascribed to the enhancement of the alcohol dehydrogenase enzyme (ADH) activity at lower pH, which is responsible for the aforementioned reactions (Eqs. 3 and 4) (Catalanotti et al., 2013). The final pH value dropped from 5.69±0.51 to 3.98±0.31 at initial 17

pH variation from 7.0 to 5.0, due to acetate accumulation, as presented in Fig. 7b. These experimental results were fitted to the Gompatz model (Eq. 1), resulting maximum hydrogen production rates (Rm) of 4.44 and 12.44 mL/h at pH values of 7.0 and 5.0, respectively (Table S3, Supplementary). In addition, significant increase in Rm from 0.42 to 4.44 mL/h was achieved with sludge heat pretreatment. 3.4.Three-dimensional regression analysis for cumulative H2 and CH4 production rates Gaussian equation (Eq. 2) was applied in the three-dimensional regression analysis of the experimental results. The combined effect of initial MEG (y) and NH4-N / PO4-P concentrations (x) on the cumulative HP and MP rates was extensively assessed. Accordingly, the following equations were obtained, as a function of the independent variables (x and y), to fit the observed data (Eqs. 7-10). HPMEG-N = 753.44 exp (-0.5 (((x-25.69) / 75.19) 2 + ((y-8657.21)/ 3975.16) 2)),

(7)

HPMEG-P = 459.76 exp (-0.5 (((x- 10.65) / 13.94) 2 + ((y- 7273.72)/ 3672.12) 2)),

(8)

MPMEG-N = 2497.14 exp (-0.5 (((x- 89.38) / 39.89) 2 + ((y- 5948.81)/ 2388.09) 2)), (9) MPMEG-P = 2578.50 exp (-0.5 (((x- 16.11) / 6.76)

2

+ ((y- 5859.04)/ 2249.23) 2))

(10) The predicted values for HP and MP rates were statistically compared to the observed values using ANOVA (analysis of variance), as shown in Table S4 (Supplementary). The results revealed that the four models were statistically fitted to the experimental data, as R2 values ranged from 0.85 to 0.98. However, by analyzing 95% confidence level (pvalue<0.05), only the combined effect of initial MEG and PO4-P on cumulative HP rate was determined to be insignificant (p > 0.05). This indicated that initial MEG and PO4-P 18

had low impact on the related dependent variable (cumulative HP). All the F-values were higher than p-values (low probability), indicating the significance of prediction. The three-dimensional plots supplemented by contour plots (xy-planes) for cumulative HP and MP rates, interacting with two independent variables (MEG-N and MEG-P) are shown in Fig. 6. The figure depicts a clear curvature in each surface and contour plots, except Fig. 6a, indicating that maximum HP and MP rates could be achieved within the experimental region. All the contour plots were also elliptical except for Fig. 6a where the shape of plot was circular and the optimum HP and MP rates can be attained at the center. As shown in Fig. 6a-c, the optimum MEG (as COD)-N concentrations were 8750-51.00 mg/L and 6250-89.28 mg/L for maximum HP and MP rates of 711.74 and 2477.35 mL/Lsubstrate, respectively. The optimal MEG (as COD)-P concentrations were 7500-11.10 mg/L and 5625-15.79 mg/L, resulting in the HP and MP rates of 458.66 and 2561.70 mL/Lsubstrate, respectively. Meanwhile, the maximum predicted HP and MP rates, based on the optimum initial MEG, NH4-N and PO4-P concentrations, were high as compared to the recorded values by 1.53 and 1.10-fold, respectively. 3.5.Energy balance and potential economic benefits The economic and environmental aspects of the two applied AD strategies (i.e., H2/CH4 and H2 production) were assessed as presented in Table 2. The calculations were performed based on an average wastewater discharge of 2000 m3/d containing MEG with average concentration of 2000 mgCOD/L. The equivalent energy production (kWh/year) was calculated assuming 70% efficiency for energy conversion and the aforementioned heating conversion values (i.e., 120 MJ/kg H2 and 50 MJ/kg CH4). The energy costs and revenues were considered as 0.106 $/kWh, while the COD removal add-value was assumed 19

to be 0.149 $/kgCODremoved (Molinos-Senante et al., 2010). The environmental cost of carbon pollution, as CH4 combustion byproduct, was calculated based on 2.74 kgCO2/kgCH4 (Annamalai and Puri, 2006) and about 80$/ton CO2 (Wagner and Weitzman, 2015). The cost of sludge heat-pretreatment (used for single H2 production), prior to start operation, was neglected as it can be considered as a part of the digester capital costs. Gas purification cost of 28.5 $/year for 1 m3 produced CH4/H2 per day, was considered based on a case study of full-scale digester, reported by Ofori-Boateng and Kwofie (2009). As shown in Table 2, the first strategy where maximum CH4 production along with corresponding H2 production can be achieved, has higher net profit as compared to the second strategy for maximum single H2 production, using treated sludge. Notwithstanding, the latter has the highest H2-based bio-energy (i.e., 2.26-fold). Therefore, alternative strategy can be proposed seeking the best combination of the potential energy recovery that was obtained in this study. Two-phase AD system (in series) consisting of single-phase H2 production followed by single-phase CH4 production, is recommended. As a result, the net profit achieved by combining the maximum profits obtained by the other two strategies, resulted in the higher net profit (139,996 $/year), as depicted in Table 2. 4. Conclusions Hythane production from MEG contaminated wastewater via thermophilic dark fermentation is a promising approach. However, the competition between acidogenesis and methanogenesis for substrate utilization is mainly dependent on ISR, N/P ratio, sludge pretreatment and pH. The maximum MY and HY of 151.86±10.8 and 22.27±1.1 mL/gCODinitial were observed at ISRs of 5.29 and 3.78 gVSS/gCOD, respectively. The optimum N/P for methane and hydrogen production were 5.5 and 4.6, respectively. The 20

MY substantially improved by increasing initial ammonia from 51 to 95 mg/L. The hydrogen production improved 1.64-fold using heat-pretreated sludge at initial pH of 5.0, achieving optimum HY of 47.55±4.32 mL/gCODinitial. Acknowledgements The first author would like to acknowledge Ministry of Higher Education (MoHE) of Egypt for providing a scholarship to conduct this study as well as the Egypt-Japan University of Science and Technology (E-JUST) and Tokyo Institute of Technology for offering all facilities and tools needed to conduct this investigation. References 1. Ali Shah, F., Qaisar, M., Shah, M.M., Pervez, A., Ahmed Asad, S., 2014. Microbial Ecology of Anaerobic Digesters: The Key Players of Anaerobiosis. Sci. World J. 1–21. 2. Annamalai, K., Puri, I.K., 2006. Combustion science and engineering. CRC press. 3. APHA, (American Public Health Association), 2005. Standard Methods for the Examination of Water and Waste Water, 25th ed. ed. Washington DC, USA. 4. Britz, T., Noeth, C., Lategan, P., 1988. Nitrogen and phosphate requirements for the anaerobic digestion of a petrochemical effluent. Water Res. 22, 163–169. doi:10.1016/0043-1354(88)90074-7 5. Catalanotti, C., Yang, W., Posewitz, M.C., Grossman, A.R., 2013. Fermentation metabolism

and

its

evolution

in

algae.

Front

Plant

Sci

4,

150.

doi:10.3389/fpls.2013.00150 6. Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: A review. Bioresour. Technol. 99, 4044–4064. doi:10.1016/j.biortech.2007.01.057 7. Darlington, C., Kennedy, K.J., 1998. Biodegradation of aircraft deicing fluid in an 21

upflow anaerobic sludge blanket (UASB) reactor. J. Environ. Sci. Heal. Part A 33, 339– 351. doi:10.1080/10934529809376735 8. Demirel, B., Scherer, P., 2008. The roles of acetotrophic and hydrogenotrophic methanogens during anaerobic conversion of biomass to methane: A review. Rev. Environ. Sci. Biotechnol. 7, 173–190. doi:10.1007/s11157-008-9131-1 9. Demirer, S.U., Taskin, B., Demirer, G.N., Duran, M., 2013. The effect of managing nutrients in the performance of anaerobic digesters of municipal wastewater treatment plants. Appl. Microbiol. Biotechnol. 97, 7899–7907. doi:10.1007/s00253-012-4499-9 10. Dwyer, D.F., Tiedje, J.M., 1983. Degradation of Ethylene Glycol and Polyethylene Glycols by Methanogenic Consortia. Appl. Environ. Microbiol. 46, 185–190. 11. El-Bery, H., Tawfik, A., Kumari, S., Bux, F., 2013. Effect of thermal pre-treatment on inoculum sludge to enhance bio-hydrogen production from alkali hydrolysed rice straw in a mesophilic anaerobic baffled reactor. Environ. Technol. 34, 1965–72. doi:10.1080/09593330.2013.824013 12. Elbeshbishy, E., Hafez, H., Nakhla, G., 2010. Enhancement of biohydrogen producing using

ultrasonication.

Int.

J.

Hydrogen

Energy

35,

6184–6193.

doi:10.1016/j.ijhydene.2010.03.119 13. Elbeshbishy, E., Nakhla, G., 2012. Batch anaerobic co-digestion of proteins and carbohydrates. Bioresour. Technol. 116, 170–8. doi:10.1016/j.biortech.2012.04.052 14. Elreedy, A., Tawfik, A., 2015. Effect of Hydraulic Retention Time on hydrogen production from the dark fermentation of petrochemical effluents contaminated with Ethylene Glycol. Energy Procedia 74, 1071–1078. doi:10.1016/j.egypro.2015.07.746 15. Elreedy, A., Tawfik, A., Enitan, A., Kumari, S., Bux, F., 2016. Pathways of 3-biofules 22

(hydrogen, ethanol and methane) production from petrochemical industry wastewater via anaerobic packed bed baffled reactor inoculated with mixed culture bacteria. Energy Convers. Manag. 122, 119–130. doi:10.1016/j.enconman.2016.05.067 16. Elreedy, A., Tawfik, A., Kubota, K., Shimada, Y., Harada, H., 2015. Hythane (H2 + CH4) production from petrochemical wastewater containing mono-ethylene glycol via stepped anaerobic baffled reactor. Int. Biodeterior. Biodegradation 105, 252–261. doi:10.1016/j.ibiod.2015.09.015 17. Fang, H.H.P., Li, C., Zhang, T., 2006. Acidophilic biohydrogen production from rice slurry. Int. J. Hydrogen Energy 31, 683–692. doi:10.1016/j.ijhydene.2005.07.005 18. Fang, H.H.P., Liu, H., 2002. Effect of pH on hydrogen production from glucose by a mixed culture. Bioresour. Technol. 82, 87–93. doi:10.1016/S0960-8524(01)00110-9 19. Farghaly, A., Tawfik, A., Danial, A., 2015. Inoculation of paperboard sludge versus mixed culture bacteria for hydrogen production from paperboard mill wastewater. Environ. Sci. Pollut. Res. doi:10.1007/s11356-015-5652-7 20. Gallert, C., Bauer, S., Winter, J., 1998. Effect of ammonia on the anaerobic degradation of protein by a mesophilic and thermophilic biowaste population. Appl. Microbiol. Biotechnol. 50, 495–501. doi:10.1007/s002530051326 21. Hassani, A.H., Borghei, S.M., Samadyar, H., Ghanbari, B., 2013. Utilization of moving bed biofilm reactor for industrial wastewater treatment containing ethylene glycol: kinetic

and

performance

study.

Environ.

Technol.

35,

499–507.

doi:10.1080/09593330.2013.834947 22. Kim, K.N., Hoffmann, M.R., 2008. Heterogeneous photocatalytic degradation of ethylene glycol and propylene glycol. Korean J. Chem. Eng. 25, 89–94. 23

doi:10.1007/s11814-008-0015-4 23. Lay, C.H., Chang, F.Y., Chu, C.Y., Chen, C.C., Chi, Y.C., Hsieh, T.T., Huang, H.H., Lin, Y., 2011. Enhancement of anaerobic blohydrogen/methane production from cellulose using heat-treated activated sludge. Water Sci. Technol. 63, 1849–1854. doi:1021661wsl2011.390 24. Lei, Z., Chen, J., Zhang, Z., Sugiura, N., 2010. Methane production from rice straw with acclimated anaerobic sludge: Effect of phosphate supplementation. Bioresour. Technol. 101, 4343–4348. doi:10.1016/j.biortech.2010.01.083 25. Leite, W.R.M., Gottardo, M., Pavan, P., Belli Filho, P., Bolzonella, D., 2016. Performance and energy aspects of single and two phase thermophilic anaerobic digestion

of

waste

activated

sludge.

Renew.

Energy

86,

1324–1331.

doi:10.1016/j.renene.2015.09.069 26. Leskovac, V., Trivić, S., Pericin, D., 2002. The three zinc-containing alcohol dehydrogenases from baker’s yeast, Saccharomyces cerevisiae. FEMS Yeast Res. 2, 481–494. doi:10.1111/j.1567-1364.2002.tb00116.x 27. Liu, Z., Li, Q., Zhang, C., Wang, L., Han, B., Li, B., Zhang, Y., Chen, H., Xing, X.-H., 2014. Effects of operating parameters on hydrogen production from raw wet steamexploded cornstalk and two-stage fermentation potential for biohythane production. Biochem. Eng. J. 90, 234–238. doi:10.1016/j.bej.2014.06.013 28. Liu, Z., Zhang, C., Lu, Y., Wu, X., Wang, L., Wang, L., Han, B., Xing, X.H., 2013. States and challenges for high-value biohythane production from waste biomass by dark fermentation

technology.

Bioresour.

doi:10.1016/j.biortech.2012.10.027 24

Technol.

135,

292–303.

29. Mao, C., Feng, Y., Wang, X., Ren, G., 2015. Review on research achievements of biogas from anaerobic digestion. Renew. Sustain. Energy Rev. 45, 540–555. doi:10.1016/j.rser.2015.02.032 30. Marin, J., Kennedy, K.J., Eskicioglu, C., 2010. Characterization of an anaerobic baffled reactor treating dilute aircraft de-icing fluid and long term effects of operation on granular

biomass.

Bioresour.

Technol.

101,

2217–23.

doi:10.1016/j.biortech.2009.11.055 31. Mohammadi, T., Akbarabadi, M., 2005. Separation of ethylene glycol solution by vacuum

membrane

distillation

(VMD).

Desalination

181,

35–41.

doi:10.1016/j.desal.2005.01.012 32. Molinos-Senante, M., Hernández-Sancho, F., Sala-Garrido, R., 2010. Economic feasibility study for wastewater treatment: a cost-benefit analysis. Sci. Total Environ. 408, 4396–4402. doi:10.1016/j.scitotenv.2010.07.014 33. Moset, V., Al-zohairi, N., Møller, H.B., 2015. The impact of inoculum source, inoculum to substrate ratio and sample preservation on methane potential from different substrates. Biomass and Bioenergy 83, 474–482. doi:10.1016/j.biombioe.2015.10.018 34. Ofori-Boateng, C., Kwofie, E.M., 2009. Water Scrubbing : A Better Option for Biogas Purification for Effective Storage. World Appl. Sci. 5, 122–125. 35. Orecki, A., Tomaszewska, M., Karakulski, K., Morawski, A., 2006. Separation of ethylene glycol from model wastewater by nanofiltration. Desalination360–358 ,200 . doi:10.1016/j.desal.2006.03.364 36. Raposo, F., Borja, R., Martín, M.A., Martín, A., de la Rubia, M.A., Rincón, B., 2009. Influence of inoculum–substrate ratio on the anaerobic digestion of sunflower oil cake 25

in batch mode: Process stability and kinetic evaluation. Chem. Eng. J. 149, 70–77. doi:10.1016/j.cej.2008.10.001 37. Romero-Güiza, M.S., Vila, J., Mata-Alvarez, J., Chimenos, J.M., Astals, S., 2016. The role of additives on anaerobic digestion: A review. Renew. Sustain. Energy Rev. 58, 1486–1499. doi:10.1016/j.str.2014.12.012 38. Schonberg, J.C., Bhattacharya, S.K., Madura, R.L., Mason, S.H., Conway, R.A., 1997. Evaluation of anaerobic treatment of selected petrochemical wastes. J. Hazard. Mater. 54, 47–63. 39. Si, B., Liu, Z., Zhang, Y., Li, J., Shen, R., Zhu, Z., Xing, X., 2016. Towards biohythane production from biomass: Influence of operational stage on anaerobic fermentation and microbial

community.

Int.

J.

Hydrogen

Energy

41,

4429–4438.

doi:10.1016/j.ijhydene.2015.06.045 40. Siddique, M.N.I., Munaim, M.S.A., Zularisam, A.W., 2015. Effect of food to microbe ratio variation on anaerobic co-digestion of petrochemical wastewater with manure. J. Taiwan Inst. Chem. Eng. 0, 1–7. doi:10.1016/j.jtice.2015.06.038 41. Speece, R.E., 1983. Anaerobic biotechnology for industrial wastewater treatment a description of several installations. Environ. Sci. Technol 17, 416A–427A. doi:10.1139/l96-121 42. Sreela-Or, C., Plangklang, P., Imai, T., Reungsang, A., 2011. Co-digestion of food waste and sludge for hydrogen production by anaerobic mixed cultures: Statistical key factors

optimization.

Int.

J.

Hydrogen

Energy

36,

14227–14237.

doi:10.1016/j.ijhydene.2011.05.145 43. Veltman, S., Schoenberg, T., Switzenbaum, M.S., 1998. Alcohol and acid formation 26

during the anaerobic decomposition of propylene glycol under methanogenic conditions. Biodegradation 9, 113–118. doi:10.1023/A:1008352502493 44. Wagner, G., Weitzman, M.L., 2015. The cost of carbon pollution. Retrived from: http://costofcarbon.org/blog/entry/new-climate-economic-thinking 45. Wahab, M.A., Habouzit, F., Bernet, N., Steyer, J.-P., Jedidi, N., Escudié, R., 2014. Sequential operation of a hybrid anaerobic reactor using a lignocellulosic biomass as biofilm support. Bioresour. Technol. 172, 150–5. doi:10.1016/j.biortech.2014.08.127 46. Willquist, K., Nkemka, V.N., Svensson, H., Pawar, S., Ljunggren, M., Karlsson, H., Murto, M., Hulteberg, C., Van Niel, E.W.J., Liden, G., 2012. Design of a novel biohythane process with high H 2 and CH 4 production rates. Int. J. Hydrogen Energy 37, 17749–17762. doi:10.1016/j.ijhydene.2012.08.092 47. Wimonsong, P., Nitisoravut, R., Llorca, J., 2014. Application of Fe-Zn-Mg-Al-O hydrotalcites supported Au as active nano-catalyst for fermentative hydrogen production. Chem. Eng. J. 253, 148–154. doi:10.1016/j.cej.2014.05.047 48. Xu, S.Y., Karthikeyan, O.P., Selvam, A., Wong, J.W.C., 2012. Effect of inoculum to substrate ratio on the hydrolysis and acidification of food waste in leach bed reactor. Bioresour. Technol. 126, 425–430. doi:10.1016/j.biortech.2011.12.059 49. Yossan, S., O-Thong, S., Prasertsan, P., 2012. Effect of initial pH, nutrients and temperature on hydrogen production from palm oil mill effluent using thermotolerant consortia and corresponding microbial communities. Int. J. Hydrogen Energy 37, 13806–13814. doi:10.1016/j.ijhydene.2012.03.151 50. Zerva, C., Peschos, Z., Poulopoulos, S.G., Philippopoulos, C.J., 2003. Treatment of industrial oily wastewaters by wet oxidation. J. Hazard. Mater. 97, 257–265. 27

doi:10.1016/S0304-3894(02)00265-0

28

Cumulative biogas production, mL

600

2.65 gVSS/gCOD 3.78 gVSS/gCOD 5.29 gVSS/gCOD 8.82 gVSS/gCOD 13.23 gVSS/gCOD Inoculum

500 400 300 200 100 0

(a)

0

3

6

9

12

15

18

21

24

27

30

33

36

0

3

6

9

12

15

18

21

24

27

30

33

36

0

3

6

9

12

15 18 21 Time, days

24

27

30

Cumulative HP, mL

60 50 40 30 20 10 0

(b)

Cumulative MP, mL

350 300 250 200 150 100 50 0

(c)

33

36

Fig. 1. Effect of inoculum-to-substrate ratio (ISR) on the cumulative (a) total biogas, (b) hydrogen production (HP), and (c) methane production (MP) from MEG-contaminated wastewater. 29

225

30 150

20 10

75

0 50

0 250

COD removal % Hydrogen yield (HY) Methane yield (MY)

40

Soluble metabolites, mg/L NH4-Nproduced, mg/L

(c)

150

20

100

10

50

Total volatile fatty acids (TVFAs) Ethanol (EtOH) Butyrate

1500

Acetate (HAc) Propionate Final pH

0 11.0 9.0

1000

7.0

500

5.0

0 600

3.0 35

500

PO4-P initial COD/N ratio NH4-N produced

30 25 20

400 300

15

200

10 5

100 0

0 2.65

(d)

200

30

0 2000

(b)

300

Methane yield (MY)

40

MY, mL/gCODremoved & COD removal%

50

375

Final pH

HY, mL/gCODremoved

(a)

MY, mL/gMEG initial HY, mL/gMEG initial MY, mL/gCOD initial HY, mL/gCOD initial

COD/N ratio & PO4-Pinitial , mg/L

Hydrogen yield (HY)

60

3.78

5.29 8.82 ISR, gVSS/gCOD

13.23

Fig. 2. Effect of inoculum-to-substrate ratio (ISR) on: (a) hydrogen and methane yields (HY and MY), based on mL/gCODinitial and mL/gMEGinitial, (b) HY and MY, based on

30

mL/gCODremoved and COD removal efficiency, (c) soluble metabolites and final pH, and (d) ammonification process, based on net production of NH4-N.

Cumulative biogas production, mL

600

N/P = 8.5 N/P = 5.5 N/P = 4.6 Inoculum

500 400 300 200 100 0

(a)

0

3

6

9

12

15

18

21

24

27

30

33

36

0

3

6

9

12

15

18

21

24

27

30

33

36

0

3

6

9

12

15 18 21 Time, days

24

27

30

33

36

Cumulative HP, mL

80 70 60 50 40 30 20 10 0

(b) Cumulative MP, mL

350 300 250 200 150 100 50 0

(c)

31

Fig. 3. Profile of the cumulative: (a) total biogas, (b) hydrogen production (HP), and (c) methane production (MP) at different nitrogen-to-phosphorous ratios (N/P).

180 120 60

Soluble metabolites, mg/L

250

NH4-Nproduced , mg/L

Acetate (HAc) Propionate Final pH

200 150

30 11.0 9.0

5.0

50

750

90 80 70 60 50 40

7.0

100

0 900

(b)

Total volatile fatty acids (TVFAs) Ethanol (EtOH) Butyrate

100

3.0 150

NH4-N initial PO4-P initial NH4-N produced

120

600

90

450 60

300

30

150 0

0 4.6

(c)

COD removal %

240

0 300

(a)

HY, mL/gCOD initial HY, mL/gMEG initial

Final pH

300

MY, mL/gCOD initial MY, mL/gMEG initial COD removal %

5.5 N/P ratio

NH4-Ninitial & PO4-Pinitial , mg/L

Methane & hydrogen yields (MY & HY)

360

8.5

Fig. 4. Effect of nitrogen-to-phosphorous ratio (N/P) on: (a) hydrogen and methane yields (HY and MY), based on mL/gCODinitial and mL/gMEGinitial, and COD removal efficiency, (b) soluble metabolites and final pH, and (c) ammonification process, based on net production of NH4-N.

32

pH 7.0, without pretreatment pH 7.0 pH 6.0 pH 5.5 pH 5.0

Cumulative HP, mL

120 100 80 60 40 20 0 0

40 80 120 160 200 240 280 320 360 400 440 480 520 Time, hrs

100

Soluble metabolites, mg/L

85

80

70

60

55

40

40

20

25

0 3000

(b)

100

COD removal % HY, mL/gCOD initial HY, mL/gMEG initial

2500

10 11.0

Total volatile fatty acids (TVFAs) Acetate (HAc) Ethanol (EtOH) Final pH

10.0 9.0

2000

8.0

1500

7.0 6.0

1000

5.0 500

4.0

0

3.0 pH 7, without pretreatment

(c)

COD removal%

Hydrogen yield (HY)

120

Final pH

(a)

pH 7

pH 6

pH 5.5

pH 5

Initial pH

Fig. 5 Effect of initial pH and sludge thermal-pretreatment on: (a) cumulative hydrogen production (HP), (b) hydrogen yield (HY), based on mL/gCODinitial and mL/gMEGinitial, and COD removal efficiency, and (c) soluble metabolites and final pH.

33

(c)

(a) 0 200 400 600 800

P, mL/Lsubstrate Cumulative M

600

2500 2000

8000 6000

iti

2000 MEG , mgC OD/L

0

(b)

g/L

,m

00 40

G, mg

00 20

COD /L

0

0

Initia l

60

00

00 40

00 20

MEG , mgC OD/L

0

1500 1000 500 0 0 00 10

00 80

Initia lM

60

00

00 40

EG, m gC

2 2 4 223 21 20 19 18 17 16 15 14 13 12 11

PO 4 -P ,m g/L

0 80

O 4 -P

0 0 00

2 234 22 21 20 19 18 17 16 15 14 13 12 11

,m g/L

100

2000

20

OD/L

00

0

iti al

200

2500

In

, mL/Lsubstrate

300

0 500 1000 1500 2000 2500 3000

3000

Cumulative MP

400

In iti al P

, mL/Lsubstrate Cumulative HP

00 60

(d) 0 100 200 300 400 500

500

10

00 80

Initia l ME

al

4000

In

Initia l

0 0 00 10

4 -N

10000

140 130 120 110 100 90 80 70 60 50

500

NH

g/

L

0

1000

al

140 130 120 110 100 90 80 70 60 50

iti

200

1500

In

400

NH 4 -N ,m

, mL/Lsubstrate Cumulative HP

800

0 500 1000 1500 2000 2500 3000

3000

Fig. 6. Three-dimensional and contour plots of the effects of initial NH4-N and PO4-P along with the initial MEG concentrations on the HP rate (a and b) and MP rate (c and d).

34

Table 1 Simultaneous H2 and CH4 production rates from different substrate composition Initial pH

CODinitial

ISR

H2/CH4 yields

(°C)

(g/L)

(gVSS/gCOD)

(L/gCOD)

Acetate

37

7.0

4.20

0.48

NA / 0.126

(Elbeshbishy et al., 2010)

Glucose

37

6.5

10.70

0.19

0.075 / NA

(Elbeshbishy et al., 2010)

Starch

37

7.2

5.00

0.49

NA / 0.316

(Elbeshbishy and Nakhla, 2012)

Cellulose

55

7.0

10.00 b

3.0

0.188 / 0.506

(Lay et al., 2011)

Protein

37

7.2

5.00

0.49

NA / 0.246

(Elbeshbishy and Nakhla, 2012)

Substrate type

Temp.

a

d

References

PWW + AM

55

NA

32.00

0.50

NA / 0.017

Propylene glycol

35

NA

2.00

NA

NA / 1.910

(Siddique et al., 2015) (Veltman et al., 1998)

Acrylic

35

7.5

30.33

NA

(0.041)

(Schonberg et al., 1997)

Ethylene oxide

35

7.5

30.33

NA

(0.062) c

(Schonberg et al., 1997)

MEG

55

7.0

5.00

5.29

0.064 / 0.464

This study

NA, not available; a based on gVS/L; b based on gCellulose/L; c values between brackets refer to the total gas production rate; d based on L/gVS; PWW, petrochemical wastewater; AM, activated manure

35

Table 2 Economic analysis of hythane production from MEG contaminated wastewater Energy recovery Design

COD removal add-value ($/year)

CH4 production (ton/year)

H2 production (ton/year)

Bio-energy production

CO2-released environmental cost

Energy for thermophilic condition (55°C)

MWh/year

$/year

ton/year

$/year

MWh/year

$/year

Gas purification cost ($/year)

Net profit ($/year)

Simultaneous CH4 and H2 production a

158,347

145.6

2.76

1479.95

156,875

278.88

-22,310 c

1309.3

-138,786

-19,719

134,407

Single H2 production b

71,962

-

6.24

145.60

15,434

-

-

1309.3

-138,786

-5,412

-56,802

Two-stage system

158,347

145.6

6.24

1561.15

165,482

278.88

-22,310

1309.3

-138,786

-22,737

139,996

a b

at maximum CH4 production conditions: ISR = 5.29 gVSS/gCOD, pH = 7.0 and N/P = 5.5, considering its corresponding H2 production; at maximum H2 production conditions: ISR = 3.78 gVSS/gCOD, pH = 5.0, N/P = 4.6 and pretreated sludge; c Negative values represent costs; Positive values for profit

36

37

Highlights • Hythane production from anaerobic digestion of MEG-based substrate was examined •

Optimum H2 and CH4 production depended on inoculum-to-substrate ratio (ISR) and N/P

•

CH4 yield peaked (152 mL/gCODinitial) at ISR of 5.29 gVSS/gCOD and N/P ratio of 5.5

•

H2 yield at optimum ISR and N/P, further peaked at pH 5.0 using pre-treated sludge.

•

Regarding energy recovery, substantial economic/environmental profits were achieved

38