Wastewater Treatment and Biohydrogen Production in Attached and Suspended-sludge UASB Reactors During the Start-up Period by Malasse

XIAO Chang-hong，LIU Ming-yi，ZHANG Ling-xing，ZHOU Yu，LI Yong-feng

(School of Forestry, Northeast Forestry University, Harbin 150040, China)

Abstract: Conventional up-flow anaerobic sludge blanket (UASB) reactor and modified UASB reactor were studied and compared during the start-up period in the removal of chemical oxygen demand (COD) and fermentative biohydrogen production. Maximum COD removal rate of 82 and 68.4% and maximum hydrogen production rate (HPR) of 188.99 and 117.03 mmole-H2 L-1h-1 for modified UASB and conventional UASB reactors were observed, respectively. Besides, the maximum specific hydrogen production rate (SHPR) of modified UASB was 45.03 mmol H2/g VSS-d at HRT of 6 h while the maximum SHPR of UASB was 30.36 mmol H2/g VSS-d at hydraulic retention time (HRT) of 8 h. Ethanol-type fermentation was predominant in modified UASB while acetic acid-type fermentation was predominant in conventional UASB. Modified UASB showed stable biohydrogen production and managed to with stand the washout by retaining a high amount of biomass.

Key words:Hybrid; Biohydrogen; Fermentation; Washout; and Biomass

1 Introduction

Fossil fuel has become a major source of pollution and has forced the search for a new source of energy which is environmentally friendly(Sharma, Parnas and Li, 2011). Hydrogen gas has been considered as an ideal alternative source of energy which is clean and possesses a high energy yield (122 kJ/g) when combusted and produces only water as a waste product. However, conventional hydrogen production processes that have been used including electrolysis of water and thermal processes are not environmentally friendly and energy input required during the processes is high(Van Ginkel and Logan, 2005; Wang and Wan, 2009). Anaerobic biological hydrogen production by dark fermentation process is reported to be efficient, simple to control, and has low production cost(Levin, Pitt and Love, 2004). Moreover, the fermentative process could treat wastewater and produce volatile liquid organic acids and alcohols(Manish and Banerjee, 2008).

The anaerobic process is divided into two main processes which are acidogenic and methanogenic. During the acidogenic process, volatile fatty acids (VFAs) and hydrogen gas are produced while the methanogenic process produces methane gas. Pretreatment methods (Heat-shock, Acid or base) and pH control during operation time have been used to inhibit and deactivate methane producers which tend to consume H2intermediate products as substrate(Yin, Hu and Wang, 2014).

An up-flow anaerobic sludge blanket (UASB) reactor is widely utilized in anaerobic stabilization of wastes in water and biogas generation because it has a simple structure, could withstand high organic loading rate, and has high treatment efficiency. However, the common problem of the UASB system is a washout of microorganisms during shorter HRT. To overcome this problem, UASB suspended-sludge system(Chang, 2003; Gavala, Skiadas and Ahring, 2006) has been modified to attached-sludge on carrier materials(Fang, Liu and Zhang, 2002; Kumar and Das, 2001). Comparison studies on the effect of suspended-sludge UASB system and attached-sludge UASB system on hydrogen production are very rare. Gavala et al. (Gavala, Skiadas and Ahring, 2006) in their research at shorter residence time they observed the hydrogen production rate in the attached-sludge UASB reactor was altogether higher contrasted with that of the suspended-sludge CSTR. Attached microbial growth on carrier materials has the following advantages over suspended-sludge systems. It maintains a sufficient amount of hydrogen microbes in the biohydrogen production system while operating at a high organic loading rate (OLR) and low HRT (Barros et al., 2010). Changing the water flow state in the area, thereby increasing the contact of the organic matter in wastewater and microorganisms adsorbed on the carrier, which enhances fully contact and degradation of wastes. The surface of the carriers can adsorb a large number of microorganisms and gradually develop into an anaerobic biofilm, which significantly increases the biomass of the area. It is conducive to the formation of granular sludge in the UASB reactor. Under the agitation of gas and water, the extensive biofilm is detached from the surface of the filler and returned to the sludge bed, which can be rapidly developed into granular sludge. It has been reported that the solid surface with the polar (hydrophilic) surface could help to form better biofilm compared to the hydrophobic solid surface(Sfaelou, Karapanagioti and Vakros, 2015).

The present study examines and compares the efficacy of immobilized microbial growth system on hydrophilic fibrous fiber carrier materials in modified UASB and suspended-sludge system UASB during the start-up period. The study examined the removal of COD, HPR, HY, SHPR, and production of volatile fatty acids and alcohol by using glucose as a sole source of carbon and sewage sludge as the inoculum seed sludge for both bioreactors. During operation time, different HRTs (18-5 h) and different organic loading rates (OLR) range from 18-5 kg·m-3·d-1were used. This research relied upon to give some particular facts toward a more advanced, powerful UASB bioreactor system for the destabilization of wastes and biohydrogen generation at shorter HRT

2 Material and method

2.1 Experimental devices

In this study, two UASB reactors were used. Conventional UASB reactor (UASB), which was a control reactor and modified UASB reactor (HUASB). The configuration of two reactors was all the same except that the hydrophilic fibrous fibers were embedded in the sludge suspended zone in the HUASB reactor (Fig.2). Each reactor had an effective volume of 25 L. The temperature controllers were used to keep the temperature of both bioreactors at 35(±2)℃.

2.2 Inoculum, substrate and operation conditions

The inoculum was taken from the Harbin Wenchang wastewater treatment plant. During cultivation, to inhibit methanogenic bacteria the inoculum was heat-treated at 100℃ for 15 minutes. The seed sludge was subjected to 20 L continuous tank reactor (CSTR) operated in batch mode and after three weeks hydrogen gas was produced(Yang et al., 2006). Subsequently, the Inoculum which contained 33.28 g-VSS L-1of volatile suspended solids (VSS) was transferred into each UASB reactor. Soluble synthetic wastewater with COD range of 3 000-2 000 mgL-1was used to feed both reactors continuously with the following components (gL-1): Glucose 12, KH2PO40.125, NH4Cl 5.2, MgCl.6H2O 0.015, FeCl3.7H2O 0.03, Peptone 0.02, MgCl2.6H2O 0.02 and Cu (NO3)2.3H2O 0.005. NaHCO35 g/L was added to control the pH range 6 to 7.

The two reactors were operated for 46 days, and Table 1 shows the changes in HRT and OLR during operation time.

Tab.1 Operation time and change of HRT and OLR

Fig.1 Schematic diagram for (a) modified UASB and (b) conventional UASB

Analyses

In this study, the COD and VSS were analyzed according to standard methods (APHA/AWWA/WEF, 2012). The concentrations of volatile fatty acids (VFAs), alcohols were determined by gas chromatography (GC-122, Shanghai Analytical Apparatus Corporation, China) which was equipped with FID and a 2.0 m steel column filled with GDX-103(60/80 mesh). Operational temperatures for the injection port, the oven, and the detector were 190, 220 and 220, respectively. Nitrogen, hydrogen, and Argon were used as the carrier gases at a flow rate of 60, 50 and 490 mL·min-1, respectively.

Hydrogen was determined with a gas chromatograph (sc-2, shanghai Analytical Apparatus Corporation, China) which was equipped with a thermal conductivity detector and a 2.0 m steel column(2 m×5 mm) filled with Porapak Q (50-80 meshes). Nitrogen was used as the carrier gas at a flow rate of 40 mL·min-1. The pH and ORP were measured by using pH meter (pHS-3C, Nanjing T-Bota SciTech Instruments Co., Ltd).

Results and discussion

Comparison of COD removal rate

The COD removal rate of the two reactors is illustrated in Fig. 2 and for the first 5 days of the startup the average removal rate for UASB was about 10%, and the effluent COD concentrations are unchanged. The reason was sludge has just been inoculated into a substrate medium and it takes time to adapt to the new environment. On the other side, the COD removal rate in 5 days of operation for HUASB was slightly higher than that of UASB, which was about 15%. It is because fibrous fibers adsorbed some of the organic materials. Ren et al. (Ren et al., 2010) reported similar results when activated carbon adsorbed most of the organic material to 66% of the influent COD on the first day of the experiment in a continuous stirred tank reactor (CSTR). On the Day 6 to Day 10, the hydraulic retention time of the reactor was shortened by 2 h, and the removal rate of HUASB began to increase rapidly, the effluent COD concentration decreased, and the highest COD removal rate reached to 50% on the Day 10, because the microorganisms already adapted to the new environment and their number increased rapidly. However, in the UASB, the COD concentration in the effluent from Day 6 to Day 10 was also decreasing, but the reduction was still very inadequate, and the maximum COD removal rate was less than 20%. It was caused by the absence of fibrous fillers which would accelerate the growth of microbial flora(Karadag et al., 2015) by providing a suitable habitat.

The operation of the two reactors proceeded from Day 11 to Day 30 by the change of HRT as shown in Table 1. Figs 2(a) and 2(b) shows that after each increase of OLR, the COD removal rate was reduced, however, after a buffer period of almost two days, the COD removal rate was restored. The decrease in the COD removal rate was probably due to the organic shock load received during the abrupt change of the OLR(Musa et al., 2018). At this stage, the effluent quality of the HUASB was still better than that of the conventional UASB, and the COD removal rate rises rapidly, mainly because the embedded fibrous filler adsorbs a large number of active bacteria at this stage, and gradually developed into a biofilm. From the Day 33 to the Day 40, sufficient organic substrate promoted the large-scale reproduction of the acid-producing and hydrogen-producing bacteria in the HUASB, and the COD removal rate was stable around 80%. At this time, OLR was 9 kg·m-3·d-1corresponding to HRT of 6 h. According to the effluent water quality analysis results, the HUASB was successfully started. However, the COD removal rate of UASB was still rising, and the maximum COD removal value was 68.46% at HRT of 6 h. The OLR was increased to 10.8 kg·m-3·d-1on the Day 41, which corresponds to HRT of 5 h. The rising flow rate caused sludge to flow out of the two reactors (termed as a washout). The amount of sludge lost by UASB was much more significant than that of HUASB since the later possesses fibrous fiber at the suspended sludge zone, which prevents the excess loss of sludge. The washout caused the COD removal rate of the two reactors being reduced from 82.9% to 74%, and 68% to 60% for HUASB and UASB, respectively.

Fig.2 COD removal rate of (a) conventional UASB reactor (b) modified UASB reactor

Comparison of the effect of HRT on H2 production

The gas produced only consists of H2and CO2and the proportion of hydrogen gas is illustrated in Fig.3(d). The hydrogen gas content in both bioreactors was significant and HUASB H2content was higher than that detected in UASB. The reason is obvious, the presence of fibrous fibers in HUASB caused the higher growth rate of hydrogen-producing microorganisms and the formation of biofilm on these fibers in the suspended zone of bioreactor means microorganisms were retained in high amount even after washout at shortest HRT of 5 h. Hydrogen production rate (HPR) and hydrogen yield (HY) as illustrated in Fig.3(a) and 3(b) was observed at a maximum when HRT of 6 h and 8 h for UASB and HUASB, respectively. Maximum HPR of 188.99 and 119.55 mmole-H2L-1h-1, maximum HY of 6 872 and 4 347 mmol H2/mole glucose of HUASB and UASB, respectively, were observed. At shorter HRT of 5 h caused the HPR and HY to decrease from both reactors due to washout of hydrogen-producing organisms, HPR of 169 and 86.68 mmole-H2L-1h-1for HUASB and UASB, respectively, was observed. Shorter HRT increases the OLR and degradation of the substrate but can result in negative effects(Zhang et al., 2007). Substrate degradation is illustrated in Fig 3(c). HUASB showed a high ability to degrade substrate and produced high-quality effluent for discharge with the removal of 95% of the substrate when a steady-state was reached. Anaerobic treatment by using UASB produced effluent which needs post-treatment and 74% was the maximum removal of the substrate. The substrate in HUASB promised to be used in Hydrogen production and not directed to bacterial growth and repair(Zhang et al., 2007).

Fig.3 Variation in (a) Hydrogen production rate (b) Hydrogen yield (c) substrate degradationefficiency (%) and (d) H2 content in biogas of UASB reactor and HUASB reactor

Comparison of biomass activity on biohydrogen production

Fig.4 illustrates the gradual increase of volatile suspended solids (VSS) and SHPR of two reactors from the first day of the study. The SHPR was observed to increase with VSS gradually in both reactors until Day 24 when SHPR in HUASB and UASB reached 40.36 and 27.99 mmol H2/g VSS-d, which corresponds to VSS of 170.60 and 165.51g, respectively, at HRT of 10 h. The SHPR was consistent with VSS formation from day 1 to day 24 in both reactors. It determines that in the biomass (VSS), the hydrogen-producing organisms were dominant. Another speculation is that the metabolic pathway which favored hydrogen production occurred(Chang, 2003). UASB reached higher and maximum SHPR of 30.36 mmol H2/g VSS-d on Day 29 at which 175 g VSS of biomass was produced and HRT was 8 h. SHPR on Day 31 was reduced to 26.82 mmol H2/g VSS-g for HUASB and 38.74 mmol H2/g VSS-g for UASB, respectively, due to organic shock load caused by shorter HRT of 7 h. However, no significant change of VSS was observed in the bed zone for both reactors (Fig.4). It shows that the organic shock load lowered SHPR due to changes in the fermentation metabolic pathway. At 6 HRT, both reactors managed to recover, and SHPR was increased to 29.72 and 45.03 mmol H2/g VSS-d for UASB and HUASB, respectively. When HRT was shortened to 5 h on Day 42, washout of biomass was observed from both reactors, and VSS was lowered to 159.30 g and 176.79 g for UASB and HUASB, respectively. Since HUASB possesses fibrous fibers, VSS lowered in HUASB was 3.1 times less than that lost by conventional UASB. The washout of biomass probably reduced hydrogen-producing microorganisms in both reactors; consequently, SHPR of both reactors was reduced to 24.39 and 40.28 mmol H2/g VSS-d for UASB and HUASB, respectively.

Fig.4 Biomass concentration with SHPR for UASB and HUASB

Comparison of liquid phase end products (VFAs)

Biohydrogen production is always accompanied by the production of liquid products (acetate, butyrate, propionate, ethanol, etc.) which constitute different types of fermentation(Puyol et al., 2017). In the present study, acetate, ethanol, propionate, and butyrate were observed in a considerable amount from the effluent of both reactors. Table 2 shows the effect of HRT on the change of the pH, which directly influenced the type of metabolite pathway produced in the reactors at different times. Fermentation types are incredibly touchy to pH changes, according to biohydrogen production theory(Li et al., 2019). In both reactors, as the HRT was shorted from 18 to 6 h, the pH of the effluent decreased from 6.9 to 4.5 as the VFAs accumulates in the reactor. In UASB, the acetate was the predominant metabolite, which constitutes 44.1% to 34.5% of the effluent metabolites from Day 9 to Day 46. Acetate metabolic pathway has been reported as favorable for biohydrogen production (Ding, Yang and He, 2016). The second metabolite which was predominant in UASB was propionate, which constitutes 31% of effluent metabolite at HRT of 18 h on Day 4 to 18.4% at HRT of 6 h on Day 39. The propionate pathway(Eq. (1)) reported consuming hydrogen and tended to lower specific hydrogen production rate(Saady, 2013)in UASB. Other liquid end products from UASB were in lower percentage, Ethanol 8.3% at 18 h HRT to 22.9% at 6 h HRT, butyrate 24.6% at 18 h HRT to 12.8% at 8 h HRT.

HUASB which contained fibrous fibers in suspended sludge zone was dominated with ethanol type fermentation from Day 14 to Day 46 with little change in pH. Generally, the short HRT favored the ethanol pathway in reactor II, and the percentage of ethanol increased from 37.3% on Day 14 at 14 h HRT to 48.8% on Day 39 at 6 h HRT. The ethanol-type fermentation has been reported to be a superior and progressively stable metabolic pathway for biohydrogen generation, dependent on the redox reactions among NAD+ and NADH inside bacterial cells(Manish and Banerjee, 2008; Yin, Hu and Wang, 2014). The second metabolite, which was predominant in HUASB is acetate. The acetate fermentation was predominant in HUASB on initial days of experiment and on Day 4 reached 41.2% of the metabolites, before lowered to 24.4% on Day 14 at 14 h HRT, when ethanol-type fermentation started to be dominant. Acetate-type fermentation (Eq.(2)) and ethanol-type fermentation(Eq.(3)) which were dominant in HUASB are pathways for efficient biohydrogen production(Zhang et al., 2017).

(Eq.1)

(Eq. 2)

(Eq. 3)

The presence of stringy fibers at the suspended sludge zone in HUASB caused the production of a suitable culture for biohydrogen production. Acetate and ethanol pathway bacteria were formed in biofilm formed onto these fibers, and stable hydrogen production was maintained very quickly in HUASB even after the washout of suspended bacteria at shorter HRT of 5 h on Day 42. The washout caused acetate and ethanol-type fermentation to decrease to 29.9 and 16.6%, respectively.

Tab.2 Distribution of VFA and ethanol in the UASB and HUASB effluent at different HRT

I=UASB, II=HUASB

Oxidation reduction potential

The oxidation-reduction potential (ORP) change (Fig. 5) for both reactors were similar. It can be seen at the initial stage, the anaerobic degree is low in both systems, and the ORP is relatively high. With the anaerobic environment gradually formed, the ORP value in the UASB has gradually stabilized around -300 mV at pH range 4.5-5 and the predominant metabolic pathway is acid-type fermentation. The fermentation type of the HUASB system was changed to ethanol-type fermentation. The ORP value was stable below -425 mV at the pH range of 4.8 to 6.0. The high amount of biomass maintained in the HUASB by fibrous fibers created an anaerobic environment suitable for bacteria growth and stable pathways for hydrogen production were maintained. The pH and ORP jointly controlled fermentation types in both reactors.

Fig.5 Redox potential change for the UASB and HUASB during the start-up period

3 Conclusion

The attached sludge system-HUASB demonstrated its potential in wastewater treatment and biohydrogen gas production. During the operation time, the HRT was reduced from 18 to 5 h before the modified UASB was successfully started. The acetic acid and ethanol-type fermentation were predominant for conventional UASB and modified UASB, respectively. The COD removal rate, HPR, HY, and SHPR with biomass (VSS) activity were also compared. The following results were observed, maximum COD removal rate of 82 and 68%, maximum HPR of 188.99 and 119.55 mmole-H2L-1h-1, maximum HY of 6 872 and 4 347.5 mmol H2/mole glucose, and maximum SHPR of 45.03 and 30.36 mmole H2/g VSS-d of modified UASB and conventional UASB, respectively. Optimal operational parameters of HUASB and conventional UASB were as follows; pH range 4.8 to 6.0 and 4.5-5.0, HRT of 6 h and 8 h, OLR 9 and 4.5 kg m-3d-1, ORP-100 to -300 mV and -100 to -450 mV, respectively.

doi:ISBN 9780875532356.

doi:10.1016/j.ijhydene.2010.01.108.

doi:10.1016/s0360-3199(03)00082-x.

doi:10.1016/B978-0-08-100455-5.00011-4.

doi:10.1002/bit.10174.

doi:10.1016/j.ijhydene.2005.09.009.

doi:10.1016/j.procbio.2014.11.005.

doi:10.1016/S0141-0229(01)00394-5.

doi:10.1016/S0360-3199(03)00094-6.

doi:10.1063/1.5024723.

doi:10.1016/j.ijhydene.2007.07.026.

doi:10.3390/ijerph15102220.

doi:10.3389/fmicb.2016.02106.

doi:10.1016/j.ijhydene.2009.05.010.

doi:10.1016/j.ijhydene.2013.07.122.

doi:10.1155/2015/734384.

doi:10.1016/j.ijhydene.2010.12.040.

doi:10.1016/j.watres.2005.07.021.

doi:10.1016/j.ijhydene.2008.10.010.

doi:10.1016/j.ijhydene.2005.11.018.

doi:10.1016/j.ijhydene.2014.01.145.

doi:10.1016/j.biortech.2016.11.096.

doi:10.1016/j.ijhydene.2006.08.017.