Hydrogen production using an anaerobic baffled reactor

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 2 1 5 4 e1 2 1 6 1

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Hydrogen production using an anaerobic baffled reactor e Mass balances for pathway analysis and gas composition profiles Lars Ju¨rgensen a,c,*, Ehiaze Augustine Ehimen b,1, Jens Born c, Jens Bo Holm-Nielsen a a

Aalborg University Esbjerg, Niels Bohrsvej 8, 6700 Esbjerg, Denmark Flemish Institute of Technological Research (VITO), Boeretang 200, 2400 Mol, Belgium c Flensburg University of Applied Science, Kanzleistr. 91, 24937 Flensburg, Germany b

article info

abstract

Article history:

This study investigated pathways for anaerobic hydrogen production from biomass at low

Received 15 April 2015

pH values, also known as dark fermentation. A 200 l bench scale anaerobic baffled reactor

Received in revised form

with four internal compartments was used for hydrogen production from wheat starch.

30 June 2015

The liquid fermentation products and hydrodynamic characteristics were analyzed using

Accepted 13 July 2015

high performance liquid chromatography, tracer studies, and gas analysis. A mean resi-

Available online 5 August 2015

dence time of 29 h and a feed strength of 4 gCOD/l resulted in a total gas production of 230 l/ d containing 42% hydrogen and 11% methane. The gas collected from the different com-

Keywords:

partments highly differed in composition showing a partial phase separation, with

Anaerobic digestion

maximum H2 concentrations of up to 60% observed in the first compartment. 49% and 44%

Hydrogen

of the total H2 produced were derived during the formation of acetic and butyric acid

Anaerobic baffled reactor (ABR)

respectively. Just 8% of the H2 was produced during propionic acid synthesis. Concentra-

Pathways

tions up to 1 g/l lactic acid built by the bifidum pathway was also observed.

Biogas

Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction _ in Hydrogen is a key component in chemical industry, e.g. crude oil processing or ammonia production, as well as in pharmaceutical and food production processes. Today, more than 5$109 kg hydrogen is produced worldwide, mainly by the steam reforming of methane [1]. Towards a sustainable future and a 100% renewable energy system, new processes for

hydrogen production have to be developed to become compatible [2,3]. Besides the well known production by electrolysis, the biological conversion of biomass seems to offer promising options [4e7]. The microbial conversion of biomass for hydrogen production can be divided into the main process categories: photo- and dark fermentation, thermophyilic and enzymatic digestion. Dark fermentation comes with the advantage of using mixed cultures stabilized against contamination of biological species entering the process by

* Corresponding author. West Coast University of Applied Sciences, Fritz-Thiedemann-Ring 25, 25746 Heide, Germany. Tel.: þ49 481 8555 375. E-mail address: [email protected] (L. Ju¨rgensen). 1 Future Analytics Consulting, 23 Fitzwilliam Square, Dublin 2, Ireland. http://dx.doi.org/10.1016/j.ijhydene.2015.07.068 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.


the feed stream (favorable waste streams from other processes) [8]. Dark fermentation is also known as the acidogenesis step of the conventional anaerobic digestion (AD) [9]. Hydrogen production by anaerobic digestion has been discussed extensively in scientific papers during the last decades. Anaerobic digestion is already known to be an economical process for the treatment of complex waste streams, agricultural by-products, or energy crops for biogas production [10]. Biogas is a mixture of the main components: CO2 and CH4 with minor trace concentrations of NH4, H2, and H2 S and other gases [11]. Today, the main use is for cooking when produced in small scale household digesters (mainly in developing countries) [12], combined heat and power (CHP) generation, and the production of substituted natural gas (SNG) by gas separation [13], or via the Sabatier process using hydrogen from renewable energy systems [14]. In recent years, the possibility of hydrogen production by anaerobic digestion has been increasingly discussed in scientific literature [15]. A wide range of studies have been performed using model substrates, such as sugar [16] or starch [17], up to very complex waste streams like agricultural waste [18,19] for demonstrating the feasibility of bio-H2 production. H2 production by dark fermentation is possible due to the nature of the anaerobic degradation of polymers via four steps: hydrolysis, acidogenisis, acetogenesis, and methanogenesis (see Fig. 1). By lowering the pH value below 6.0e6.5, the methanogenic bacteria can be inhibited, thus H2 production via the hydrolysis and acidogenesis steps can be released to the gas phase [17]. The H2 is produced during the formation of the volatile fatty acids (VFA), for example acetic and butyric acid synthesis from glucose (as presented in equations (1) and (2)) [21]: C6 H12 O6 þ 2H2 O/2CH3 COOH þ 2CO2 þ 4H2

(1)

C6 H12 O6 /CH3 ðCH2 Þ2 COOH þ 2CO2 þ 2H2 :

(2)

Hydrogen plays a key role in anaerobic digestion: Released during the acidogenesis, it is consumed quickly by the acetoclastic/hydrogenotrophic methanogens producing CH4 and CO2 [22]. The presence of methanogens is the main disadvantage with the use of mixed cultures [23]. The molecular hydrogen dissolved in the liquid phase is oxidized by an enzyme called hydrogenases. Its activity decreases with decreasing pH, which is assumed to be the reason for inhibition of the methanogens at low pH [22]. The VFAs formed by the acidogenesis are degraded by the acetogenesis and methanogenesis to CH4. This can be expressed by the general reaction [24]: Cn Ha Ob þ ðn a=4 b=2Þ H2 O/ðn=2 a=8 þ b=aÞ CO2 þ ðn=2 þ a=8 b=aÞ CH4

(3)

Furthermore, the H2 and CO2 process components present in the digestion reactor would also potentially react to form CH4 [25]: 4H2 þ CO2 /CH4 þ H2 O:

(4)

With regards to the optimisation of proposed H2 production using anaerobic digestion systems, the inhibition of this methanogenic step is therefore important to favor an accumulation of H2 in the final gas phase.

12155

In addition to the common fermentation VFAs (i. e. acetic, butyric, and propionic acid), the formation of another process intermediate, lactic acid is also of interest in this study. This is due to the fact that lactic acid formation has been reported to have a negative impact on H2 production by its hydrogen neutral synthetic route [26]. In literature, the production of lactic acid can be explained by three hydrogen neutral pathways, the homofermentative (5), heterofermentative (6), and the bifidum (7) pathways [26]: C6 H12 O6 /2CH3 CHOHCOOH

(5)

C6 H12 O6 /CH3 CHOHCOOH þ CH3 CH2 OH þ CO2

(6)

2C6 H12 O6 /2CH3 CHOHCOOH þ 3CH3 COOH:

(7)

Although the conversion of the digested macromolecules to lactic acid intermediates would be preferentially limited for the proposed H2 production via anaerobic digestion, lacitc acid can also be considered as a valuable chemical output with a lot of interest shown for its role in proposed biorefinery systems [27]. With regards to the reactor set-up used to facilitate the anaerobic digestion process, hydrolysis has often been carried out in continuous stirred tank reactors (CSTR) with low residence times due to its fast kinetics [28]. In recent years, upflow anaerobic sludge blanket reactors (UASBR) [29], packed bed reactors [30], and biotrickling reactors have been used for biohydrogen production [15]. The main difference of these systems compared to the CSTR being the fixation of the active biomass achieved by balancing the solid retention time (SRT) and hydraulic retention time (HRT). In conventional CSTRs both are the same, thus a wash-out of the bacterial consortium from the reactor when applying short (mean) residence times was encountered. By using different compartments, the different phases of AD can be separated, i. e. the acidogenesis takes place in the first compartments, while the acetogenesis and methanogenesis steps occur place in the later compartments and at different pH [31]. Additional advantages with such systems are the simple and robust design without any moving parts [32] and good shock loading resistance [33]. For this study, an anaerobic baffled reactor (ABR) is used for H2 production. The ABR system can simply be seen as a series of UASBRs. Considering such a series, it is possible to utilize the effluent from the H2 producing stage for further methane production [34]. The common model for the ABR is a well fluidized sludge bed in the upflow section, where the main biological activity takes places. In the downflow, the sludge bed is settled and no biological activity is assumed to take place. Some models even neglect the volume of the sludge bed in the downflow section [32]. Thanwised et al. (2012) described the effects of hydraulic retention time on H2 production and chemical oxygen demand removal from tapioca (a type of starch) wastewater using an ABR on lab scale (24 l). In their study, specific H2 yields up to approximately 880 ml/(l$d) were observed. Inspired by that paper, we investigated the H2 production from wheat starch in detail, using a 200 l bench scale ABR. This was carried out to provide more details on potential H2

12156


Fig. 1 e Anaerobic digestion of organic matter to biogas: depending on the substrate the bacteria have to be distinguished into four to five bacteria groups (gray background) (according to [20]).

production and an understanding of the biomass degradation process using such systems which have not been explored by previous studies in the literature. Using residence time distribution analysis (RTD), high performance liquid chromatography (HPLC), and gas analysis for each of the compartment outputs, it was possible to investigate the degradation of the biomass in detail. The HPLC use allowed for closed mass balances and to draw inferences from the VFAs build-up and the potential pathways for H2 production.

Methods and materials Experimental setup A 200 l bench-scale ABR (see Fig. 2) made from stainless steel was used to carry out the experiments. The series of hanging and standing baffles divides the reactor into down- and upflow sections. One down- and one up-flow section is what constitutes one compartment. Using this setup, the flow can be considered as being intermediate between plug-flow and perfectly mixed [35]. During fermentation the upflowdownflow-ratio was 1:1. The ground of the compartments was angled at approximately 30 to amplify the mixing effect. The temperature of each compartment was measured by a Pt100 temperature sensor and controlled by electrical heat foils stuck to the reactor surface. The operating temperature was set to 38 C. By this individual close-loop control of each compartment, a uniform temperature profile along the reactor was avoided. The pH was monitored by sampling from the sample ports as shown in Fig. 2. The initial pH of 7 in all chambers was allowed to drop, so a self-established pH profile along the reactor could be observed (see VFA concentration profile). The produced biogas was sampled from five ports in the reactors head space enabling separate measurements of the gas flows from each compartment. The total gas flow was measured continuously using a Ritter counter. Commercial

wheat starch (Penny GmbH, Germany) was used as model substrate for feed preparation. Due to the highly standardized and monitored food processes, wheat starch is considered an ideal and constant (with regards to its composition) model substrate. The starch was sieved and mixed with water in a well stirred feed tank. The feed concentration was held constant at 3.6 gCOD/l. The feed was pre-heated to 35 C before entering the ABR. A nutrient solution [36] was added to avoid nutritional limitations, containing the most essential nutrients for hydrogen-production, i. e. ammonia and iron. Iron is important, because it is a co-factor for the hydrogenase [37]. Seed sludge collected from a 5000 m3 CSTR, located at the municipal waste water treatment plant in Flensburg, Germany, was used as inoculum. This CSTR was used for treatment of sewage sludge aimed at biogas production. Seed sludge from anaerobic digester contains a diverse micro-flora that are more robust against contamination and environmental stress [37]. After inoculation the reactor was placed in stand by for one week to allow for settling.

Residence time distribution analysis RTD studies were carried out in the empty reactor (i. e. no biomass) using NaCl solution (1 g/l). Two pulses (t ¼ 20 min and t ¼ 200 min) and a step function (t ¼ 200 min) experiment was conducted. The conductivity of the effluent was then measured to determine the RTD. To analyze the derived data, a dispersion model was used, which describes the ABR as a series of plug flow reactors and CSTRs [38].

Start up To facilitate the start-up of the investigated process, the chemical oxygen demand (COD) based organic loading rate (OLR) was increased by decreasing the HRT at constant feed strength (3.6 gCOD/l) progressively from 0.6 to 3.2 g/(l$d) until day 59. To accelerate the start up, a split feed procedure was


used during the first 31 days as in Ref. [39], i. e. a fraction of the feed was fed to every compartment.

Analysis The COD was determined by Hach Lange cuvette tests (LCK014 and LCK114, according to the sample concentration). For the preliminary investigation of the produced gas, a biogas analyzer VISIT 03 (Messtechnik EHEIM GmbH) was used. The gas was collected in bags and fed to the analyzer. A liquid chromatography system (Merck Hitachi) with interface module D-6000A, a L-6200A intelligent pump and the RI-detector L-2490 was used. Furthermore, an ICE-Coregelcolumn 87H3 from the company Transgenomic Inc., Omaha, USA, was applied. The column packing material consisted of cross linked, sulphurized polystyrene with a particle size of 9 mm. 4 mM sulphuric acid (H2 SO4) in deionized water was used as the eluent. The HPLC samples were taken every 10 days. Data processing was carried out using the D-7000 HPLC System Manager (LaChrom) software.

Results and discussion Hydrogen production and gas separation For the entire reactor operation period, the H2 fraction of the total gas yield was produced in high concentration (see Fig. 3) between 30 and 49%. Gas compositions appeared to stabilize after day 72 with approximately 41% H2. The CO2 concentration was seen to be slightly higher than the H2 concentration. The CH4 concentration in the produced biogas decreased from 25% at the reaction start-up to 9% after 72 days, indicating a strong inhibition of the methanogens. The gas compositions can be observed to derivate with time before stabilizing. This can be explained by the adaptation of the micro-flora to the new environmental conditions and that since it takes some time before the activity of the hydrogen consuming bacteria is

12157

minimized. The mechanisms involved here are however not clear [37]. Also of interest in this study is the constituent gas profiles obtained from the different reactor compartment units as previously discussed. Fig. 4 shows the composite gas profiles along the reactor with increasing CH4 concentrations and decreasing H2 concentrations. The gas concentrations from GO 3, 4 and 5 did not differ much, with only minor gas flows (