Bioresource Technology 129 (2013) 85–91

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Hydrogen production from the organic fraction of municipal solid wastein anaerobic thermophilic acidogenesis: Influence of organic loading rateand microbial content of the solid waste

S. Zahedi a,⇑, D. Sales a, L.I. Romero b, R. Solera a

a Department of Environmental Technologies, Faculty of Marine and Environmental Sciences (CASEM), University of Cádiz, Pol. Río San Pedro s/n, 11510 Puerto Real (Cádiz), Spainb Department of Chemical Engineering and Food Technology, Faculty of Sciences, University of Cádiz. Pol. Río San Pedro s/n, 11510 Puerto Real (Cádiz), Spain

h i g h l i g h t s

" We studied the hydrogen production from municipal solid waste." Different operating conditions were assayed in this study." Increasing organic loading rates resulted in an increase in hydrogen production." It is possible to treat solid waste within an optimal range for a liquid substrate." Microbial concentration in the reactor is influenced by the microbial content of the waste.

a r t i c l e i n f o

Article history:Received 23 July 2012Received in revised form 23 October 2012Accepted 1 November 2012Available online 17 November 2012

Keywords:Hydrogen productionMicrobial activitySolid wasteThermophilic condition

0960-8524/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.biortech.2012.11.003

⇑ Corresponding author. Tel.: +34 956 016157.E-mail address: soraya.zahedi@uca.es (S. Zahedi).

a b s t r a c t

Hydrogen production (HP) from the organic fraction of municipal solid waste (OFMSW) under thermo-philic acidogenic conditions was studied. The effect of nine different organic loading rates (OLRs) (from9 to 220 g TVS/l/d) and hydraulic retention times (HRTs) (from 10 d to 0.25 d) was investigated. Normally,butyrate was the main acid product. The biogas produced was methane- and sulfide-free at all tested OLR.Increasing the OLR resulted in an increase in both the quantity and quality of hydrogen production,except at the maximum OLR tested (220 g TVS/l/d). The maximum hydrogen content was 57% (v/v) atan OLR of 110 g TVS/l/d (HRT = 0.5 d). HP was in the range of 0.1–5.7 l H2/l/d. The results have clearlyshown that the increase in OLR was directly correlated with HP and microbial activity. The bacterial con-centration inside the reactor is strongly influenced by the content of microorganisms in the OFMSW.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The generation and management of waste is a serious environ-mental problem for modern societies. However, properly managedwastes become resources that contribute to savings in raw materi-als, conservation of natural resources and climate, and sustainabledevelopment. Hydrogen is currently considered one of the majorenergy carriers of the near future because it is clean, recyclableand efficient and can be used to generate electricity at the sametime as not contributing to greenhouse-gas production. Severalauthors have studied its production by means of biological pro-cesses (Kapdan and Kargi, 2006; Valdez-Vazquez et al., 2005,2006), including anaerobic acidification of MSW (Shin et al.,2004; Shin and Youn, 2005). A basic description of the anaerobicacidogenesis process includes three steps. In the first step

ll rights reserved.

(hydrolysis), complex organic polymers are hydrolyzed into sim-pler soluble organic compounds. Acidogenesis and acetogenesisare the second and third steps, respectively, and produce volatilefatty acids (VFA), H2, CO2 and other intermediates.

Acidogenesis of municipal solid waste (MSW) needs short HRTsand low pH to prevent the growth of hydrogen consumers as meth-anogens. Several authors have suggested that the optimal pH forenhanced hydrolytic and acidogenic bacteria activity ranges be-tween 5 and 6 (Verrier et al., 1987). To date, the lowest HRT for aci-dogenic of MSW is 1.2 d (Ueno et al., 2007), operating undercontinuous mode and thermophilic conditions and inoculatingthe acidogenic reactor with seed microflora. Furthermore, thermo-philic conditions have been reported to have an inhibitory effect onmethanogenesis (Ueno et al., 1996, 2007).

H2 production from MSW is also influenced by the compositionof the waste. Changes in waste composition can be produced byseveral factors, including climate, collection frequency and sea-sonal practices (Tchobanoglous et al., 1997). Some reports have

Fig. 1. Schematic diagram of the CSTR.

86 S. Zahedi et al. / Bioresource Technology 129 (2013) 85–91

obtained continuous hydrogen production treating garbage andindustrial wastewater using natural populations of microorgan-isms without sterilization. These reports suggest that the selectionof suitable microflora and their cultivation under appropriate con-ditions would enable stable hydrogen fermentation if the processwas operated at an industrial level (Ueno et al., 2007).

Understanding the functioning of the acidogenic reactor re-quires quantitative information on the microbial number andactivities of the bacteria involved in the process. Microbial activitywill correlate with number only as long as environmental condi-tions remain constant. Any change in substrate and operating con-ditions in the reactors will alter these parameters. Microbialnumber and activity represent distinct ecological parameters.Though normally correlated to each other, they should not be usedinterchangeably. Direct count procedures yield the highest esti-mates of numbers of microorganisms. Epifluorescence microscopywith fluorometric stains are widely used for direct counting of bac-teria. In particular, 40, 60-diamidine-2-phenylindole (DAPI) is themost widely used staining protocol (Kepner and Pratt, 1994). In aprevious study, determining activity ‘‘in situ’’ without extrapolat-ing from the reactor effluent was shown to have an advantage overthe activity test, as it permits direct measurement of activity(Montero et al., 2009).

Previous studies had demonstrated links between digesteroperating conditions, physical and chemical performance parame-ters and microbial population dynamics (Montero et al., 2008). Theresults had clearly indicated that the relative abundance of micro-organisms was directly correlated with OLR, volatile solid removaland methane production by anaerobic reactor treatment of theOFMSW. However, there have been few studies on the correlationsbetween HP and the number of microorganisms or their activitiesin the reactor and the influence of the microbial content of theOFMSW on the process.

The aim of the present study was to optimize the operating con-ditions, HRT and OLR of the hydrolytic-acidogenic step of theanaerobic process of OFMSW in a thermophilic semi-continuousstirred tank reactor. The effect of the variations in operatingparameters and OFMSW on chemical oxygen demand (COD),VFA, total volatile solids (TVS), microbiological concentration andhydrogen production to improve the aforementioned stage wasstudied at laboratory scale. Relationships between OLR, hydrogengeneration and both total number and activity of microorganismswere also considered.

Table 1Operating conditions applied during the experimental test.

HRT (d) 10 6.6 4.4 3 2 1 0.75 0.5 0.25OLR (g TVS/l/d) 9 13 19 28 43 55 73 110 220

2. Methods

2.1. Experimental equipment

A laboratory-scale continuously stirred tank reactor (CSTR) wasused in these studies (Fig. 1). The equipment employed consists ofa 5.5 l capacity stainless steel reactor, heated by recirculatingwater through a thermostatic jacket. A PRECISTERM 6000142/6000389 (SELECTA S.A.) bath was used, with a maximum capacityof 7 l of water. The stainless steel reactor lid has a diameter of200 mm and contains three openings, one for the biogas outlet, afeed inlet and another opening for the stirring system. The bottomof the reactor has a discharge valve with a 40 mm i.d., used forsampling. The biogas is collected in 40 l capacity Tedlar (a polyvi-nyl fluoride plastic polymer) bags. The bags are 29.8 cm wide and45.7 cm long. The stirring system consists of an IKA EUROSTARPower Control visc-P4 overhead stirrer coupled to a stainless steelblade with scrapers which allows homogenization of the waste at aspeed of 23 rpm. In a CSTR without recycling of solids, the solidretention time (SRT) and HRT are equal.

2.2. Start-up and operating conditions

A semi-continuous reactor and thermophilic conditions werechosen for this study. Anaerobic digester effluent from single-phase dry-thermophilic anaerobic digestion of OFMSW was usedas inoculum for the hydrogen reactor. Total and volatile solid con-centrations in the inoculum were 67 g/kg and 33 g/kg, respectively.pH was maintained stable within the 5.0–5.8 range for 27 d oper-ating at an HRT of 15 d (OLR = 6 g TVS/l/d), which is not suitable formethanogenic microorganisms (Moosbrugger et al., 1993). pH wasadjusted using 10 N NaOH and 8 M HCl solutions.

Different HRTs were tested after 27 days of operation (initialphase), when H2 production was detected and methanogens wereinhibited, i.e. when no methane production was observed. The ini-tial OLR was 9 g TVS/l/d, corresponding to an HRT of 10 d. The OLRsand corresponding HRTs assayed are shown in Table 1.

2.3. Feeding regime

As regards the feeding regime, the reactor was fed once a dayfor an HRT of between 10 d and 2 d; twice a day for an HRT of be-tween 1 d and 0.75 d; three times a day for an HRT of 0.5 d; andevery three hours for an HRT of 0.25 d. Each HRT was maintainedfor at least three periods in order to achieve stable operation.The overall duration of the experiments was 178 days. The changein HRT was applied when the reactor was stable, hydrogen produc-tion being the parameter chosen to demonstrate the stability of thereactor (Fig. 2a and b).

Fig. 2. (a) Evolution of HP (l H2/l/d) from t = 0 to t = 141 d. (b) Evolution of HP (l H2/l/d) from t = 141 to t = 178 d.

S. Zahedi et al. / Bioresource Technology 129 (2013) 85–91 87

2.4. Feeding

The tested substrate was the OFMSW from the 30 mm trommelof the MSW treatment plant in Cadiz, Spain.

Two different batches of OFMSW (Table 2) were used: Batch Ifor HRTs from 10 d to 2 d and Batch II for HRTs of less than 1 d.The OFMSW was stored in 25 kg drums at �4 �C to avoid anaerobicdegradation by the microorganisms found in the solid waste itself.The total solid concentration of the feed was adjusted to 20%(which is characteristic of dry anaerobic digestion) by adding tapwater. NaOH 10 M was added to the substrate when the pH ofthe effluent was below 5.3.

The substrate used for HRTs of 10 d–2 d (Batch I) had a highercontent in organic matter and microorganisms than the substrate

Table 2Composition of the substrate used in the experiment tested for HRTs of 10 d to 2 d(Batch 1) and HRTs of 1 d to 0.25 d (Batch II).

Parameter Batch I Batch II

Average Range Average Range

pH 5.5 (0.7) 4.8–6.2 5.8 (0.3) 5.4–6.2CODD (g/l) 23 (4) 19–27 15 (1) 14–16CODT (g/l) 75 (4) 71–79 45 (6) 39–51TVFA (g acetic/l) 3.1 (1.0) 2.0–4.0 0.7 (0.1) 0.6–0.8Acetate (g/l) 1.7 (1.1) 2.8–0.6 0.6 (0.1) 0.5–0.7Propionate (g/l) 0.4 (0.7) 0.1–1.7 0.0 (0.0) 0.0–0.0Butyrate (g/l) 0.1 (0.0) 0.1–0.2 0.1 (0.0) 0.0–0.2TVS (g/kg) 84 (9) 75–93 55 (7) 48–62DAPI (107 cells/ml) 243 (19) 220–261 90 (10) 80–100

used for HRTs of 1 d–0.25 d (Batch II). This is due to seasonal var-iability in the OFMSW.

2.5. Chemical and microbial analyses

The volume and composition of biogas were determined daily,the biogas produced being quantified using a gas flow meter (RitterTG1) and a gas suction pump (KNF Laboport). Gas chromatographywas used to analyze the different components of the biogas. Thegases analyzed were: H2, CH4, CO2, O2, N2 and H2S (GC-2010 Shi-madzu). The first five components were analyzed by means of athermal conductivity detector (TCD) using a Supelco Carboxen1010 Plot column. A Supelco Supel-Q Plot column and a flame pho-tometric detector (FPD) were used for H2S. Samples were takenusing a 1 ml Dynatech Gastight gas syringe under the followingoperating conditions: Split = 100; constant pressure in the injec-tion port (70 kPa); 2 min at of 40 �C; ramped at 40 �C/min until200 �C; 1.5 min at 200 �C; detector temperature: 250 �C; injectortemperature: 200 �C. The carrier gas was helium employing a flowrate of 266.2 ml/min.

A commercial mixture of H2, CH4, CO2, O2, N2 and H2S (AbellóLinde S.A.) was used to calibrate the system.

The following analytical determinations were performed tomonitor and control the process in the substrate and the effluent:TVS, pH, total chemical oxygen demand (CODT), dissolved chemicaloxygen demand (CODD) and VFA. The pH was measured daily usinga Crison 20 Basic pH-meter. TVS, CODT, CODD and VFA were ana-lyzed three times a week for a 10 d–4.4 d HRT; daily in the 3 and

88 S. Zahedi et al. / Bioresource Technology 129 (2013) 85–91

1.5 d HRT; and several times a day for HRTs of less than 1.5 d.These determinations were performed according to APHA (1995)and Fdez-Güelfo et al. (2010).The percentage of solubilizationwas calculated according to the following equation:

ðCODD effluent � CODD substrateÞCODD substrate

� 100 ð1Þ

Organic matter removal was calculated as the percentage differ-ence between the TVS of the substrate and the TVS of the effluentwithin the substrate TVS. Total acidity was calculated by additionof the individual fatty acids.

Microbial population dynamics were studied during the stabil-ization periods at each HRT. Epifluorescence microscopy employ-ing DAPI (40, 60-diamidine-2-phenylindole) as fluorochrome wasused to determine the total number of microorganisms in thereactors.

3. Results and discussion

3.1. Characterization of the reactor

3.1.1. pHpH remained approximately stable (5.3 ± 0.5), although a slight

decrease in its value was observed when the OLR was increased to110 g TVS/l/d. This was the optimal pH for enhanced hydrolyticand acidogenic bacteria activity (Verrier et al., 1987; Bouallaguiet al., 2004; Demirer and Chen, 2004; De la Rubia et al., 2009).

3.1.2. VfaFig. 3 shows the VFA composition, expressed as a percentage of

total volatile fatty acids (TVFA), for all HRTs assayed. The concen-tration of each VFA is shown in Table 3. TVFA concentration in thisreactor (12 g acetic/l) increased (22 g acetic/l) when the HRT de-creased from 10 d to 2 d. However, below an HRT of 2 d, TVFA con-centrations decreased.

In general, the dominant fermentation product was butyric acidfor all the tested HRTs. The dominant fermentation products forhigher OLRs (from 55 to 220 g TVS/l/d) were butyric acid and aceticacid, ranging from 63–72% and 24–32%, respectively, while thesum of produced acetic acid and butyric acid was relatively similar,approximately 95–99% of the TVFA. The ratio of butyrate to acetate(ratio B/A) varies (0.4–2.1) with each HRT, but is similar to thoseobtained by Ueno et al. (1996) and Shin et al. (2004). However,the acidogenic reactor continued working normally when this rela-tion was lower (HRT 4.4 d) or higher (HRT 1 d), and even for a 4.4 dHRT, when high concentrations of caproic acid were observed(Fig. 3, Table 3). Although high concentrations of caproic acid can

Fig. 3. Volatile fatty acid composition for all HRT assayed, expressed as percentageof TVFA production in g acetic/l.

be lethal to acidogenic microorganisms (Rinzema et al., 1994), thiswas not found to be the case in the present study. Most likely, themicroorganisms were well acclimatized to this substrate. The high-est B/A ratio, 2.1, was obtained at an HRT of 1 d when the change inOFMSW was produced.

Propionic acid concentration in the reactor effluent was not sig-nificant from a very short period of time onward, its concentrationremaining stable or even decreasing when the HRT decreased.

The increase in OLR did not produce any increase in the TVFA,below an HRT of 2 d, as might be expected (De la Rubia et al.,2009; Ponsá et al., 2008b). Nevertheless, the acidogenic effluentis suitable for the methanogenic step in all cases, as evidencedby the distribution of acid fermentation products obtained (Ucisikand Henze, 2008).

3.1.3. TVS and CODD

The TVS and CODD of each tested HRT are shown in Table 3.CODD concentration in this reactor (28 g CODD/l) increased (42 gCODD/l) when the HRT decreased from 10 d to 2 d. These resultsare in line with those obtained by De la Rubia et al. (2009). How-ever, the percentage of solubilized organic matter remains con-stant between 3 d and 0.5 d, with a CODD solubilization of55% ± 6%. TVS removal percentages decreased in HRTs below0.5 d. The decomposition of TVS was over 23% in the acidogenicprocess, although it did not exceed 11% at HRTs below 0.5 d. Thelow removal of volatile solid is due to the early stages of anaerobicdegradation, when there is practically no consumption of organicmatter. These results are in line with those obtained by Ponsáet al. (2008a) in the hydrolytic-acidogenic anaerobic digestionstage of sewage sludge.

3.2. Biogas and hydrogen production

The results of gas production (GP), percentage of hydrogen inbiogas (% H2), HP and specific hydrogen production (SHP) areshown in Table 4. The biogas produced was composed of hydrogenand carbon dioxide, with no methane or sulfide being detected un-der any of the tested conditions. The highest values of GP, % H2, HPand SHP were obtained at an OLR of 110 g TVS/l/d and a 0.5 d HRT.In this period, the B/A ratio was 1.6.

The influence of the OLR on the HP was evaluated. A linear rela-tionship was observed between the concentration of OLR and theHP from 9 g TVS/l/d to 110 g TVS/l/d (Fig. 4). The correlation coef-ficient, R2, was greater than 0.99, indicating a perfect fit to exper-imental data.

Below an HRT of 2 d, the % H2 in the biogas was clearly higherthan the values reported in the literature, ranging typically from25–40% (De la Rubia et al., 2009; Zhu et al. 2008; Valdez-Vazquezet al., 2005).

The SHP values obtained by Kim et al. (2008) and Cavinato et al.(2011) in anaerobic digestion (AD) of biowaste employing CSTRand OLR (35 g TVS/l/d, 21 g TVS/l/d) were 25.8 l H2/kg TVS and2.7 l H2/kg TVS, respectively. These SHP results are lower thanthose obtained in this research work (Table 4).

In this study, it was possible to work at HRT as low as those se-lected in low solid content AD. AD of feedstock at sucrose and glu-cose (Zhang et al., 2007) allows working at lower HRT than withhigh solid content feedstocks (Shin et al., 2004) and even moreso in systems where the SRT is not equal to the HRT (Kim et al.,2008). Shin and Youn (2005) compared the H2 fermentation perfor-mance of substrate waste at HRTs of 2 d, 3 d, and 5 d, an HRT of 5 dshowing the highest SHP. This is because the optimal HRT wasmuch longer when treating solid-type feedstock, seeing as a hydro-lysis step is required. In the present study, however, it was possibleto treat a solid-type feedstock at the typical HRT of a liquid-typesubstrate (8–12 h) (Hawkes et al., 2002). The contribution of

Table 3Characterization of effluents for different HRTs assayed.

HRT (d) 10 6.6 4.4 3 2 1 0.75 0.5 0.25

pH 5.7 ± 0.1 5.3 ± 0.4 5.7 ± 0.3 5.8 ± 0.4 5.7 ± 0.4 5.1 ± 0.0 5.0 ± 0.2 4.8 ± 0.2 5.1 ± 0.0CODD (g/l) 28 ± 3 35 ± 3 35 ± 3 42 ± 2 41 ± 2 23 ± 3 23 ± 3 24 ± 3 21 ± 1TVS (g/kg) 50 ± 8 70 ± 8 63 ± 4 65 ± 4 67 ± 3 40 ± 2 41 ± 1 49 ± 4 52 ± 1TVFA (g acetic/l) 12 ± 2 15 ± 1 17 ± 1 22 ± 1 21 ± 3 8 ± 1 7 ± 1 7 ± 0 5 ± 1Acetate (g/l) 3.7 ± 0.3 3.5 ± 0.4 4.6 ± 0.4 4.7 ± 0.5 6.0 ± 1.0 1.8 ± 0.2 2.1 ± 0.2 2.1 ± 0.1 1.6 ± 0.2Propionate (g/l) 0.6 ± 0.1 0.4 ± 0.1 0.5 ± 0.1 1.0 ± 0.1 0.7 ± 0.1 0.1 ± 0.0 0.1 ± 0.0 0.0 ± 0.0 0.0 ± 0.0Butyrate (g/l) 3.8 ± 0.4 5.0 ± 1.0 2.0 ± 0.1 6.4 ± 0.9 7.4 ± 0.8 3.7 ± 0.0 2.9 ± 0.0 3.3 ± 0.3 2.3 ± 0.7Caprionate (g/l) 1.0 ± 0.1 1.8 ± 0.3 3.9 ± 0.8 3.2 ± 0.5 0.5 ± 0.1 0.1 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0

Table 4Analysis of hydrogen production in the acidogenic anaerobic digestion of OFMSW.

HRT GP H2 HP SHPl/l/d % l H2/l/d l H2/kg TVSa

10 0.37 ± 0.07 27 ± 4 0.10 ± 0.02 13 ± 36.6 0.75 ± 0.20 36 ± 5 0.27 ± 0.07 20 ± 54.4 1.07 ± 0.18 39 ± 6 0.42 ± 0.07 23 ± 43 2.29 ± 0.49 41 ± 2 0.93 ± 0.20 31 ± 72 3.12 ± 0.43 47 ± 5 1.45 ± 0.20 33 ± 51 5.19 ± 0.20 50 ± 0 2.59 ± 0.10 47 ± 10.75 6.20 ± 0.07 54 ± 1 3.35 ± 0.04 46 ± 10.5 10.48 ± 0.37 55 ± 2 5.66 ± 0.14 52 ± 10.25 2.10 ± 0.61 49 ± 3 1.01 ± 0.30 5 ± 1

a Total volatile solids in the substrate.

Fig. 4. Relationship between OLR (g TVS/l/d) and HP (l H2/l/d).

S. Zahedi et al. / Bioresource Technology 129 (2013) 85–91 89

indigenous flora in the substrate could allow such low HRT to beachieved (Table 2). For this reason, the HRT achieved was muchlower even than those obtained in continuous systems (Uenoet al., 2007) and in systems with biomass retention (Kim et al.,2008). Moreover, a review of the literature (Jung et al., 2011) indi-cates that the shortest HRT achieving a stable, high-performanceprocess was 1.2 d, for the treatment of a mixture of substrate wasteand paper waste operated under thermophilic conditions (60 �C)(Ueno et al., 2007). It was found in the present study that the bestresult was achieved at an HRT of 0.5 d, obtaining higher values ofHP and SHP. Table 5 shows a comparison between these resultsand those obtained by Ueno et al.

Table 5Comparison between this study and the one carried out by Ueno et al.

Feedstock Feed/Draw HRT(h)

OLR (kg COD/m3/d

Garbage amended with shreddedpaper waste

Continuous 28.8 97

OFMSW Semi-continuous 12.0 90

3.3. Microbial population dynamics

Anaerobic digester effluent from single-phase dry-thermophilicanaerobic digestion of the organic fraction of municipal solid wastewas used as inoculum. The evolution of the microbial populationwas studied during the stabilization periods in thermophilic-dryanaerobic digestion in all the tested OLR. Results of microorganismconcentrations are shown in Table 6. The concentration of microor-ganisms in the starting inoculum is 482�107 cells/ml. In the first27 days, before establishing an HRT of 10 d, the reactor content isacidified in order to inhibit the methanogenic population, whichis subsequently removed (washed) from the system. During thisinitial phase, the microorganism population inside the reactor thusdecreases to 200�107 cells/ml.

In the first period (HRT from 10 d to 2 d), the microbial commu-nity increased slightly with increasing OLR. In the second period(HRT from 1 d to 0.25 d), a significant decrease in the microbialpopulation took place as the result of the low content of microor-ganisms in the waste (Table 2).

The observed increase in microorganisms between the HRT of1 d and 0.5 d is due to increased microbial activity, which allowsthese microorganisms to have higher growth rates.

Two correlations were obtained between hydrogen productionand microorganism concentrations. The results of microbial popu-lation concentrations afforded a positive linear correlation with HPin the digester for both OFMSW (R2 = 0.875, R2 = 0.848). These rela-tionships depend on the characteristics of the microbial content ofthe substrate: one linear relationship was observed for the highmicrobial content OFMSW (Batch I), and another for the low micro-bial content OFMSW (Batch II) (Fig. 5a and b).

The change in OFMSW did not affect biogas or hydrogen pro-duction. In fact, both the production of biogas and its percentagein hydrogen increase with the applied organic load. These resultsseem to show that the activity of anaerobic microorganisms inthe reactor could be more related to the OLR than to microbial con-centrations, as will be discussed later.

Microbial activity was determined by comparing the amount ofhydrogen generated for each loading rate with the size of popula-tion in the acidogenic reactor by DAPI staining. The results areshown in Table 6. There is a high correlation between OLR andmicrobial activity throughout the entire study period (Fig. 6). Thecorrelation coefficients were greater than 0.9, indicating a perfectfit to experimental data. Microbial activity increased throughout

) pH Tª HP (l H2/l/d) SHP (l H2/kg COD) Reference

5.8–6.0 60 5.4 46 Ueno et al. (2007)

4.7–5.0 55 5.7 ± 0.1 63 ± 1 This paper

Table 6Microbial community in a thermophilic-dry anaerobic reactor for different OLR assayed. (All the results shown are the average for each HRT assayed).

HRT 10 6.6 4.4 3 2 1 0.75 0.5 0.25

DAPI (107 cells/ml) 155 ± 5 228 ± 9 270 ± 67 324 ± 78 360 ± 46 165 ± 36 230 ± 42 250 ± 30 142 ± 50Activity (l H2�10�13/cells/d) 0.6 1.2 1.6 2.9 4.0 15.7 14.5 22.6 7.1

Fig. 5. (a) Relationship between HP (l H2/l/d) and total microorganisms (cells/ml)for HRT between 10 d and 2 d. (b) Relationship between HP (l H2/l/d) and totalmicroorganisms (cells/ml) for HRT from 1 d to 0.25 d.

Fig. 6. Relationship between OLR (g TVS/l/d)) and microbial activity (l H2/cell/d).

90 S. Zahedi et al. / Bioresource Technology 129 (2013) 85–91

the assayed period with OLR, being much higher when the micro-bial content in the reactor decreased due to the low microbial con-tent of the OFMSW. In systems with no biomass retention, adecreased HRT is reflected by faster rates of dilution and henceby a greater number of microorganisms exiting the system dailyin the effluent. Consequently, a large amount of substrate is con-sumed to keep the population size steady as a result of a young,very active population inside the reactor. This situation is intensi-fied when the systems work under no limiting substrateconditions.

Although the determination of the number of microorganismsis important in many microbial ecology studies (Zabriskie and

Humphrey, 1978; Hanning et al., 2007), these papers do not as-sess the activities associated with the total population. Undersome conditions, microbial number and activity show propor-tional correlations, whereas this is not the case under many real-istic circumstances. This requires caution and critical thinkingwhen one parameter is calculated or estimated from another.

This study shows that the increase in microbial activity insidethe reactor is directly proportional to the OLR (or inversely propor-tional to the HRT) and inversely proportional to the size of themicrobial population in the system.

4. Conclusions

The optimum operating condition for semi-continuous hydro-gen production from the OFMSW was obtained at 110 g TVS/l/d(90 g CODT/l/d), 0.5 d HRT and pH 5.5, where the H2%, HP, SHP were55%, 5.66 l/l/d, and 52 l H2/kg TVS, respectively.

The treatment may be carried out in an optimal range for aliquid-type substrate (8–12 h) thanks to the contribution of activemicroorganisms in the waste.

The increase in microbial activity is directly proportional to theOLR.

The bacterial population inside the reactor is strongly influ-enced by the microbial content of the OFMSW.

Acknowledgements

This work was supported by the Spanish Ministry of Scienceand Innovation (MICINN) (CTM2007-62164) and the Innovation,Science and Enterprise Department of the Regional Governmentof Andalucia (P07-TEP-02472). Both projects were co-financed byEuropean Regional Development Funds (ERDF). S. Zahedi gratefullyacknowledges funding from the Spanish Ministry of Science andInnovation (MICINN) (AP2008-01213).

References

APHA, AWWA, WEF, 1995. Standard Methods for the Examination of Water andWastewater, nineteenth ed. Washington, DC, New York, USA.

Bouallagui, H., Torrijos, M., Godon, J.J., Moletta, R., Ben Cheikh, R., Touhami, Y.,Delgenes, J.P., Hamdi, M., 2004. Two-phases anaerobic digestion of fruits andvegetable wastes: bioreactors performance. Biochem. Eng. J. 21, 193–197.

Cavinato, C., Bolzonella, D., Fatone, F., Cecchi, F., Pavan, D., 2011. Optimization oftwo-phase thermophilic anaerobic digestion of biowaste for hydrogen andmethane production through reject water recirculation. Bioresour. Technol.102, 8605–8611.

De la Rubia, M.A., Raposo, F., Rincón, B., Borja, R., 2009. Evaluation of the hydrolytic–acidogenic step of a two-stage mesophilic anaerobic digestion process ofsunflower oil cake. Bioresour. Technol. 100, 4133–4138.

Demirer, G.N., Chen, S., 2004. Effect of retention time and organic loading rate onanaerobic acidification and biogasification of dairy manure. J. Chem. Technol.Biotechnol. 79, 1381–1387.

Fdez-Güelfo, L.A., Álvarez-Gallego, C.J., Sales, D., Romero, L.I., 2010. Dry-thermophilic anaerobic digestion of organic fraction of municipal solid waste:methane production modeling. Waste Manage. 32, 382–388.

Hanning, C., Hanning, M., Rehmer, O., Braun, G., Hellwing, E., Al-Ahmad, A., 2007.Fluorescence microscopic visualization and quantification of initial bacterialcolonization on enamel in situ. Arch. Oral Biol. 52, 1048–1056.

Hawkes, F.R., Dinsdale, R., Hawkes, D.L., Hussy, I., 2002. Sustainable fermentativehydrogen production: challenges for process optimization. Int. J. Hydrog.Energy 27, 1339–1347.

Jung, K.W., Kim, D.H., Kim, S.H., 2011. Bioreactor design for continuous darkfermentative hydrogen production. Bioresour. Technol. 102, 8612–8620.

Kapdan, I.K., Kargi, F., 2006. Bio-hydrogen production from waste materials. EnzymeMicrob. Technol. 38, 569–582.

S. Zahedi et al. / Bioresource Technology 129 (2013) 85–91 91

Kepner, R.L., Pratt, J.R., 1994. Use of fluorochromes for direct enumeration of totalbacteria in environmental samples: past and present. Microbiol. Rev. 58, 603–615.

Kim, S.H., Han, S.K., Shin, H.S., 2008. Optimization of continuous hydrogenfermentation of food waste as a function of solids retention time independentof hydraulic retention time. Process Biochem. 43, 213–218.

Montero, B., García-Morales, J.L., Sales, D., Solera, R., 2008. Evolution ofmicroorganisms in thermophilic-dry anaerobic digestion. Bioresour. Technol.99, 3233–3243.

Montero, B., García-Morales, J.L., Sales, D., Solera, R., 2009. Analysis of methanogenicactivity in a thermophilic-dry anaerobic reactor: use of fluorescent in situhybridization. Waste Manage. 29, 1115–1144.

Moosbrugger, R.E., Wentzel, M.C., Ekama, G.A., Marais, G.R., 1993. Weak acid/basesand pH control in anaerobic systems-review. Water SA. 19, 1–10.

Ponsá, S., Ferrer, I., Vázquez, F., Font, X., 2008a. Optimization of the hydrolytic–acidogenic anaerobic digestion stage (55 �C) of sewage sludge: Influence of pHand solid content. Water Res. 42, 3972–3980.

Rinzema, A., Boone, M., Van Knippenberg, K., Lettinga, G., 1994. Bactericidal effect oflong chain fatty acids in anaerobic digestion. Water Environ. Res. 66, 40–49.

Shin, H.S., Youn, J.H., Kim, S.H., 2004. Hydrogen production from food waste inanaerobic mesophilic and thermophilic acidogenesis. Int. J. Hydrog. Energy 29,1355–1363.

Shin, H.S., Youn, J.H., 2005. Conversion of food waste into hydrogen by thermophilicacidogenesis. Biodegradation 16, 33–44.

Tchobanoglous, G., Hilary, T., Vigil, S.A., 1997. Integrated Solids Waste Management:Engineering Principles and Management Issues. McGraw-Hill, Inc.

Ucisik, A.S., Henze, M., 2008. Biological hydrolysis and acidification of sludge underanaerobic conditions: the effect of sludge type and origin on the production andcomposition of volatile fatty acids. Water Res. 42, 3729–3738.

Ueno, Y., Otsuka, S., Morimoto, M., 1996. Hydrogen production from industrialwastewater by anaerobic microflora in chemostat culture. J. Ferment. Bioeng.82, 194–197.

Ueno, Y., Fukui, H., Goto, M., 2007. Operation of a two-stage fermentation processproducing hydrogen and methane from organic waste. Environ. Sci. Technol. 41,1413–1419.

Valdez-Vazquez, I., Rios-Leal, E., Esparza-Garcia, F., Cecchi, F., Poggi-Varaldo, H.,2005. Semi-continuous solid substrate anaerobic reactors for H2 productionfrom organic waste: mesophilic versus thermophilic regime. Int. J. Hydrog.Energy 30, 1383–1391.

Valdez-Vazquez, I., Ríos-Leal, E., Muñoz-Páez, K.M., Carmona-Martínez, A., Poggi-Varaldo, H.M., 2006. Effect of inhibition treatment, type of inocula andincubation temperature on batch H2 production from organic solid waste.Biotechnol. Bioeng. 95, 342–349.

Verrier, D., Roy, F., Albagnac, G., 1987. Two-phase methanization of solid vegetableswastes. Biol. Wastes 22, 163–177.

Zabriskie, D.W., Humphrey, A.E., 1978. Estimation of fermentation biomassconcentration by measuring culture fluorescence. Enzyme Microb. Technol.35, 337–343.

Zhang, Z.P., Show, K.Y., Tay, J.H., Liang, D.T., Lee, D.J., Jiang, W.J., 2007. Rapidformation of hydrogen-producing granules in an anaerobic continuous stirredtank reactor induced by acid incubation. Biotechnol. Bioeng. 96, 1040–1050.

Zhu, H., Stadnyk, A., Bèland, M., Seto, P., 2008. Co-production of hydrogen andmethane from potato waste using a two-stage anaerobic digestion process.Bioresour. Technol. 99 (11), 5078–5084.