Bioresource Technology 144 (2013) 513–520

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

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

Feasibility of anaerobic co-digestion of poultry blood with maizeresidues

0960-8524/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biortech.2013.06.129

⇑ Corresponding author. Tel.: +34 660736200.E-mail addresses: xagomb@unileon.es, marta.otero@ua.pt (X. Gómez).

M.J. Cuetos a, X. Gómez a,⇑, E.J. Martínez a, J. Fierro a, M. Otero b,c

a Chemical and Environmental Bioprocess Engineering Department, Natural Resources Institute (IRENA), University of León, Avda Portugal 41, 24009 León, Spainb Centre for Environmental and Marine Studies (CESAM), Department of Chemistry, University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugalc Department of Applied Chemistry and Physics, University of León, Campus de Vegazana, 24071 León, Spain

h i g h l i g h t s

� Poultry blood and maize were digested under batch and semi-continuous conditions.� TG, SEM and FTIR analyses were useful in evaluating digestion performance.� Methane yields of up to188 ± 21 L CH4/kg VS were obtained under batch tests.� Accumulation of organic material was observed under semi-continuous operation.

a r t i c l e i n f o

Article history:Received 8 May 2013Received in revised form 27 June 2013Accepted 29 June 2013Available online 12 July 2013

Keywords:Co-digestionResidual poultry bloodMaize residuesFTIRKinetic analysis

a b s t r a c t

The potential of anaerobic digestion for the treatment of poultry blood was evaluated in batch assays atlaboratory scale and in a mesophilic semi-continuously fed digester. The biodegradability test performedon poultry blood waste showed a strong inhibition. Maize residues were used as co-substrate toovercome inhibition thanks to nitrogen dilution. Under batch operation, increasing the maize concentra-tion from 15% to 70% (volatile solids (VS) basis) provided an increase of biogas from 130 ± 31 to188 ± 21 L CH4/kg VS. In the semi-continuous mesophilic anaerobic digester, the biogas yield was165 ± 17 L CH4/kg VS fed, as a result of strong volatile fatty acid (VFA) accumulation. Although physicalmodifications of maize particles were observed by Scanning Electron Microscopy (SEM), an incompletedegradation was confirmed from analysis of digestates. Furthermore, Fourier Transform Infrared (FTIR)spectroscopy analysis demonstrated that along with VFA build-up, an accumulation of non-degradedmaterials took place.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction Few studies in literature deal with the digestion of residual

Anaerobic co-digestion of slaughterhouse wastes has been con-sidered a feasible alternative for increasing biogas potential inconventional digesters. However, the use of slaughterhouse wastesas co-substrates must fulfil requirements established by AnimalBy-Products Regulations (ABPR, EC 1069/2009; 142/2011). Slaugh-terhouse wastes are characterized by presenting high nitrogenconcentrations. This fact makes imperative the addition of a co-substrate in order to achieve mixtures with balanced C/N ratio thatallows a decrease in nitrogen concentration and enhance biogasyields as consequence (Cuetos et al., 2009; Lehtomäki et al.,2007). Although nitrogen is an essential nutrient for anaerobicmicroorganisms, ammonia inhibition has been frequently reportedwhen treating wastes with high nitrogen content and inhibitorylevels are accepted to be around 4 g N/L (Lobato et al., 2010; Reschet al., 2011).

blood (Salminen et al., 2000, 2001; Wang and Banks, 2003). Differ-ent mixtures containing blood and slaughterhouse wastes as co-substrates have been recently studied due to their high biochemicalmethane potential. In this line, Alvarez and Lidén (2008) studied thedigestion of mixed fractions of cattle rumen, stomach content andgut fill of swine and blood (with blood cow and blood swine being34% by total weight) with fruit–vegetable wastes and manure. Re-cently Marcos et al. (2010) determined the optimal operational con-ditions when digesting mixtures of wastewater with blood andsolid wastes from the meat industry (98% wastewater plus 2% bloodand 97% wastewater and blood plus 3% solid offal, respectively (v/v)). In a similar way, Zhang and Banks (2012) reported operatinglimits for the organic loading rates when studying the co-digestionof sheep blood, mechanically recovered organic fraction of munici-pal solid wastes and pig intestine with flotation fat.

The combination of energy crops and slaughterhouse wastes of-fers the possibility of increasing biogas production of existing facil-ities not only by the effect of balancing nutrients but also by theincrease in organic loading rate (Nges et al., 2012). Maize,

514 M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520

sunflower, grass and sudan grass are the most commonly usedenergy crops (Amon et al., 2007a), with maize being the most dom-inating crop for biogas production (Amon et al., 2007b). Althoughthere are several studies about conversion of energy crops asmono-substrates to biogas (Demirel, 2009; Klocke et al., 2007),the problems associated with the lack of micro and macro nutri-ents and buffering capacity lead to obtaining more stable systemswhen co-digesting crop biomasses with manures (Cuetos et al.,2011; Lehtomäki et al., 2007).

On the other hand, biochemical methane potential (BMP) teststhat are performed to evaluate the biogas potential of wastes aretime consuming. The use of fast and reliable methods that mayindicate the adequacy of mixing certain substrates may be a usefultool when deciding the selection of co-substrates and preventingproblems associated to degradation of complex compounds andthe time needed for complete digestion of substrates in order to re-duce post-digestion emissions. This is the case of the use of ther-mal analysis and Fourier Transform Infrared (FTIR) spectroscopy.These methods have been widely used as a way for evaluatingthe quality of organic material in an attempt to predict the matu-rity of biologically stabilised samples (Cuetos et al., 2009; Tintneret al., 2012).

This paper focus, for the very first time, on the anaerobic diges-tion of poultry blood and mixtures of blood with leaves of maizeplants in combination with the proposal of using thermogravimet-ric kinetic analysis as a ool for the characterisation of substratesprior to digestion experiments. The anaerobic degradation of maizewas studied along with the assessment of increments in biogasyield with the addition of blood as co-substrate.

2. Methods

2.1. Inoculum and substrate sources and characterisation

The inoculum used was obtained from an anaerobic slaughter-house waste digester adapted to an ammonium-rich environment(Cuetos et al., 2009). This laboratory digester treated a mixture ofpoultry blood and OFMSW and was operated during 75 days atan HRT of 36 days. The total solid (TS) content of the inoculumwas 41 g/L and volatile solid (VS) content was 29 g/L. The poultryblood was obtained from a local poultry slaughterhouse in León(Spain) and then pasteurised (60 min, 70 �C) prior to its use indigestion experiments. The maize used (Zea mays L.) was harvestedand dried at room temperature. Leaves of maize plants wereground to a particle size around 3 mm to increase the superficialarea and the accessibility for microbial action (Palmowski andMüller, 2000).

2.1.1. Analytical techniques used in the characterisation of substratesKjeldahl nitrogen and organic carbon analysis were performed

according to MAPA (1994). Protein content was calculated fromthe Kjeldahl-N content using a conversion factor of 6.25. Lipid con-tent was determined using a Standard Soxhlet method (APHA,1998).TS, VS and ammonia were determined in accordance withStandard Methods (APHA, 1998). Fibre characterisation was carriedout by determination of cellulose, hemicellulose and lignin analy-sis as described by Van Soest et al. (1991). Acid detergent fibre(ADF), neutral detergent fibre (NDT) and crude fibre were deter-mined using an ANKOM200 fibre analyser.

2.1.2. Thermogravimetric kinetic analysisThermal analysis was carried out in a TA Instruments SDT2960

equipment. Samples were further crushed using a 200MM ball millRetch. The heating rates (b) applied were of 5, 10, 25 and 50 K/minup to 1000 K under inert atmosphere. These dynamic runs were

carried out on a pan containing approximately 5 mg of the sampleand a reference crucible containing the same mass of calcined cal-cium oxide. Three replicates were run to calculate mean values.During temperature-programmed runs, a continuous flow of100 mL/min of Nitrogen (purity P99.9994%) at a manometricgauge pressure of 1 atm (101 kPa) was fed into the furnace.

The description of the rate of heterogeneous solid-state reac-tions can be found in Otero et al. (2011). In this work, two differentisoconversional models were applied to non-isothermal thermo-gravimetric data from the temperature-programmed combustionof the samples. Using dynamic integral TG curves obtained at fourdifferent b, in the same way as described elsewhere (Otero et al.,2008), and applying these models, the activation energy valuesand reaction order associated to the pyrolysis of maize and bloodsamples were determined. In addition, the use of the independentparallel first-order reactions (IPR) model (Sørum et al., 2001) wasapplied to DTG curves obtained from substrates. In this model,the decomposition of biomass is described by three independentparallel reactions that can be associated to the decomposition ofthe constituent components.

2.2. Anaerobic digestion tests

2.2.1. Batch digestion experimentsBatch tests were performed at different proportions of maize

(15%, 40% and 70% VS) in the mixture with blood. Experimentswere performed in 100 mL Erlenmeyer flasks incubated at34 ± 1 �C in a water bath under stirring conditions (200 rpm). Theinoculum to substrate ratio was kept in the range of 1–2 to avoidthe addition of alkali solution for pH correction and volatile fattyacids (VFAs) overloading. Reactors were denoted as M_15, M_40and M_70 based on the proportion of maize added to the mixture.For each assay 20 replicates were initially set and two replicateswere withdrawn from the water bath at days 1, 3, 5, 8, 11, 15,18, 22, 25 and 30. The volume of biogas produced was measuredby means of liquid displacement bottles. Values obtained were cor-rected to standard temperature and pressure.

The digestion of blood was also carried out in batch assays. Inthis case, 16 replicates were run during 20 days. Two replicateswere withdrawn from the bath for liquid-phase analysis at days1, 3, 7, 9, 11, 15 and 20.

In addition, control assays were run in parallel to measure thebackground methane production from the inoculum and the biogasproduction from maize. The residual biogas was subtracted fromthe total production in each case. The biogas production obtainedfrom maize was used for comparing results obtained from thermo-gravimetric analysis and to evaluate results from co-digestion as-says. Table 1 shows a description of the batch and semi-continuous experiments carried out.

Cumulative biogas curves were fitted to a modified GompertzEq. (1), which is a suitable model for describing the process ofcumulative biogas production in batch experiments (Nopharatanaet al., 2006):

PðtÞ ¼ Pmax: exp � expRmax:ePmax

ðk� tÞ þ 1� �� �

ð1Þ

where P(t) is the cumulative biogas production (l); Pmax is the biogasproduction potential (l), Rmax is the maximum biogas productionrates (l/d) y k is lag-phase time (d) and e is 2.71. The software Origin6.0 was used to fit the equation and determine Pmax, Rmax and k. Thebiogas yield (expressed as L/kg VS) and the maximum specificbiogas production (SBP) rate (expressed as L/kg VS d) were obtainedby dividing P and Rmax by the VS content of the substrate.

Table 1Description of batch and semi-continuous experiments performed to evaluate the digestion of poultry blood and maize leaves.

Experiment Substrate ratio Maize: Blood (%) Digestion time (d) Configuration

Poultry blood 0:100 20 BatchMaize leaves 100:0 35M_15 15:85 30M_40 40:60 30M_70 70:30 30Reactor 60:40 HRT: 36 d Semi-continuous

Table 2Characteristics of substrates used in digestion experiments.

Parameters Blood Maize

Total organic carbon (% db) 34.2 47.3Kjeldahl nitrogen (% db) 12.0 0.8C/N ratio 2.8 59.1Total ammonia (g/L) 16.6 NATotal fat (% db) 0.2 NATotal protein (% db) 7.5 NATS (%) 10.8 90.0VS (%) 9.3 84.0Cellulose (% db) NA 32.8Hemicellulose (% db) NA 44.1Lignin (% db) NA 1.9

NA: not analysed; db: dry basis.

M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520 515

2.2.2. Semi-continuous anaerobic digestionSemi-continuous co-digestion was carried out in a completely

mixed stirred digester, with a working volume of 3 L and ther-mostatised at 34 ± 1 �C. The reactor had a hydraulic retention time(HRT) of 36 d and an organic loading rate (OLR) of 3.1 g VS/L d,which were selected on the basis of previous results on the anaer-obic digestion of similar substrates (Cuetos et al., 2009, 2011). Thereactor was manually fed every day with a mixture of poultryblood and maize (with 60% of VS of the mixture being providedby maize). The content of TS and VS of the feed were 12.6% and11.3%, respectively. The reactor was operated for 75 days. Dailygas production was measured using a reversible device with liquiddisplacement and a wet-tip counter.

2.2.3. Monitoring and evaluation of the anaerobic digestionThe following parameters where analysed during digestions as-

says: gas volume, pH, TS, VS, alkalinity, ammonia, chemical oxygendemand (COD), VFA and biogas composition by gas chromatogra-phy, as described in Cuetos et al. (2009). Fibre digestion was eval-uated by Scanning Electron Microscopy (SEM). With this purpose,samples were dried in a furnace before being doubly coated withcarbon and gold in high vacuum (0.05–0.07 mbar) with a BlazersSCD 004 cathodic sputter coater and examined via secondary elec-trons at an accelerating voltage of 15 kV by a Jeol JSM-6480LVscanning electron microscope.

Fourier Transform Infrared (FTIR) spectroscopy analysis wasalso used for the evaluation of the digestion process. Samples weredried at 105 �C for 48 h. 2 mg of milled samples were ground upwith 200 mg KBr (FTIR grade) and homogenised in an agate mortar.KBr pellets were compressed under vacuum in a standard deviceunder pressure of 6000 kg/cm2 during 10 min. Infrared spectrawere recorded using a FTIR Perkin–Elmer 2000 spectrophotometerover the 4000–400 cm�1 range at a rate of 0.5 cm/s. Fifty scanswere collected, averaged for each spectrum and corrected againstambient air as background. Mean values were obtained from threereplicates for each sample. Spectra were vector-normalised forcomparison following the procedure proposed by Meissl et al.(2007).

3. Results and discussion

3.1. Characterization of substrates

The characteristic parameters of poultry blood and maize leavesused in this work are presented in Table 2. It may be seen that thepoultry blood is a substrate with a very low C/N ratio, while maizehas a much higher C/N ratio. Maize leaves have high content of or-ganic carbon (%) and VS (%), apparently indicating a high biogas po-tential. The leaves of maize plants were chosen due to their lowlignin content (1.9%), while the total content of lignocellulosicmaterial was quite high (78%). Since the content of lignin definesthe degree of biodegradability and thus the methane yield thatcan be produced through anaerobic digestion, it is expected a highdegradability of maize leaves under batch anaerobic tests. This factcombined with the high value of C/N ratio of maize leaves makestheir mixture with poultry blood adequate for digestion.

3.2. Thermogravimetric kinetic analysis

TG curves corresponding to the poultry blood and maize wastesare shown in Fig. 1 for the dynamic runs carried out at the heatingrates (b) of 5, 10, 25 and 50 K/min. Six percentages of conversion(a) with respect to the initial mass of the dry samples (after mois-ture mass loss) have been marked crossing these TG curves. Tem-peratures at which these a occur were considered for the kineticanalysis. As it may be observed (for both blood and maize), on rais-ing the temperature, thermal degradation takes place along with amass loss. The faster the heating rate, the higher the temperatureat which weight loss starts. The mass corresponding to the residueobtained at the end of the analysis remains stable, with the maizepresenting a slightly lower fraction of initial mass at the end of theheating process than the blood.

The plots of ln b vs. 1/T and ln b/T2 vs. 1/T corresponding to theseveral a considered together with the linear fittings to the modelproposed by Ozawa–Flynn–Wall and to that by Vyazovkin, respec-tively, are presented as Supplementary Material (Fig. S1). Activa-tion energy (AE) values obtained by the Ozawa–Flynn–Wall andby the Vyazovkin models, which are given in Table 3, presented agreat accordance, although, in all cases, the first one gave slightlyhigher values. As it may be seen, AE calculated was higher for poul-try blood than for maize. These values are comparatively lowerthan those reported for the pyrolysis of cardoon stems and leaves(Damartzis et al., 2011). Also, EA values here obtained were lowerthan those reported by Sonobe and Worasuwannarak (2008) forcellulose, even though maize leaves used in this study presenteda high content in cellulosic material, which may indicate that inter-actions among cellulose, hemicellulose and lignin should be affect-ing thermal degradation and this may in turn affect thedegradation under anaerobic digestion.

Regarding the use of the IPR model, three different reactions canbe ascribed to the thermal decomposition process (see Fig. 2). In thecase of maize leaves, these reactions are associated to hemicellulose(reaction 1), cellulose (reaction 2) and lignin (reaction 3) compo-nents (Damartzis et al., 2011). However, in the case of blood thisassignation is not possible, so reaction 3 can be translated into thedifficulty that may be encountered when digesting this substrate,since the contribution of this reaction to the DTG curve is high when

Fig. 1. TG curves corresponding to the temperature programmed combustion of (a)residual poultry blood and (b) maize at different heating rates. Six different a: 10%,20%, 30%, 40%, 50% and 60% are indicated by lines crossing over the TG curves.

Table 3Activation energy (AE) values obtained by the Vyazovkin and the Ozawa–Flynn–Wallkinetic models and reaction order (n) as a function of temperature for thetemperature programmed combustion of poultry blood and maize wastes used inthis work.

Sample AE Vyazovkina (kJ/mol)

AE Ozawa–Flynn–Walla (kJ/mol)

nb

Poultryblood

163.1 172.6 0.24

Maizeleaves

149.8 159.8 0.27

a Calculated as arithmetic average of AE values obtained for the different aconsidered.

b Calculated as arithmetic average of n values obtained at the different Tconsidered.

516 M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520

compared with the other two reactions in the same sample. In addi-tion, the contribution of reaction 3 to the DTG curve obtained frommaize sample also indicates interactions during the thermaldegradation process. The content in lignin obtained for this sampleis particularly low but the contribution of this reaction to the com-plete peak is about 32%. If this value is used as a measurement ofcarbon availability, this result indicates that about 53.6% of theorganic material is available for degradation (taking intoconsideration values reported in Table 2 on the cellulose, hemicellu-lose and lignin contents).

3.3. Anaerobic digestion

3.3.1. Batch digestion experimentsThe production of biogas obtained from the digestion assay of

poultry blood was 46.5 L/kg VS. This system presented high valuesof pH during the whole digestion assay (around 8.8) along with aconcentration of ammonium around 4500 mg/L, which resulted

in the presence of high levels of free ammonia (average value of1813 mg/L). Even though the inoculum used in the experimentwas adapted to high values of ammonium concentration, microor-ganisms were confronted with extremely high concentrations offree ammonia to which they were not used. These conditionstranslated into severe inhibitory problems explaining thus the poorperformance of the process.

In the case of maize leaves, the batch digestion test resulted in abiogas production of 262 L/kg VS (157.1 L CH4/kg VS), while thetheoretical value calculated using cellulose and hemicellulosecontent was 315 L CH4/kg VS, thus indicating low availability ofcarbon. Results obtained indicate that the characterisation of or-ganic matter in terms of lignocellulosic content is a poor approachfor evaluating the real biodegradability of the wastes used. If this isused to adjust the C/N ratio, it is far from realistic that these frac-tions are equivalent to inert and biodegradable fractions. Thermalanalysis demonstrated in this case that about half of the organicmaterial was available to microorganisms.

On the other hand, the addition of maize leaves in different pro-portions resulted in a significant improvement in the specificmethane production (SMP) when compared to results obtainedfrom blood digestion test. The increase in the content of maize inthe feeding mixture resulted in 4, 5 and 6 times more biogasthanks to adjusting C/N ratio. Values of AE obtained from kineticanalysis were higher for maize leaves, thus results from kineticanalysis may explain the increase in SMP with the increase inthe proportion of maize in the mixture. In addition, difficultiesfound in the digestion assay of blood may keep a relation withthermal degradation behaviour of this substrate.

Results from chemical parameters measured in batch tests arepresented in Table 4 and cumulative curves of specific biogas pro-duction (SBP) are shown in Fig. 3a and b. Values obtained for theSMP in this study were in accordance with results obtained by dif-ferent authors. In this line, Raposo et al. (2006) reported values of211 L/kg VS for the digestion of maize under batch conditions.However, these values are lower to those reported in previous re-search work when digesting maize residues and swine manure(Cuetos et al., 2011). The biogas production obtained for co-diges-tion system M_70 was higher than that of maize, thus indicating asynergistic effect. Biogas production patterns were similar to theones reported by Salminen et al. (2000) for batch degradation ofslaughterhouse waste. The most important difference observedfor all tests (Table 4) was the significant reduction of the initiallag phase compared to that obtained by the cited authors. The re-duced lag phase obtained in the present work may be explained bythe use of inoculum adapted to ammonium-rich medium. Thesetests showed a rapid hydrolysis/acidogenesis and methane produc-tion thanks to co-substrate addition and balanced C/N ratio. Theshape of the cumulative curve indicates that inhibitory stages

Reaction 1

Reaction 2

Reaction 3

(a)

Reaction 1Reaction 2

Reaction 3

(b)

Fig. 2. Comparison between experimental and calculated DTG curves using the IPRmodel for (a) residual poultry blood and (b) maize at a heating rate of 10 K/min.

M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520 517

can be discarded during the initial phase of the assays. This meansthat the effect of VFA build-up (Fig. 3c–e) observed at the begin-ning of the experiments (especially in the case of the M_70 sys-tem), was attenuated thanks to the high alkalinity and pH valuesof reactors, which were always above 7.5 during the experiments.

Table 4Results from batch experiments on the anaerobic digestion of maize and blood mixtures.

Parameters M_70

Gompertz model

P’max (L/kg VS) 311.4 ± 4.5R’max (L/kg VS d) � 10�3 20.6 ± 1.3k (d) 0.1 ± 0.3SMP (L CH4/kg VS) 188 ± 21Final values

pH 8.01 ± 0.1Alkalinity (mg/L) 7200 ± 282COD (g/L) 23.2 ± 1.2TS (g/L) 22.3 ± 0.9VS (g/L) 16.2 ± 1.5%VS removal 31.7 ± 2.3NH4

+ (mg/L) 1925 ± 129Free NH3 (mg/L) 267.5 ± 15.0

3.3.2. Semi-continuous digestion testsThe reactor operated under semi-continuous conditions was fed

with a mixture in which maize provided 60% of VS, based on re-sults obtained in batch experiments, which presented higher SMPvalues with the increase in the content of maize leaves. The mix-ture tested at 40% VS (M_40 batch experiments) presented VFA val-ues that could be easily overcome by methanogenic microflora. Inorder to avoid VFA build-up a mixture with a proportion lower to70% VS was decided to be tested under semi-continuous operation.In addition, in previous work reported by Cuetos et al. (2011),experiments of co-digestion with maize at a ratio of 50% VS pre-sented complete degradation of VFA after 70 days of operation.For these reasons, the value of 60% was considered appropriatefor the semi-continuous experiment in an attempt to fulfil thecompromise between possible VFA accumulation and C/N ratio.

This system presented a low biogas production at the beginningof the study, which increased progressively. The daily volume ofbiogas produced was 10 times higher at the end of the first 15 days.Once a period equivalent to an HRT had elapsed, the biogas profilebecame stable, with an average value of 2.7 L/d (Fig. 4a). However,the methane content in biogas was slightly lower than 60%. Withregard to VS content, there was an accumulation during the firstmonth. The concentration of VS rapidly increased from 13 to30 g/L during the initial stage of the experiment. Although the rateof accumulation decreased from day 20 onwards, the increasingtrend was maintained during the whole experiment. Despite theevidence of accumulation of maize particles, final VS removalreached a value of 72%.

The reactor presented an initial accumulation of acetic acid andposteriorly a high build-up of VFAs during the operation (Fig. 4b).This behaviour is another evidence that reactor was incapable ofreaching steady state conditions even though inoculum used wasadapted to digest poultry blood. C3–C5 acid forms presented anincreasing trend after 30 days of operation. Although the presenceof iso-forms is associated to instabilities, biogas production duringthe experiment presented a steady behaviour with pH values in therange of 7.0–7.5 and alkalinity values over 15 g/L. The ratio VFA-to-alkalinity was close to 1.0 at the end of the study, which leads toconsider the digestion as a failed anaerobic process. However,thanks to the buffering capacity provided by the residual blood,stable pH values were obtained in the process.

The trend observed for isovaleric and isobutyric acids keep aclose relation to the behaviour of acetic and propionic acid profiles.An acetic acid concentration greater than 1 g/L has been reportedas the threshold limit for inhibition of microorganism responsibleof butyric acid degradation, with total inhibition being observedat a concentration of 5 g/L. In this study, isovaleric and isobutyricacids presented an increasing trend during the experiment. The

M_40 M_15

251.7 ± 2.7 192.7 ± 1.718.6 ± 1.0 22.6 ± 1.0

2.6 ± 0.2 0.9 ± 0.2150 ± 32 103 ± 31

8.1 ± 0.1 8.1 ± 0.16650 ± 640 7150 ± 80817.6 ± 0.7 24.9 ± 1.616.1 ± 0.3 13.0 ± 1.210.7 ± 0.3 9.07 ± 1.337.3 ± 2.5 12.2 ± 0.8

2175 ± 250 2944 ± 287252.2 ± 34.4 409 ± 39.9

Fig. 3. (a) Cummulative curves of specific biogas production for maize leaves and (b) co-digestion systems and R2 values obtained from fitting Gompertz models toexperimental data obtained under batch mesophilic conditions. VFA concentration from anaerobic digestion batch experiments (c) M_70, (d) M_40 and (e) M_15.

518 M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520

acetic acid concentration measured in the reactor was above 2 g/Lduring the entire duration of the experiment, the exception being abrief period around 10th day of operation.

Fig. 4c represents the evolution of total ammonium nitrogen(TAN) and free ammonia (FA) in the semi-continuous process.The reactor showed increasing values of these parameters duringthe first HRT studied with TAN values becoming approximatelyconstant when the process was close to an end. In this study FAvalues were much lower than toxic limits reported in literaturefor digestion of rich-nitrogen wastes (Angelidaki and Ahring,1993b; Hansen et al. 1998). The low concentration of FA reachedin the system (<250 mg/L) was associated with pH values, whichwere in the range of 7.0–7.5.

Biogas yield was 280 ± 20 L/kg VS fed (methane yield of165 ± 17 L CH4/kg VS fed). This value was calculated from data ob-tained during the last HRT evaluated. In spite of reactor instability,the methane yield was similar to that obtained under batch condi-tions although a severe VFA build-up was taking place. In this line,even though anaerobic microflora was adapted to severe inhibitoryconditions caused by the presence of VFAs, a much higher yield of

methane would have been obtained if complete degradation of theorganic material had been reached. Similar peak values of aceticacid were measured during the batch digestion test denoted asM_70, thus indicating that methane yields obtained from experi-ments were affected by inhibitory VFA conditions. Theoreticalmethane yields calculated from chemical characteristics reportedin Table 2 are 352 and 342 L CH4/kg VS for the two mixtures at70% and 60% proportion of VS of maize. However, these valuesare reduced to 193 and 190 L CH4/kg VS when considering carbonavailability proposed by IPR model and are in accordance withthose obtained from experiments. This indicates that experimentalresults are far from attaining the complete degradation of the or-ganic material. In spite of this, results obtained in this study werein the same range of those reported by Hejnfelt and Angelidaki(2009) when evaluating a mixture of wastes containing blood.However, the values here obtained were much lower than those re-ported by other authors on the anaerobic digestion of slaughter-house wastes (Karlsson and Ejlertsson, 2012; Peu et al., 2012). Inthis work, the mixture of blood and maize presented high biogaspotential under batch tests but instability occurred during

Fig. 4. (a) Daily biogas production and VS removal, (b) VFA concentration and (c)ammonium concentration data obtained from semi-continuous co-digestion ofpoultry blood and maize (maize at 60% ratio).

M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520 519

semi-continuous digestion, even when the inoculum used camefrom a laboratory reactor adapted to residual blood as substrate.It is possible that incomplete adaptation may have contributed tothe poor performance of the semi-continuous reactor. However,it seems most feasible that the difficulties encountered are relatedto inhibitory effects associated to the high VFA levels in the reactor.It must be highlighted that the use of small particle size of maizeleaves during the different experiments performed implies a great-er accessibility of microflora to organic matter, which could havehad something to do with the high VFA levels reached. In any case,this is a hypothesis that should be tested by further work at a pilotscale and under different HRT.

Photographs obtained from SEM analysis of samples taken fromthe effluent of the digestion system demonstrated the incompletedegradation of fibre particles, as it may be seen in Fig. S2a and b(Supplementary material). In addition FTIR spectra obtained fromthe digested sample, which have been included as Supplementarymaterial (Fig. S2c), show the characteristic absorbance ascribed toproteins and cellulose-carbohydrate material of the substrates.When degradation of organic matter takes place, readily oxidable

compounds are degraded and the volatile content of the substrateis expected to be reduced whereas the degree of aromaticityshould increase (Cuetos et al., 2009). The failure of the digestionprocess is then reflected in FTIR spectra by the permanence of sig-nals ascribed to the substrate samples and high relative intensitiesof signals ascribed to carbohydrates (1070–1020 cm�1), C@O, C@Cbonds (1650–1620 cm�1) when compared to 1550 cm�1 ofN-containing compounds. This result indicates that in addition toVFA build-up, there was also an accumulation of organic materialtaking place. Difficulties becoming evident during this semi-continuous digestion experiments may keep a relation with thethermal profile obtained from thermal analysis and description ofthermal behaviour described by the IPR model.

4. Conclusions

TGA was used for the substrates characterisation, predictingpossible difficulties in their anaerobic digestion. Under batch con-ditions, the anaerobic co-digestion of maize and blood was suc-cessful but under semi-continuous operation it presentedinhibitory problems. In this last case, the reactor suffered volatilefatty acids (VFAs) build-up when operated at low organic loadingrate (3.1 g VS/L d). Ammonium accumulation was discarded asthe origin of the inhibition since low levels of free ammonia wereobtained during the experimentation. The high VFA levels duringthe digestion were accompanied by accumulation of non-degradedorganic material as demonstrated by SEM and FTIR analysis.

Acknowledgements

This research was made possible through the projects PROBIO-GAS PSE-120000-2007-57 supported by Ministry of Education andFEDER funds. Marta Otero acknowledges support from the SpanishMinistry of Science and Innovation (RYC-2010-05634).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2013.06.129.

References

Alvarez, R., Lidén, G., 2008. Semi-continuous co-digestion of solid slaughterhousewaste, manure and fruit and vegetable waste. Renew. Energ. 33, 726–734.

Amon, T., Amon, B., Kryvoruchko, V., Machmüller, A., Hopfner-Sixt, K., Bodiroza, V.,Hrbek, R., Friedel, J., Pösch, E., Wagentristl, H., Schreine, R.M., Zollitsch, W.,2007a. Methane production through anaerobic digestion of various energycrops grown in sustainable crop rotations. Bioresour. Technol. 98, 3204–3212.

Amon, T., Amon, B., Kryvoruchko, V., Zollitsch, W., Mayer, K., Gruber, L., 2007b.Biogas production from maize and dairy cattle manure-Influence of biomasscomposition on the methane yield. Agric. Ecosyst. Environ. 118, 173–182.

Angelidaki, I., Ahring, B.K., 1993. Thermophilic anaerobic digestion of livestockwaste: the effect of ammonia. Appl. Microbiol. Biotechnol. 38, 560–564.

APHA, 1998. Standard Methods for the Examination of Water and Wastewater,20th ed. American Public Health Association, American Water WorksAssociation and Water Environment Federation, Washington, USA.

Cuetos, M.J., Morán, A., Otero, M., Gómez, X., 2009. Anaerobic co-digestion ofpoultry blood with OFMSW: FTIR and TG-DTG study of process stabilization.Environ. Technol. 30, 571–582.

Cuetos, M.J., Fernández, C., Gómez, X., Morán, A., 2011. Anaerobic co-digestion ofswine manure with energy crop residues. Biotechnol. Bioprocess Eng. 16, 1044–1052.

Damartzis, T., Vamvuka, D., Sfakiotakis, S., Zabaniotou, A., 2011. Thermaldegradation studies and kinetic modeling of cardoon (Cynara cardunculus)pyrolysis using thermogravimetric analysis (TGA). Bioresour. Technol. 102,6230–6238.

Demirel, B., 2009. Laboratory investigations on continuous bio-methanization ofenergy crops as mono-substrate without supplementation. Biomass Bioenergy33, 988–993.

Hejnfelt, A., Angelidaki, I., 2009. Anaerobic digestion of slaughterhouse by-products.Biomass Bioenergy 33, 1046–1054.

520 M.J. Cuetos et al. / Bioresource Technology 144 (2013) 513–520

Karlsson, A., Ejlertsson, J., 2012. Addition of HCl as a means to improve biogasproduction from protein-rich food industry waste. Biochem. Eng. J. 61, 43–48.

Klocke, M., Mähnert, P., Mundt, K., Souidi, K., Linke, B., 2007. Microbial communityanalysis of a biogas-producing completely stirred tank reactor fedcontinuously with fodder beet silage as mono-substrate. Syst. Appl.Microbiol. 30, 139–151.

Lehtomäki, A., Httunen, S., Rintala, J.A., 2007. Laboratory investigations on co-digestion of energy crops and crop residues with cow manure for methaneproduction: effect of crop to manure ratio. Resour. Conserv. Recy. 51, 591–609.

Lobato, A., Cuetos, M.J., Gómez, X., Morán, A., 2010. Improvement of biogasproduction by co-digestion of swine manure and residual glycerine. Biofuels 1,59–68.

MAPA, 1994. Official Methods of Analysis. Ministry of Agriculture, Fisheries andFood, General Direction of Food Politics, vol. III, Madrid.

Marcos, A., Al-Kassir, A., Mohamad, A.A., Cuadros, F., López-Rodríguez, F., 2010.Combustible gas production (methane) and biodegradation of solid and liquidmixtures of meat industry wastes. Appl. Energy 87, 1729–1735.

Meissl, K., Smidt, E., Schwanninger, M., 2007. Prediction of humic acid content andrespiration activity of biogenic waste by means of Fourier transform infrared(FTIR) spectra and partial least squares regression (PLS-R) models. Talanta 72,791–799.

Nges, I.A., Escobar, F., Fu, X., Björnsson, L., 2012. Benefits of supplementing anindustrial waste anaerobic digester with energy crops for increased biogasproduction. Waste Manage. 32, 53–59.

Nopharatana, A., Pullammanappallil, P.C., Clarke, W.P., 2006. Kinetics and dynamic-modelling of batch anaerobic digestion of municipal solid waste in a stirredreactor. Waste Manage. 27, 595–603.

Otero, M., Gomez, X., García, A.I., Morán, A., 2008. Non-isothermalthermogravimetric analysis of the combustion of two different carbonaceousmaterials: coal and sewage sludge. J. Therm. Anal. Calorim. 93, 619–626.

Otero, M., Lobato, A., Cuetos, M., Sánchez, M.E., Gómez, X., 2011. Digestion of cattlemanure: thermogravimetric kinetic analysis for the evaluation of organicmatter conversion. Bioresour. Technol. 102, 3404–3410.

Palmowski, L., Müller, J., 2000. Influence of size reduction of organic waste on theiranaerobic digestion. Water Sci. Technol. 41, 155–162.

Peu, P., Picard, S., Diara, A., Girault, R., Béline, F., Bridoux, G., Dabert, P., 2012.Prediction of hydrogen sulphide production during anaerobic digestion oforganic substrates. Bioresour. Technol. 121, 419–424.

Raposo, F., Banks, C.J., Siegert, I., Heaven, S., Borja, R., 2006. Influence of inoculum tosubstrate ratio on the biochemical methane potential of maize in batch tests.Process Biochem. 41, 1444–1450.

Resch, C., Wörl, A., Waltenberger, R., Braun, R., Kirchmayr, R., 2011. Enhancementoptions for the utilisation of nitrogen rich animal by-products in anaerobicdigestion. Bioresour. Technol. 102, 2503–2510.

Salminen, E., Rintala, J., Lokshina, L.Ya., Vavilin, V.A., 2000. Anaerobic batchdegradation of solid poultry slaughterhouse waste. Water Sci. Technol. 4, 33–41.

Salminen, E., Einola, J., Rintala, J., 2001. Characterisation and anaerobic batchdegradation of materials accumulating in anaerobic digesters treating poultryslaughterhouse waste. Environ. Technol. 22, 577–585.

Sonobe, T., Worasuwannarak, N., 2008. Kinetic analyses of biomass pyrolysis usingthe distributed activation energy model. Fuel 87, 414–421.

Sørum, I., Grønli, M.G., Hustad, J.E., 2001. Pyrolysis characteristics and kinetics ofmunicipal solid wastes. Fuel 80, 1217–1227.

Tintner, J., Smidt, E., Böhm, K., Matiasch, L., 2012. Risk assessment of an old landfillregarding the potential of gaseous emissions – A case study based onbioindication, FT-IR spectroscopy and thermal analysis. Waste Manage. 32,2418–2425.

Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary fiber, neutraldetergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J.Dairy Sci. 74, 3583–3597.

Wang, Z., Banks, C.J., 2003. Evaluation of a two stage anaerobic digester for thetreatment of mixed abattoir wastes. Process Biochem. 38, 1267–1273.

Zhang, Y., Banks, C.J., 2012. Co-digestion of the mechanically recovered organicfraction of municipal solid waste with slaughterhouse wastes. Biochem. Eng. J.68, 129–137.