Bioresource Technology 125 (2012) 226–232

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

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

Biochemical methane potential and anaerobic biodegradabilityof non-herbaceous and herbaceous phytomass in biogas production

Jin M. Triolo ⇑, Lene Pedersen, Haiyan Qu, Sven G. SommerInstitute of Chem. Eng., Biotechnology and Environmental Tech., Faculty of Engineering, University of Southern Denmark, Niels Bohrs Allé 1, DK-5230 Odense M, Denmark

h i g h l i g h t s

" The Biochemical methane potential (BMP) of non-herbaceous phytomass was 159.3–249.5 CH4 NL kg VS�1." Wood cuttings had half the BMP of lawn cuttings but a similar BMP to cow manure." Wild plants showed a clear trend for lower biodegradability than lawn cuttings." Lawn cuttings from gardens proved to be the most suitable substrate." Lignin of 100 g kg VS�1 was the critical biodegradability point of phytomass.

a r t i c l e i n f o

Article history:Received 2 July 2012Received in revised form 16 August 2012Accepted 18 August 2012Available online 31 August 2012

Keywords:LignocelluloseLignificationLawn wastePlant biomassCrystallinity

0960-8524/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.biortech.2012.08.079

⇑ Corresponding author. Tel.: +45 4117 8867; fax: +E-mail address: jmt@kbm.sdu.dk (J.M. Triolo).

a b s t r a c t

The suitability of municipal plant waste for anaerobic digestion was examined using 57 different herba-ceous and non-herbaceous samples. Biochemical methane potential (BMP) and anaerobic biodegradabil-ity were related to the degree of lignification and crystallinity of cellulose. The BMP of herbaceous gardenplants (332.7 CH4 NL kg VS�1) was high, although lower than that of energy crops (400–475 CH4

NL kg VS�1). Herbaceous wild plants from natural grassland contained most lignocelluloses, leading torelatively low BMP (214.0 CH4 NL kg VS�1). Non-herbaceous phytomass had a high degree of lignificationand a high concentration of crystalline cellulose, but due to the content of non-woody parts with a lowconcentration of lignocellulose the BMP was relatively high, 199.9 and 172.0 CH4 NL kg VS�1 for hedgecuttings and woody cuttings, respectively. There were indications that a plant lignin concentration of100 g kg VS�1 is the critical biodegradability point in anaerobic digestion of phytomass.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Biogas production is one of the most socio-economically cost-efficient renewable energy technologies, using organic waste andplant biomass as feedstock (Nielsen et al., 2002). It is also an effi-cient method of reducing greenhouse gas emissions (GHG) (Som-mer et al., 2004; Dhingra et al., 2011). The European Commissionhas set mandatory national targets for renewable energy to ac-count for 20% of energy consumption by 2020 and to reduceGHG emissions (EREC, 2010). In accordance with the EuropeanCommission targets, the Danish government has set a target forincreasing the use of animal slurry as feedstock for biogas produc-tion from the current level (5%) up to 40% by 2020 (Green Growth,2009). However, biogas production using only animal manure isnot economically sustainable and addition of biomass from othersources is needed (Møller et al., 2007; Triolo et al., 2011). Energy

ll rights reserved.

45 6550 7354.

crops have been widely used as co-substrate in Germany and Aus-tria, particularly maize, sunflower, grass and Sudan grass (Amonet al., 2007). In Denmark industrial organic waste is co-digestedwith animal manure, but domestic sources of industrial organicwaste have been exhausted and alternative biomass resourcesneed to be explored (Møller et al., 2007; Raven and Gregersen,2007). One source is imported organic industrial waste, which atpresent produces 0.68 petajoule (PJ) biogas in Denmark (Jørgensen,2009). There is still the potential for 15.9 PJ biogas production fromcrop residues and municipal plant litter, which can significantlycontribute to achievement of the Danish target on non-fossil fuelenergy production. Cultivating energy crops for biogas productionmay not be viable due to production costs (Tilman et al., 2006; Rajuet al., 2011). Moreover, there is a risk of increasing food prices if thearea of agricultural land used for food and feed production is re-duced (Hensgen et al., 2011). Herbaceous phytomass from naturalgrasslands, gardens and parks is cheap and use of this biomass forenergy production would not affect food prices. However, the en-ergy yield of such biomass is variable because the harvested plant

J.M. Triolo et al. / Bioresource Technology 125 (2012) 226–232 227

material is diverse in terms of both plant species and organic com-position. Furthermore, the biogas production potential may be lowbecause the fibre content is high (Bühle et al., 2012).

In Denmark, 557,000 tons of green garden waste are deliveredannually to local municipal recycling centres. This correspondedto 17% of total household waste and 7% of total domestic wastein 2005 (Waste Statistics, 2005). The amount of garden waste in-creased 10-fold (96%) from 1994 to 2005 (Waste Statistics, 2005).Green garden waste consists of both herbaceous and non-herba-ceous phytomass such as lawn grass, lawn weeds and residuesfrom hedge and tree trimming, and is currently recycled after beingcomposted and not used for biogas production (Hensgen et al.,2011). The reason could be that biogas plants are reluctant to useorganic waste with a high content of lignocellulose, which is noteasily transformed to biogas (Menon and Rao, 2012). However,since they are facing ‘‘aggressive biogas growth’’ with no availableco-substrates, biogas producers are becoming increasingly inter-ested in using herbaceous garden waste as a cheap feedstock. Toour knowledge, no previous study has examined the biochemicalcharacteristics of a wide range of green garden waste in terms ofmethane potential, anaerobic biodegradability and crystallinity ofcellulose in order to assess the quality of this biomass as a feed-stock for biogas production. Hence, in this study we investigatedthe biogas production potential of different forms of municipalherbaceous phytomass, as well as non-herbaceous material, i.e.green garden waste, energy crops, grass from natural grasslandand woody waste. Biochemical methane potential (BMP) wasdetermined and a distributional analysis was made of each organiccomponent using data obtained from physicochemical analyses.These included Van Soest characterisation and determination ofthe crystallinity index of cellulose in different lignocellulosic mate-rials using X-ray powder diffraction (XRPD). Anaerobic biodegrad-ability was assessed using the concentration of lignocellulosic fibrein each biomass group and the BMP was correlated to the anaero-bic biodegradability.

2. Methods

2.1. Biomass samples collected

In total, 57 samples were collected from a wide variety ofsources in Funen, Denmark, during the period July–October 2011(Table 1).

The samples obtained were separated into two groups, herba-ceous plants that do not have a persistent woody stem and non-herbaceous material such as hedge and tree trimmings. The mate-rial was then subdivided into five groups: (1) herbaceous lawnwaste from private gardens or public parks (‘lawn cuttings’); (2)hedge trimmings from the sites where lawn waste was collected(‘hedge cuttings’); (3) tree trimmings from the sites where lawnwaste was collected and willow from wetland (‘wood cuttings’);(4) herbaceous plants from natural grassland (‘wild plants’); and(5) crops/crop residues from agricultural land (‘crops’).

Most of the lawn waste consisted of lawn grass clippings andlawn weeds, e.g. smooth meadow-grass, clover, short bluegrass,etc. The hedge clippings were from common species such asoval-leaved privet and beech hedge and also contained hedgeweeds. The tree trimmings were from birch and coniferous trees(Lawson’s cypress), as well plane trees from the roadside, and thewillows from wetland included two dominant species in Denmark,namely weeping willow and sharpleaf willow. The hedge and woodcutting samples consisted of leaves, fruit bodies and branchesaround 20 cm in length. The wild plant samples from natural grass-land were obtained from green areas beside roads or from wetland,for example, green areas beside streams and lakes, etc., and most

were either perennial grasses or perennial flowering plants. Thecrop samples consisted of maize, sugar beet and wheat straw.

Apart from grass samples, which consisted of relatively homo-geneous leaves, all the other plant samples were heterogeneousand consisted of a distinct plant body such as stem, leaves, fruitbodies and occasionally branches. Hence, to improve analyticalprecision the samples were well homogenised by milling to a max-imum size of 1 mm after drying at 60 �C prior to analysis. A controlBMP test carried out on eight randomly chosen samples showedthat the drying/milling process did not affect BMP and other biogasproduction characteristics (p > 0.05). On the other hand, from therepeatability test, the lower relative standard deviation was foundin the dried/milled 2.61(±1.61)%, while that in fresh samples wasmuch higher at 9.18(±8.19)%, showing that milling could improvethe precision in BMP tests.

2.2. BMP assay

The BMP of samples was determined according to VDI 4630(2006) using 1.0-L (working volume) batch infusion digesters.Inoculum was obtained from the Fangel biogas plant, which oper-ates at mesophilic conditions (37 �C), and is fed a mixture of 80%animal slurry and about 20% organic industrial waste. The inocu-lum was degassed for 2 weeks at 37 �C. The average pH of the inoc-ulum was 8.1 and the volatile solids (VS) concentration was 67.3%of dry matter. The average methane (CH4) concentration in biogasreleased from the inoculum was 61.7%. When fermenting the plantresidues, the inoculum:substrate ratio was set to 3:1 on a total sol-ids (TS) basis. Following the recommendation in the standard pro-tocol (ISO, 1995; VDI, 2006), 100 mL of anaerobic buffer solutionwith medium was added to the inoculum substrate mixture. Eachreactor was flushed with nitrogen gas to ensure an anaerobicatmosphere.

All assays were performed in triplicate. Fermentation was car-ried out under mesophilic conditions, at 37 �C. Digestion was ter-minated when daily biogas production per batch was less than1% of cumulative gas production according to VDI 4630 (VDI,2006), which corresponded to batch fermentation for approxi-mately 60 days. A couple of times during a working day, the reac-tors were thoroughly mixed by hand shaking to avoid dry layersand to encourage degassing. The gas volume measured was cor-rected to a dry gas basis by excluding the water vapour contentin wet biogas. Pressure and temperature for a norm litre (NL) ofgas were corrected into standard temperature and pressure (STP)conditions (273 K, 1.013 hPa), according to VDI 4630 (2006). TheCH4 concentration in the biogas was determined by a gas chro-matograph (HP 6890 series), equipped with a thermal conductivitydetector and a 30 mm � 0.320 mm column (J&W 113-4332). Thecarrier gas was helium (30 cm/s), and injection volume was0.4 mL. Injector temperature was 110 �C, and detector and oventemperature was 250 �C.

Biomethane was quantified assuming that the dry biogas wascomposed of CO2 + CH4 alone. Consequently CH4 production vol-ume was calculated according to VDI 4630 (2006) by multiplyingthe dry gas production by the CH4/CO2 + CH4 ratio.

The BMP of cellulose (Avicel PH-101 cellulose) was determinedas a control and inoculum as a blank. The BMP of cellulose was393.3(±4.1) CH4 NL (kg VS)�1 and the ratio of BMP to theoreticalBMP (TBMP) was 94.8%. The TBMP of cellulose is 415 CH4 NL(kg VS)�1.

2.3. Physicochemical analysis

Dry matter (TS), volatile solids (VS), crude lipid (XL), totalammoniacal nitrogen (TAN = NH3 + NH4

+) and total Kjeldahl nitro-gen (TKN) were determined according to standard procedures

Table 1Phytomass samples included in the study, with cultivar and harvest date; L, lawn cuttings; H, hedge cuttings; W, wood cuttings; WP, wild plants; C, crops.

No. Type Material Latin name Harvest No. Type Material Latin name Harvest

1 L Meadow grass Poa pratensis July 29 W Weeping willow Salix babylonica October2 L Meadow grass Poa pratensis July 30 W Weeping willow Salix babylonica October3 L Meadow grass Poa pratensis October 31 W Sharpleaf willow Salix acutifolia October4 L Meadow grass Poa pratensis August 32 W Sharpleaf willow Salix acutifolia October5 L Grass mixture – July 33 W Sharpleaf willow Salix acutifolia October6 L Grass mixture – August 34 W Cypress Chamaecypari lawsoniana October7 L White clover Trifolium repens October 35 WP Northern bluegrass Poa alpina August8 L Short bluegrass Poa abbreviata July 36 WP Green foxtail Setaria viridis July9 L Short bluegrass Poa abbreviata July 37 WP Bamboo Bambuseae October10 H Oval-leaved privet Ligustrum ovalifolium July 38 WP Bamboo Bambuseae October11 H Oval-leaved privet Ligustrum ovalifolium October 39 WP Common reed Phragmites October12 H Oval-leaved privet Ligustrum ovalifolium August 40 WP Common reed Phragmites October13 H Ivy Hedera canariensis August 41 WP Tufted hair-grass Deschampsia cespitosa October14 H Beech hedge Fagus sylvatica October 42 WP Tufted hair-grass Deschampsia cespitosa October15 H Black chokeberry Aronia melanocarpa October 43 WP Reed canary grass Phalaris arundinacea July16 H Red chokeberry Aronia arbutifolia October 44 WP Reed canary grass Phalaris arundinacea October17 H Ground-elder Aegopodium podagraria July 45 WP Reed canary grass Phalaris arundinacea October18 H Ground-elder Aegopodium podagraria July 46 WP Reed canary grass Phalaris arundinacea August19 W Birch tree Betula October 47 WP Reed canary grass Phalaris arundinacea August20 W Birch tree Betula October 48 WP Chrysanthemum Helianthus salicifolius July21 W Birch tree Betula October 49 WP Chrysanthemum Matricaria chamomilla July22 W Plane tree, fallen Platanus August 50 WP Dandelion Taraxacum July23 W Plane tree, fallen Platanus October 51 WP Dandelion Taraxacum platycarpum August24 W Plane tree (green) Platanus October 52 C Maize (grain) Zea mays July25 W Plane tree (green) Platanus October 53 C Maize (grain) Zea mays October26 W Weeping willow Salix babylonica October 54 C Maize (leaves) Zea mays July27 W Weeping willow Salix babylonica October 55 C Maize (leaves) Zea mays October28 W Weeping willow Salix babylonica October 56 C Wheat straw Triticum aestivum October

57 C Sugar beet Beta vulgaris October

228 J.M. Triolo et al. / Bioresource Technology 125 (2012) 226–232

(APHA, 2005). Neutral detergent fibre (NDF) was determinedaccording to Mertens et al. (2002). Acid detergent fibre (ADF) andacid detergent lignin (ADL) were determined according to ISO stan-dards (ISO 13906:2009). Crude protein (XP) was determined bymultiplying the difference between TKN and TAN by 6.25. Hemi-celluloses, cellulose and lignin were determined in accordancewith Van Soest characterisation for fibre analysis (Van Soest,1963; Goering and Van Soest, 1970) using the results of NDF,ADF and ADL analyses. Hemicellulose was determined as the dif-ference between NDF and ADF and cellulose by subtracting ADLfrom ADF.

2.4. XRPD analysis

The crystallinity of grass and woody biomass was comparedusing the crystallinity index (CI) determined by XRPD (SiemensD5000 Crystalloflex diffractometer). The X-ray unit was operatedat 40 kV and 35 mA. The 2h region scanned was from 5� to 30�,with a step size of 0.02�. The CI of cellulose was calculated accord-ing to Park et al. (2010) from the height ratio of the intensity of thecrystalline peak to the total intensity after subtraction of the back-ground signal in the amorphous phase (Eq. (1))

CI ð%Þ ¼ 100 � ½ðI002 � IamorphousÞ=I002� ð1Þ

where I002 is the intensity for the crystalline portion of cellulose atabout 2h = 22 and Iamorphous is the peak for the amorphous portion atabout 2h = 18.0.

2.5. Data analysis

In assessment of BMP, a simple stoichiometric calculation of themaximum potential for anaerobic gas production is used to assessthe theoretical or stoichiometric methane potential (Møller et al.,2004; Raposo et al., 2011; Triolo et al., 2011). However, TBMP de-rived from the stoichiometric method is larger than actual meth-ane potential, because recalcitrant carbon affects the latter

(Mauseth, 1988; Triolo et al., 2011). Lignin is non-degradable andthe content of lignin suppresses degradation of lignocellulosic fi-bres such as hemicellulose and celluloses (Triolo et al., 2011).Hence, TBMP is not used as a gauge of biogas production potentialbut as a yardstick to assess the biodegradability of substrate by cal-culating the ratio between BMP and TBMP, as in this study. Herethe transformation of VS for bacterial growth was not taken intoaccount, based on the assumption that microbial cellular yield isnegligible. The theoretical CH4 production of each major compo-nent in the VS fraction was calculated using the stoichiometricequation (Symons and Buswell, 1933). TBMP (CH4 NL kg VS�1)and anaerobic biodegradability were calculated using Eqs. (2)and (3), respectively, including lipid, protein, carbohydrate and lig-nin as g kg VS�1 in the calculations:

TBMP ¼ ðC57H104O6 � 1014þ C5H7O2N � 496þ C6H10O5 � 415

þ C10H13O3 � 727Þ � 0:001 ð2Þ

where TBMP is expressed in CH4 NL kg VS�1 and C57H104O6,C5H7O2N, C6H10O5 and C10H13O3 is the empirical formula for lipid,protein, carbohydrate and lignin, respectively, expressed asg kg VS�1 (Triolo et al. (2011).

Anaearobic biodegradability ¼ ðBMP=TBMPÞ � 100% ð3Þ

All statistical analyses were performed using the SAS softwarepackage (SAS Institute, 1992).

3. Results and discussion

3.1. Specific methane potential

The mean concentration of CH4 in biogas from all samples was56.6%, with very low variations between samples (standard devia-tion: 2.4%). Among the sample groups, CH4 concentration in biogasfrom lawn cuttings and crops was highest, 58.3(±1.1)% and60.1(±3.7)%, respectively. As expected, crop samples showed the

J.M. Triolo et al. / Bioresource Technology 125 (2012) 226–232 229

greatest methane production potential, approximately 5-fold great-er than samples with the lowest methane production potential(Fig. 1 and Table 2).

After crop samples such as maize and sugar beet, lawn cuttingsgave the highest BMP of the biomasses tested, 288.7(±3.5) to388.9(±10.2) CH4 NL kg VS�1, which is at a similar level to theBMP of pig manure (Møller et al., 2004; Triolo et al., 2011).

The BMP of the wild plants tested ranged from 104.0(±12.9) to302.3(±6.1) CH4 NL kg VS�1 (Fig. 1), and varied more than the BMPof the other biomass groups. The high variation in BMP and thelarge and variable amount of lignocelluloses in wild plants con-firmed findings by Bühle et al. (2012), who demonstrated the prob-lems of using grassland plants as feedstock for biogas production. Amost interesting result in the present study is that some of the wildplant samples had lower BMP than the wood cuttings.

Non-herbaceous plant samples produced less methane than theother groups but the production was still significant, ranging from146.6(±12.0) to 249.5(±6.3) CH4 NL kg VS�1 for hedge cuttings and142.5(±15.3) to 239.9(±23.1) CH4 NL kg VS�1 for wood cuttings. Interms of the different wood cuttings, BMP decreased in the order:birch > willow > plane tree. However, the differences in BMP betweenthe different wood wastes were small compared with the variationwithin a particular wood type. Thus the standard deviation of themean BMP obtained for birch cuttings was ±22.2, for willow cuttings±4.7 and for plane tree cuttings ±31.6. Wood cuttings had approxi-mately half the BMP of lawn cuttings but a similar BMP to cow manure(148–223 CH4 NL kg VS�1; Møller et al., 2004; Triolo et al., 2011).

The BMP of samples was influenced by time of harvesting, withsummer (July–August) lawn cuttings having higher BMP than au-tumn cuttings (352.5(±27.8, n = 4) and 301.2(±102, n = 2) CH4 -NL kg VS�1, respectively). The wild plant material also showed aseasonal variation in BMP, with plants cut in October having lowvalues and plants cut in July having higher BMP. The results werein accordance with the previous studies that reported seasonal de-cline of biogas production. The reason for decline is probably dueto the advancing vegetation stage (Prochnow et al., 2009). Themeans of using the plant materials may cause significant seasonalfluctuation not only in energy efficiency, but also instability of theanaerobic environment in biogas reactors. Hence, a suitable rota-tion of feedstock against seasonal variations is necessary for opti-mal biogas operation.

3.2. Fractional lignocellulose and degree of lignification

A major obstacle to using phytomass in biogas production is ahigh degree of lignification, which is a particular problem when

Fig. 1. Profile of measured BMP (CH4 NL (kg VS)�1) for the different types of biomass testriplicate values.

using stem wood. Lignification is a physiological process influ-enced by genetic and environmental factors. The degree of lignifi-cation reflects the extent of lignin deposition in the plant cell well,which can be assessed either by the concentration of lignin in lig-nocellulose or the ratio of lignin to cellulose.

Lignocelluloses were the most abundant organic compound inthe samples tested here (Table 2 and Fig. 2). However, the concen-tration of lignocelluloses varied widely within the different plantgroups tested. Lawn cuttings had a low lignocellulose content(50.7(±10.4)% of VS), owing to a larger proportion of non-cell wallcomponents, and the concentration of lignin in lignocellulose wasmostly very low (4.4(±2.4)% of VS), which is in good agreementwith the BMP of this biomass. Wild plants had the highest contentof lignocellulose (72.2(±3.9)% of VS) and their lignin concentration(15.2(±4.7)% of VS) was considerably higher than that of lawn cut-tings. These differences are partly due to the different biochemicalcharacteristics of the dominant plant community in natural grass-land, but also to their greater age than lawn waste. Frequent lawnmowing shortens physiological vegetation age, which reduces thedegree of lignification. The degree of lignification (lignin/lignocel-lulose) of herbaceous plant samples was low, 8.2% for lawn cut-tings and 15.8% for wild plants.

The lignocellulose concentration in non-herbaceous sampleswas low, 53.0(±12.7)% of VS for hedge cuttings and 50.6(±4.6)% ofVS for wood cuttings. The low lignocellulose concentration of thenon-herbaceous samples seemed to be due to a large amount offruit bodies in these samples giving high methane potential andlow lignocellulose concentration. In contrast, we found a very highdegree of lignification, with lignin/lignocellulose comprising29.4(±11.2)% of lignocellulose in hedge cuttings and 39.7(±7.5)%of lignocellulose in wood cuttings.

3.3. X-ray diffraction

The access by hydrolytic enzymes to cellulose is affected notonly by the degree of physical and chemical links between ligninand hemicelluloses in the plant cell, but also the crystallinity of cel-lulose. Hence, the low biodegradability of lignocellulosic biomassin an anaerobic environment is linked to lignin and biodegradabil-ity of the cellulose and hemicellulose included in a rigid structureof a lignocellulosic matrix and of crystalline cellulose.

Cellulose is the most abundant fibrous fraction in lignocellulosicbiomass and it is a high molecular weight linear polymer linked byb-1.4-glycosidic bonds. The amorphous and crystalline celluloseare caused by intramolecular and intermolecular hydrogen bonds(Park et al., 2010). The XRPD method used in this study is widely used

ted. NL, norm litre (273 K, 1.013 bar). Error bars represent the standard deviation of

Table 2Organic composition, specific methane potential and anaerobic biodegradability (BD = BMP/TBMP � 100%); NL = norm litre (273 K, 1.013 bar) of the different types of phytomasstested, DM = dry matter.

Waste type Lawn cuttings Hedge cuttings Wood cuttings Wild plantsNumber of samples, n n = 9 n = 9 n = 16 n = 17

TKN (g/kg DM) 22.3(10.1) 16.2(7.1) 15.17(4.11) 12.1(6.0)TAN (g/kg DM) 1.86(1.51) 1.70(0.80) 1.74(1.20) 1.17(0.84)NDF (g/kg DM) 50.7(10.4) 49.4(11.0) 56.9(4.9) 66.7(14.1)ADF (g/kg DM) 30.1(4.8) 39.2(10.7) 44.9(4.5) 43.8(8.1)ADL (g/kg DM) 4.15(2.20) 13.9(4.4) 22.5(4.0) 10.08(5.2)XL (g/kg DM) 7.38(2.94) 4.39(3.50) 3.70(3.10) 3.09(1.47)TBMP (CH4 NL kg VS�1) 501.8(27.7) 500.0(24.6) 527.3(26.6) 483.7(21.4)BMP (CH4 NL kg VS�1) 332.7(38.4) 199.9(40.8) 172.0(28.9) 214.0(58.1)BD (%) 66.6(8.5) 39.9(7.6) 32.7(5.2) 44.9(12.5)CH4 in biogas (%) 58.3(1.1) 54.7(2.0) 55.6(1.6) 56.4(1.2)

Fig. 2. (a) Mean organic composition of VS, (b) mean fraction of hemicellulose, cellulose and lignin in lignocellulose; and (c) mean lignocellulose concentration in lawncuttings, hedge cuttings, wood cuttings, wild plants and crop samples. XL, crude lipid; XP, crude protein. Error bars represent the standard deviation within the sample group.

Table 3Crystallinity index (CI) of selected lignocellulosic samples, determined by X-raydiffraction.

Sample Willow Birch Plane tree Lawn Reed canary grass

CI 43.0(4.2) 58.8(1.0) 52.0(2.8) 32.6(9.8) 44.4(5.3)

230 J.M. Triolo et al. / Bioresource Technology 125 (2012) 226–232

to compare crystallinity of cellulose by determining the relative dif-ferences between samples, but cannot be used to estimate theamount of crystalline and amorphous cellulose (Park et al., 2010).

Here, 10 samples were randomly chosen and the crystallinity ofcellulose measured using XRPD. The highest peaks for all the sam-ples were found at 2h = 22� and the plane tree and willow samplesshow clearer and stronger peaks than grass samples, comparedwith the amorphous phase at around 2h = 18�. The cellulosecrystalline peak of lawn grass was lowest of all samples (Table 3).

The cellulose crystallinity index measurements (Table 3)showed that the low biodegradability of non-herbaceous

phytomass is caused not only by a high lignin concentration, butalso by a low content of amorphous cellulose and a higher amountof crystalline cellulose.

Fig. 4. Biochemical methane potential (BMP) as a function of lignin concentrationin VS (above) and as a function of lignin in lignocellulose (below): A, correlationwhen only data on crops and lawn cuttings were used; B, correlation when data onall the biomass groups were used.

J.M. Triolo et al. / Bioresource Technology 125 (2012) 226–232 231

3.4. Anaerobic biodegradability

Anaerobic biodegradability of the samples was calculated as theratio between BMP and TBMP. It was possible to subdivide the phy-tomass into groups with high and low anaerobic biodegradabilityas related to lignin content and to the crystallinity of the cellulosein the biomass. Lawn cuttings (66.6(±8.5)%) and crops(77.8(±26.3)%) had a high anaerobic biodegradability, while hedgecuttings, wood cuttings and wild plants had a low anaerobic biode-gradability, 39.9(±7.6)%, 32.7(±5.2)% and 44.9(±12.5)%,respectively.

The influence of lignocellulose and lignin on biodegradabilitywas examined by calculating the theoretical BMP of degradablefractions not containing lignin (TBMPDE) (Eq. (4)) and the TBMPof non-lignocellulose, defined as the theoretical BMP of neutraldetergent soluble fractions (TBMPNDS) (Eqs. (4) and (5))

TBMPDE ¼ TBMP� TBMPLIGNIN ð4Þ

where TBMPDE is the TBMP of degradable fraction (CH4 NL kg VS�1)and TBMPLIGNIN is the TBMP of the lignin fraction (CH4 NL kg VS�1)

TBMPNDS ¼ TBMP� TBMPNDF ð5Þ

where TBMPNDS is the TBMP of the neutral detergent soluble com-ponents (CH4 NL kg VS�1), which corresponds to the TBMP of non-lignocellulose, and TBMPNDF is the TBMP of neutral detergent fibre(CH4 NL kg VS�1), which correspond to the TBMP of lignocellulose.

Stoichiometric calculations of theoretical BMP using Eqs. (4)and (5) showed that the TBMP of most samples tested was around500 CH4 NL kg VS�1 and the variation within the tested sampleswas very low (Fig. 3).

TBMPDE was much closer to measured BMP than TBMP, but stillconsiderably higher than both. TBMPNDS of wood and hedge cut-tings and wild plants was very similar to measured BMP, whilstthat of lawn cuttings and crops was much lower than the measuredvalues. Thus bioconversion to methane of organic components inhedge, tree and wild plant material almost corresponded to theamount of methane produced from non-lignocellulosic compo-nents in these plant groups. In contrast, BMP was much higherthan the BMPNDS of lawn waste and energy crops, which could indi-cate that more lignocelluloses were transformed to methane due toa much lower degree of lignification and probably also due to low-er crystallinity of cellulose.

Measured BMP is presented in Fig. 4 as a function of lignin con-centration in VS and lignin in lignocellulose. As can be seen fromthe diagram, the correlation between lignin concentration in VSand BMP was better (R2 = 0.761) when only data for crops and lawncutting were used rather than all the biomass groups were in-cluded (R2 = 0.643). When BMP was plotted against degree of ligni-

Fig. 3. TBMP, TBMPDE, BMP and TBMPNDS of lawn cuttings, hedge cuttings, woodcuttings, wild plants and crops.

fication, the R2 value decreased slightly to 0.720 (p < 0.05) withdata for crops and lawn cuttings only, while that for all the biomassgroups was much lower. Furthermore, there was no significant cor-relation between lignin and BMP from hedge cuttings, wood cut-tings or wild plants.

The BMP of crops and lawn cuttings was significantly affectedby the concentration of lignin (Fig. 4). However, the BMP of hedgecuttings, wood cuttings and wild plants was more strongly relatedto the concentration of lignocellulose (Fig. 3), a plant componentwith a lignin concentration above 100 g kg VS�1 (Fig. 2). Therewere indications that a plant lignin concentration of 100 g kg VS�1

is the critical biodegradability point in anaerobic digestion ofphytomass.

4. Conclusions

Biochemical methane potential was low for phytomass with alignin concentration above 100 g kg VS�1 and high for phytomasswith lower lignin concentrations. Despite its high degree of lignifi-cation and a larger amount of crystalline cellulose, the BMP of non-herbaceous phytomass was 159.3–249.5 CH4 NL kg VS�1. The BMPof wild plant material displayed the largest variability betweensamples, but showed a clear trend for lower biodegradability andlower BMP than lawn cuttings. Lawn cuttings proved to be themost suitable substrate owing to their considerably higher BMP,

232 J.M. Triolo et al. / Bioresource Technology 125 (2012) 226–232

low lignin and lignocellulose content and low crystallinity ofcellulose.

Acknowledgement

This study was supported by a grant from Energi FynsUdviklingsfond.

References

Amon, T., Amon, B., Kryvoruchko, V., Zollitsch, W., Mayer, K., Gruber, L., 2007. Biogasproduction from maize and dairy cattle manure: influence of biomasscomposition on the methane yield. Bioresour. Technol. 100, 5777–5782.

APHA, 2005. Standard Methods for the Examination of Water and Wastewater, 21sted. American Public Health Association, Washington, DC.

Bühle, L., Hensgen, F., Donnison, I., Heinsoo, K., Wachendorf, M., 2012. Life cycleassessment of the integrated generation of solid fuel and biogas from biomass(IFBB) in comparison to different energy recovery, animal-based and non-refining management systems. Bioresour. Technol. 111, 230–239.

Dhingra, R., Christensen, E.R., Liu, Y., Zhong, B., Wu, C.F., Yost, M.G., Remais, J.V.,2011. Greenhouse gas emission reductions from domestic anaerobic digesterslinked with sustainable sanitation in rural China. Environ. Sci. Technol. 45,2345–2352.

EREC, 2010. Renewable Energy in European: Market, Trends, and Technologies,second ed. European Renewable Energy Council.

Goering, H.K., Van Soest, P.J., 1970. Forage Fiber Analysis (Apparatus, Reagents,Procedures and Some Applications). Agriculture Handbook No. 379. AgricultureResearch Service, United States Department of Agriculture, Washington, DC.

Green Growth, 2009. Grøn Vækst. Danish Environmental Ministry, April 2009 (inDanish). Available from: <http://www.mim.dk/NR/rdonlyres/D5E4FC9A-B3AC-4C9A-B819-C42300F23CCA/0/GROENVAEKST_2904rapporten.pdf>.

Hensgen, F., Richter, F., Wachendorf, M., 2011. Integrated generation of solid fueland biogas from green cut material from landscape conservation and privatehouseholds. Bioresour. Technol. 102, 10441–10450.

ISO, 1995. Evaluation of the ‘ultimate’ Anaerobic Biodegradability of OrganicCompounds in Digested Sludge Method by Measurements of the BiogasProduction. International Standard, ISO/DIS 11734.

Jørgensen, P.J., 2009. Biogas-green energy. PlanEnergi and Researcher for a Day,Faculty of Agricultural Sciences, Aarhus University.

Mauseth, J.D., 1988. Plant Anatomy. The Benjamin/Cummings Publishing Company,CA.

Menon, V., Rao, M., 2012. Trends in bioconversion of lignocellulose: biofuels,platform chemicals & biorefinery concept: a review. Prog. Energy Combust. Sci.38, 522–550.

Mertens, D. et al., 2002. Gravimetric determination of amylase-treated neutraldetergent fiber in feeds with refluxing in beakers or crucibles: collaborativestudy. J. AOAC Int. 85 (6), 1217–1240.

Møller, H.B., Sommer, S.G., Ahring, B.K., 2004. Methane productivity of manure,straw and solid fractions of manure. Biomass Bioenerg. 26, 485–495.

Møller, H.B., Nielsen, A.M., Nakakubo, R., Olsen, H.J., 2007. Process performance ofbiogas digesters incorporating pre-separated manure. Livestock Sci. 112, 217–223.

Nielsen, L.H., Hjort-Gregersen, K., Thygesen, P., Christensen, J., 2002.Samfundsøkonomiske analyser af biogasfællesanlæg – med tekniske ogselskabsøkonomiske baggrundsanalyser. Food Economy Institute, Reportnumber 136, Copenhagen, 130p.

Park, S., Baker, J.O., Himmel, M.E., Parilla, P.A., Johnson, D.K., 2010. Cellulosecrystallinity index: measurement techniques and their impact on interpretingcellulase performance. Biotechnol. Biofuels 3, 10.

PN-EN ISO 13906:2009. Animal feeding stuffs. Determination of acid detergent fibre(ADF) and acid detergent lignin (ADL) contents.

Prochnow, A., Heiermann, M., Plöchl, M., Linke, B., Idler, C., Amon, T., Hobbs, P.J.,2009. Bioenergy from permanent grassland – a review: 1. Biogas Bioresour.Technol. 100 (21), 4931–4944.

Raju, C.S., Ward, A.J., Nielsen, L., Møller, H.B., 2011. Comparison of near infraredspectroscopy, neutral detergent fibre assay and in-vitro organic matterbiodegradability assay for rapid determination of the biochemical methanepotential of meadow grasses. Bioresour. Technol. 102 (17), 7835–7839.

Raposo, F., Fernández-Cegrí, V., De la Rubia, M.A., Borja, R., Béline, F., Cavinato, C.,Demirer, G., Fernández, B., Fdz-Polanco, M., Frigon, J.C., Ganesh, R., Kaparaju, P.,Koubova, J., Méndez, R., Menin, G., Peene, A., Scherer, P., Torrijos, M., Uellendahl,H., Wierinck, I., deWilde, V., 2011. Biochemical methane potential (BMP) of solidorganic substrates: evaluation of anaerobic biodegradability using data from aninternational interlaboratory study. J. Chem. Technol. Biotechnol., 11 (Publishedonline in Wiley Online Library).

Raven, R.P.J.M., Gregersen, K.H., 2007. Biogas plants in Denmark: successes andsetbacks. Renew. Sustain. Energy Rev. 11, 116–132.

SAS Institute, 1992. SAS User’s Guide. SAS Institute, Cary, NC.Sommer, S.G., Petersen, S.O., Moller, H.B., 2004. Algorithms for calculating methane

and nitrous oxide emissions from manure management. Nutr. CyclingAgroecosyst. 69, 143–154.

Symons, G.E., Buswell, A.M., 1933. The methane fermentation of carbohydrates. J.Am. Chem. Soc. 55 (5), 2028–2036.

Tilman, D., Hill, J., Lehman, C., 2006. Carbon-negative biofuels from low-input highdiversity grassland biomass. Science 314, 1598–1600.

Triolo, J.M., Sommer, S.G., Møller, H.B., Weisbjerg, M.R., Jiang, X., 2011. A newalgorithm to characterize biodegradability of biomass during anaerobicdigestion: influence of lignin concentration on methane production potential.Bioresour. Technol. 102, 9395–9402.

Van Soest, P.J., 1963. Use of detergents in the analysis of fibrous feeds. II: Arapidmethod for the determination of fiber and lignin. J. Assoc. Offic. Agric.Chem. 46, 829–835.

VDI, 2006. VDI 4630, Fermentation of organic materials. In: Characterisation ofSubstrate, Sampling, Collection of Material Data, Fermentation Tests.

Waste Statistics, 2005. 2007 Danish Environmental Protection Agency, 2005.Environmental Review No. 6, 2007. Available from: <http://www2.mst.dk/Udgiv/publications/2007/978-87-7052-581-7/pdf/978-87-7052-581-7.pdf>.