te

en

Received in revised form

Published online 21 May 2009

by-products from pigs were tested. The methane potential measured by batch assays for

fat, blood, hair, meat, ribs, rawwaste were: 225, 497, 487, 561, 582, 575,

methane potential. Dilution of the by-products had a positive effect on the specificmethane

ignific

for food consumption. During the last 60 years, slaughter-

treatment of wastes and by-products has emerged as a major

concern not only in pork industry but also in meat industry in

general. For instance, outbreak of diseases such as bovine

spongiform encephalopathy (BSE) in cattle and the dangerous

(part of infected animals, international catering etc.) and is

content) cannot be used as feedstock in composting and

biogas plants, unless they have first been rendered to the

133 C, 300 kPa, 20 min EU pressure-rendering standard(sterilization); and finally category 3, low risk material

* Corresponding author. Tel.: +45 45 25 14 29; fax: +45 45 93 28 50.

Avai lab le at www.sc iencedi rect .com

.co

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 4E-mail address: ria@env.dtu.dk (I. Angelidaki).house waste, rich in proteins and lipids, has been treated and

used for production of animal fodder. However, due to legal

restrictions and environmentally conscious consumers, the

not allowed to be treated in composting or biogas plants

under any circumstances; category 2, high risk animal by-

products (diseased animals, manure and digestive tractproduction in Denmark. More than 24 million pigs are

slaughtered annually in Denmark [1]. Due to higher meat

demand at present compared to the past, the quantity of

organic by-products from slaughterhouses has increased.

Approx. 25% of the total animal weigh slaughtered is not used

some animal by-products [2].

According to the legislation slaughterhouse waste must

be treated by different treatments depending on the cate-

gory of the animal byproduct [3]. Three categories of animal

by-products are defined; Category 1 is high risk materialKeywords:

Methane yields

Pasteurization

Sterilization

Inhibition

Pretreatment

Temperature

Co-digestion

EC regulation No. 1774/2002

1. Introduction

Pork production constitutes a s0961-9534/$ see front matter 2009 Publisdoi:10.1016/j.biombioe.2009.03.004yield with the highest dilutions giving the best results. High concentrations of long-chain

fatty acids and ammonia in the by-products were found to inhibit the biogas process at

concentrations higher than 5 g lipids dm3 and 7 gN dm3 respectively. Pretreatment

(pasteurization: 70 C, sterilization: 133 C, and alkali hydrolysis (NaOH) had no effect on

achievedmethaneyields.Mesophilic digestionwasmore stable than thermophilic digestion,

and higher methane yield was noticed at high waste concentrations. The lower yield

at thermophilic temperature and high waste concentration was due to ammonia inhibition.

Co-digestion of 5% pork by-products mixed with pig manure at 37 C showed 40% higher

methane production compared to digestion of manure alone.

2009 Published by Elsevier Ltd.

ant part of meat

human disease CreutzfeldJacob in 2001, has resulted in

increasing awareness of the need for hygiene regulations,

tighter process control, and the prohibition of utilization ofAccepted 17 March 2009 359, 619 dm3 kg1 respectively, corresponding to 50100% of the calculated theoretical11 March 2009 meat- and bone flour,Article history:

Received 13 January 2008

Anaerobic digestion of animal by-productswas investigated in batch and semi-continuously

fed, reactor experiments at 55 C and for some experiments also at 37 C. Separate or mixedAnaerobic digestion of slaugh

Anette Hejnfelt, Irini Angelidaki*

Department of Environmental Engineering, Technical University of D

a r t i c l e i n f o a b s t r a c t

ht tp : / /www.e lsev ierhed by Elsevier Ltd.rhouse by-products

mark, DTU, Building 113, DK-2800 Kgs. Lyngby, Denmark

m/loca te /b iombioe

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 4 1047(catering residues, meat, precooked foods, etc.) approved for

food consumption, must be treated to at least 70 C for 1 h ina closed system [3].

Anaerobic digestion of animal by-products constitutes

a possible method of treating the by-products and at the same

time produce energy in the form ofmethane, and utilization of

the digestion effluents as fertilizer for application on agricul-

tural fields for nutrient recovery, [4]. However, slaughterhouse

wastes are generally regarded as difficult substrates for anaer-

obic digestion, mainly because of their typically high protein

and lipid content [5]. Protein degradation releases ammonia,

which at high concentrations is suggested to be inhibitory for

the anaerobic microorganisms [6,7]. It is generally considered

that the unionized form of ammonia is the cause of inhibition

and concentrations ranging from 0.1 to 1.1 kg-Nm3 arereported as inhibiting concentrations [7]. Additionally, lipids

may also cause problems in anaerobic digestion because of

their tendency to promote floating scum and due to possible

accumulation of inhibiting degradation intermediates such as

long-chain fatty acids (LCFAs) [8,6,9]. The breakdown of LCFAs

can be the rate-limiting step in the degradation of complex

substrates [10] requiring gradual adaptation and careful dosing

of lipid rich waste products to avoid LCFA accumulation.

Already at very low concentrations (such as 0.5 kgm3), LCFAs,especially unsaturated LCFAs, are suggested to be inhibitory to

syntrophic acetogenic and methanogenic bacteria [6,11]. The

relatively high N content, the high total solids (TS) content,

often exclude the possibility of treating animal by-products in

their original undiluted form. Therefore, dilution is typically

necessary or, as amore attractive option, co-digestionwith less

concentrated organic waste types, such as manure or waste

water. In co-digestion the concentratedwaste types can highly

increase the biogas production compared to dilute wastes,

while dilute wastes can provide process stability and serve as

dilution media while also being treated.

Anaerobic digestion of animal by-products reported in the

literature include studies with by-products from poultry, e.g.

blood, meat and bones [4], rumen and cattle blood [5], blood

and category 3material frompigs [12]. In addition, co-digestion

of manure and rumen [13] and blood and rumen from cattle

and pigs have been studied in laboratory andpilot scale under

mesophilic conditions (37 C) andwith pre-treatment (heating)[14]. Methane yields of 0.520.55 m3 kg1 VS were reportedunder mesophilic conditions for solid slaughterhouse waste

in 2 dm3 continuously stirred tank reactor (CSTR) operated at

0.8 kg VS m3 d1 with 50 days HRT [4]. However, information

about methane yields of different parts of animal by-products

is still lacking. It would be expected that thermophilic

temperature would have better sanitation effect compared to

mesophilic, however, no studies on thermophilic digestion of

animal by-products are yet reported.

In the present study, themethane yields of several types of

by-products from a pig slaughterhouse were determined in

batch assays at different concentrations. Levels for inhibition

under thermophilic and mesophilic conditions were investi-

gated. The effect of heating (70 C, 1 h) and sterilization (133 C,300 kPa, 20 min) on themethane yield of mixed pig-waste was

tested. Finally co-digestion of animal by-products fromaslaughterhouseandmanurewas tested in semi-continuously

fed reactors at mesophilic and thermophilic temperatures.2. Materials and methods

2.1. Substrate

Five different types of pig slaughterhouse by-products; fat,

blood, raw waste (meat, fat and bones), intermediate product

(pressed rawwaste) andboneflourwere received fromaDanish

animal waste processing company (Daka, Denmark; Lat: N

5525045.3700 Long: E 1147054.7300). All these by-products werepretreated at the factory before delivered for experimental use.

The pretreatment at the factory includedmaceration to particle

size smaller thanapprox. 34 mmof a largeportion, followedby

homogenization in order to take a representative sample. A

smaller portion was thereafter sampled and sent to DTU for

analysis and experiments. In addition, pig hair, fresh meat and

ribs were obtained from a trainee-slaughterhouse (Roskilde,

Denmark; Lat: N 5537050.6100 Long: E 124043.1600). The pretreat-ment of the different waste products are shown in Table 1.

Finally,mixed porkwaste consisting of all non-commercial

parts of one slaughtered pig, was collected and immediately

delivered for testing from a slaughterhouse. The mixed pork

waste, after removal of bones, was homogenized in a blender

resulting in particle size less than 2 mm (Table 2).

Thermophilically digested manure from a centralized

biogas plant (Lemvig centralized biogas plant, Denmark; Lat: N

563100.9000 Long: E 818047.7900) was used as inoculum for batchand semi-continuous reactor experiments. For themesophilic

experiments inoculum fromamesophilic biogas plant (Nysted

centralized biogas plant, Denmark; Lat: N 5540039.8100 Long: E1212018.0700) was used.

Chemical analysis of the above mentioned materials was

performed immediately upon arrival at the laboratory and the

experiments were initiated. The mixed pork waste used for

long term continuous experiments was frozen down in 2 kg

portions and was thawed before use.

2.2. Pretreatment

The effect of three different pretreatments on characteristics

and methane production potential of the mixed pork waste

was evaluated (Table 1). The pretreatments tested were:

heating (70 C for 1 h); sodium hydroxide (50 or 100 g NaOHkg1 VS) treatments and finally autoclaving (133 C and300 kPa for 20 min). As mentioned the Daka products were

pretreated at Daka.

2.3. Methane yield

Assays for determination of the methane yield of the various

by-products from pigs were carried out in either 0.5 dm3 or

2 dm3 bottles. The assays were carried out at 55 C for alltested by-products except for the mixed pork waste for which

the methane yield assay was performed at both 55 C and37 C in order to elucidate the effect of temperature on theanaerobic digestion of mixed pork waste. Four different

concentrations (5%, 20%, 50%, and 100%) of substrate (weight

basis) were prepared by dilution with water. Substrate, basicanaerobic (BA) synthetic media [15] and inoculum (60% of

active volume) (retrieved 1 or 2 days before the experiments)

with TC detection and chromosorb 108 column (1.1 m 3/16

ncentration % Incubation temp. Inoculum used

5, 20, 50, 100 55 C Lab.-reactor

5, 20, 50, 100 55 C Lab.-reactor

5, 20, 50, 100 55 C Lab.-reactor

5, 20, 50, 100 55 C Lab.-reactor

5, 20, 50, 100 55 C Lab.-reactor

5 and 10 55 C Snertinge biogas plant5, 20, 50, 80 55 C Inoculum from own

reactor experiment5, 20, 50

5 and 10 55 C Snertinge biogas plant

5,

5,

5,

5

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 41048were added in the bottles under constant flushingwithN2/CO2(80%/20%). Finally, the bottles were sealed with butyl rubber

stoppers and aluminium crimps. Bottles containing only

water, BA media and inoculum (60% of active volume) were

used as controls. An overview of all the batch assays is pre-

sented in Table 1.

2.4. Continuously stirred tank reactor experiments

Co-digestion of different mixtures of manure and mixed pork

waste was investigated in three identical semi-continuously

fed CSTRs (referred to as Rt,20, Rt,5 and Rm,5), each with a total

capacity of 5 dm3 and a working volume of 3.2 dm3. R and

Table 1 Batch experiments, overview.

Waste product Pretreatment Co

Raw wastea Autoclaved in lab. at 300 kPa and

133 C for 20 min. (Jeppesen, 2003)Blooda Autoclaved in lab. At 300 kPa and

133 C for 20 min. (Jeppesen, 2003).Inter-mediate producta Heated to 85 C for 20 min.

(Jeppesen, 2003).

Bone floura Sterilized in the process,

133 C for 20 min. (Daka, 2003).Lipida Sterilized in the process at

125 C for one hour (Daka, 2003).Bones Untreated

Hair Untreated

Heated, 70 C, 1Meat (pure meat) 10 mm

50 mm

Mixed pork waste Untreated

Heated, 70 C, 1 hSterilized, 133 CNaOH

a Daka products.t,20

Rt,5 were operated at 55 C while Rm,5 was operated at 37 C.Reactors were fed by peristaltic pump at an interval of 8 h

from a continuously stirred feed bottle. On day 0, the reactors

were inoculated. The first 25 days the hydraulic retention time

(HRT) of the thermophilic reactors was 15 days, but because of

high VFA concentrations HRT was increased to 21 days. The

organic loading rate (OLR) operating on pig manure only was

6 g VS day1. Upon reaching stability as indicated by low dailyvariations in biogas production and VFA content (lower than

10%), co-digestion of manure with mixed pork waste was

initiated (day 75 for the thermophilic reactors and day 38 for

the mesophilic reactor). 0.15 dm3 (mixed pork waste and

manure) were added per day. Ratios of mixed pork waste to

manure tested were 20% waste in Rt,20 (OLR 12.5 g VS day1),5% waste in Rt,5 and Rm,5 (8.3 g VS day

1) respectively. Afterday 63 the composition of the feed in Rm,5 was changed to 2.5%

mixed pork waste in manure, in order to reveal whether

inhibition was present at 5% animal waste additions.

2.5. Analytical methods

pH was measured with Metrohm 744 pH meter immediately

after each sampling. Methane content in the biogas was

analysed using a Gas Chromatography (GC-14A) equippedmolsieve 137) [6]. The biogas produced in CSTRs was

measured through gas meters [6] Volatile fatty acids (VFAs)

were determined using GC5890-series II equipped with flame

ionisation detection (FID) and HP FFAP column, (0.53 mm/

30 m/1.00 mm). Total Kjeldahl nitrogen, lipids (Soxhlet),

carbohydrates and total solids (TS) and volatile solids (VS)

were determined according to Standard Methods [16]. Protein

content was calculated from Kjeldahl nitrogen content

multiplying with a conversion factor of Kjeldahl-N, 6.25

(for meat).00

20, 50, 80 55 C/37 C Thermo.: Lemvig20, 50, 80 Meso.: Nysted biogas plant

20, 50, 802.6. Calculation on theoretical methane yield

The theoretical methane yield at standard temperature and

pressure (0 C, 100 kPa) was calculated using the followingformula: dm3 kg1X 0.496 + Y 1.014 + Z 0.415

Where: X% proteins of VS, Y% lipids of VS, Z%carbohydrates of VS [15].

Table 2 Characteristics of different waste productsobtained from a single pig.

Composition of the mixedpork waste product

% Amount inkg

Blood 10.4 2.90

Hair 7.6 2.12

Hair and skin 0.1 0.03

Intestine content Not known 5.00a

Meat fractions not for use, lipids 26.7 7.45

Bones 16.10 4.49

Head Not known 2.78a

Manure 10.7 2.99

Intestine fat 0.5 0.14

Total 100.0 27.90

a Estimated.

mixed pork waste while the lowest methane yield were

Table

3Compositionoftheby-productsusedin

thebatchtests.

Raw

waste

Int.product

Boneflour

Blood

Fat

Hair

Meat

Bones

Untreatedmix.

pork

waste

Heatedmix.

pork

waste

Sterilizedmix.

pork

waste

Pig

manure

Volatile

solids

%37.6

36.6

72.2

16.8

99.2

35.8

25.6

41.6

23.2

26.1

33.4

4.6

Totalsolids

%45.8

49.1

95.7

17.9

99.4

39.7

28.6

58.9

26.9

28.7

37.9

6.4

Ash

es

%8.1

12.5

23.6

1.9

0.1

3.9

3.0

24.5

3.7

2.6

4.5

1.8

Waterco

ntent

%54.2

50.9

4.3

82.1

0.8

60.3

71.4

41.1

73.1

71.3

62.1

93.9

Density

Kgm

3

1.1

0.5

0.6

1.0

0.7

0.8

0.8

0.9

1.0

Dissolvedammonia

gNkg1

2.0

3.8

6.4

1.7

0.0

10.6

15.6

15.6

a0.4

0.4

0.4

2.52

KjeldahlN

gNkg1

24.2

30.8

108.5

27.1

0.0

58.0

40.9

40.9

a27.8

28.7

34.5

4.9

Organic

NgNkg1

22.3

27.0

102.1

25.4

0.0

47.4

25.2

25.2

a27.4

28.3

34.1

2.4

Lipids

%ofVS

52.3

16.7

11.5

0.3

100.0

a11.1

17.2

17.2

a23.6

23.6

23.6

6.8

Proteins

%ofVS

37.0

46.1

88.4

94.4

0.0

82.7

61.6

61.6

a74.0

74.0

74.0

31.9

Carbohydrates

%ofVS

108

37.3

0.2

5.3

0.0

6.2

21.2

21.2

a2.4

2.4

2.4

61.3

aNotmeasu

redassumedthat100%

ofVSis

lipid

andthatboneshavethesameco

mpositionasthemeat.

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 4 1049obtained for bone flour when incubated at 5% bone-flour

waste concentration. For meat and bones, 10% waste

concentrations resulted in the highest methane yield which

was 580 dm3 kg1 VS, equal to the respective theoretical

methane yields. Blood had a maximum methane yield of

490 dm3 kg1 VS at 5% concentration. All by-products testedhad high gas production per kg waste. Pork fat had the highest

methane yield per kg waste (562 dm3 kg1 waste) while lowestmethane production per kg waste of 81 dm3 kg1 waste wasobtained from blood.

3.2.1. The effect of pretreatmentResults showed that thermal treatment at 70 C andtreatment by addition of 50 or 100 g NaOH kg1 VS NaOHhad no significant effect on the biodegradability and

methane yields of mixed pork waste. Untreated mixed

pork waste had a specific methane yield of 600 dm3 kg1

which corresponded to the theoretical yield, thus

pretreatment was not expected to increase the methane

yield. Sterilization of the mixed pork waste had no effect3. Results

3.1. Characterization of by-products

Characteristics of the pig by-products are presented in

Table 3.

All by-products, except for fat, had higher protein (3794%

of VS) and VS (16.8%99.2%) content than pig manure. In

manure, 32% of the VS was protein and had a VS content of

4.6%. The high protein content of the pig by-products indi-

cates a potential for ammonia inhibition especially under

thermophilic conditions as ammonia-N load was relatively

high already when operated on pig manure alone. Among the

studied by-products, fat had the highest VS content of 99%

followed by bone flour (96%), while blood had the lowest 17%.

Especially the raw waste contained a lot of lipids (52% of the

VS), but also mixed pork waste had high content of lipids (24%

of VS). Pretreatment did not change the characteristic of the

by-products.

3.2. Batch experiments

The methane production rates and yields for the different

by-products incubated at 37 and 55 C with mesophilicand thermophilic inoculum, respectively, are presented in

Fig. 1.

Methane productions in all assays, except for pork fat

which had a lag phase of 20 days, started after an initial lag

phase of 35 days and continued for up to 3035 days.Methane

yields for the studied substrates varied with substrate

concentration, pretreatment and incubation temperature. For

most of the substrates incubation of 5% waste concentration

with 60% inoculum was found optimal with methane yields

ranging between 230 and 620 dm3 kg1 VS. These yields werecomparable to the theoretical yields of 500750 dm3 kg1 VS.Highest methane yield of 620 dm3 kg1 VS was obtained foron the methane yield, but the methane production per kg

waste was improved, as VS content increased due to

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 41050Daka intermediate product750

1000

-1 V

S

Daka raw waste

0

250

500

750

1000

dm

3 C

H4 kg

-1 V

Swater evaporation. Sterilized waste gave a production of

225 dm3 kg1 waste, while untreated, mixed pork wastewhich gave 125 dm3 kg1 waste.

3.2.2. The effect of temperatureThe results showed that there was no difference in methane

yield between thermophilic (55 C) and mesophilic (37 C)conditions when 5% pork waste was used in batch assays.

Theoretical yields (600 dm3 kg1 VS) were reached under both

conditions. However, when 50% pork waste was used, only

mesophilic digestion was possible (Fig. 2).

The theoretical value was reached, while no methane was

produced under thermophilic conditions. Ionisation of

ammonia is a function of temperature and pH. The free

ammonia concentration is increasing significantly with

Daka Blood

0

250

500

750

1000

dm

3 C

H4 kg

-1 V

S

0

250

500

dm

3 C

H4 kg

Daka bone flour

0

250

500

750

1000

0 10 20 30 40Time

dm

3 C

H4 kg

-1 V

S

Fig. 1 Cumulatedmethane production for different pig slaughte

A 5%, 10%,- 20%, :50%, C 80%, B 100% and theoreticalRibs

Meat (50mm pieces)temperature and pH [7]. Although the exact free ammonia

concentration was not estimated in the batches, we can

assume that the free ammonia concentration was higher at

55 C compared to 37 C, as the amount of animal waste andthereby the total ammonia load was equal in both the meso-

philic and thermophilic vials. This could explain the severe

inhibition at thermophilic digestion [7].

3.3. Continuously stirred reactor experiments

3.3.1. The thermophilic reactors (55 C)The process performance of the thermophilic reactors is pre-

sented in Fig. 3 and Table 4.

The two thermophilic CSTRs were operated equally for 43

days, fed only with pig manure. The biogas production during

Pig hair

Mixed pork waste

0 10 20 30 40(Days)

rhouse by-products incubated at 55 8C at different dilutions.

methane yield.

the initial phase (days 043), in both Rt,20 and Rt,5, was very

unstable (Fig. 3). During the initial phase very high VFA

concentrations were observed. Acetate was the pre-dominant

VFAwith values reaching up to 120 mM. Propionatewas found

increasing at the same time (from 10 to 40 mM). The VFA level

in reactor Rt,5 was similar to Rt,20 in this period.

During the next period (days 4375, stable period) the

daily biogas production was approx. 2200 ml per day (69%

methane) with daily variations lower than 0.2 dm3 biogas

Thermophilic vs. meso

0

200

400

600

800

1000

0 20 40Time

dm3

CH

4 kg

-1 V

S

Fig. 2 Cumulated methane production of mixed pork w

3

4

5

6

pro

du

ctio

n (d

m3)

Rt,20

a

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 4 10510

1

2

Bio

gas Rt,5

VFA, Rt,5

6080

100120

140

VF

A (m

M)

AcPropIso-buBut

b0

2040 Iso-val

Val

VFA, Rt,20

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120Time (days)

VF

A (m

M)

Ac

PropIso-but

But

Iso-val

Val

c

Fig. 3 Thermophilic anaerobic co-digestion of manure

with 5% (reactor Rt,20) and 20% (reactor Rt,5) mixed pork

waste; (a) Biogas production from Rt,5 and Rt,20, (b) VFA

concentrations of Rt,5 and (c) VFA concentrations of Rt,20.day1 (Fig. 3), corresponding to a specific methane yield ofapprox. 220 dm3 kg1 VS, which is rather low for pig

manure. At day 75 addition of mixed pork waste to manure

feed was initiated. Mixed pork waste was applied to the

manure in Rt,20 and Rt,5 in concentrations of 20% and 5%

respectively. After 1 day the gas production in reactor Rt,20stopped completely. A lot of foam was observed. In reactor

Rt,5 the gas production decreased to 0.73 dm3 day1 after 12

days (Fig. 3). In this period the acetate concentration

decreased, while the propionate concentration increased

slightly to reach 4045 mM (Fig. 3).

3.3.2. The mesophilic reactor (37 C)The process performance of the mesophilic reactor Rm,5 is

presented in Fig. 4 and Table 4.

Shortly after start up with pig manure, the biogas

production started. After an initial phase of 27 days the

daily gas production was stable and achieved

3.3 dm3 day1 (75% methane), corresponding to a specificmethane yield of approx. 350 dm3 kg1 VS which is some-what high for pig manure. After 10 days with stable

conditions co-digestion was started. Mixed pork waste was

added to the manure in a concentration of 5%. Within 5

days, gas production increased from 3.3 dm3 day1 to

5.5 dm3 day1, corresponding to an overall specific yield of489 dm3 kg1 VS or approx. 900 dm3 kg1 VS if extra

60 80 100

(days)

aste (50% diluted) incubated at C 37 and - 55 8C.philic for 50% of mixed pork wasteproduction is ascribed to pork waste VS only. The yield of

dm3 kg1 VS for pork waste was higher than determinedby methane potential batch assays, which indicates that

the higher process activity also had positive effect on the

methane production from manure. After 20 days of stable

process conditions the concentration of mixed pork waste

in feed was reduced from 5% to 2.5%. This resulted in

a decrease in biogas production from 5.5 dm3 day1 to4.5 dm3 day1. The corresponding decrease in yield was

from 489 to 417 dm3 kg1 VS.

4. Discussion

The results from the present study showed that slaughter-

house waste in general had a good methane potential

and could produce 225619 dm3 kg1, which corresponds to

Table 4 Results from the continuous reactor experiments. Meastable period and co-digestion with mixed pork waste. The va

Reactor Rt,20

Temperature 55 C% Mixed porker waste 20 %

HRT (days) 21

Stable period with manure 6 g VS d1

Methane yield (dm3 kg1) 219.0Methane (%) 69.0

% Of theoretical potential (%) 45.0

pH 8.05

Dissolved ammonia (g N dm3) 3.28Free ammonia (g N dm3) 0.99Degradation of VS % 34.0

Total VFA (mM) 111.0

Slaughterhouse waste and manure 12.5 g VS d1

Methane yield (dm3 kg1) 0.0Methane (%) 0.0

pH 8.23

Dissolved ammonia (g N dm3) 3.37Free ammonia (g N dm3) 1.33Degradation of VS %

Total VFA (mM) 97.0

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 4105250100% of the theoretical yields. Carpentier et al. [12],

investigated thermophilic batch digestion of pig slaughter-

house waste, category 3. They measured a biogas yield of

1.67 m3kg-1 VS (corresponding to 1085 dm3 kg1 assuming 65%CH4 in the biogas). This value is very high and is approx. cor-

responding to the theoretical maximum yield of lipid, indi-

cating that nearly all VS in the tested waste were lipid.

However, it is not possible to verify this assumption, as the

exact composition of thematerial they used was not reported.

Edstrom et al. [14] carried out mesophilic fed-batch digestionwith mixed, minced animal by-products. The specific

production of biogas was 760 dm3 kg1 VS (490 dm3 kg1 VS)

01

234

56

Bio

gas p

ro

du

ctio

n (d

m3)

a

0

5

10

15

20

25

30

0 10 20 30 40 50 60 70Time (days)

VF

A (m

M)

Ac

prop

Iso-but

But

Iso-val

Val

b

Fig. 4 Mesophilic anaerobic co-digestion of manure with

5% (days 3862) and 2.5% (days 6372) mixed pork waste

(Rm,5); (a) Biogas production and (b) VFA concentrations.for untreated waste, which is lower than the 619 ml dm3 kg1

VS for mixed pork waste in batch assays found in this study.

In most types of by-products the highest specific methane

yields were achieved when the animal by-products were at

the highest dilutions (5%). This indicates that components of

the animal by-products constitute potential inhibitors for the

biogas process. Even 5% of raw waste, bone flour and fat

resulted in methane yields, which were only about 50% of the

theoretical yield. It appears from Table 3 that these substrates

had the highest content of VS, lipids and protein. In addition

sured and calculated parameters for the reactors during thelues are mean values from each period.

Rt,5 Rm,5

55 C 37 C5 % 5 %

21 21

6 g VS d1 6 g VS d1

219.0 357.0

69.0 75.0

45.0 74.0

8.04 7.98

3.30 3.70

0.98 0.40

30.0 47.0

117.0 7.0

8.3 g VS d1 8.3 g VS d1

61.0 489.0

71.0 74.0

8.10 7.97

3.32 3.84

1.08 0.40

40.0 50.0

128.0 17.0animal by-products, due to their high protein and lipid

content, are prone to cause inhibition unless they are diluted.

From the correlation between process quality, as indicated by

the relative methane yield obtained compared to theoretical

yield and VFA levels at end of batch digestion, and the corre-

sponding N-concentration (Fig. 5), it can be seen that for total-

N concentrations higher than 7 gN kg1 the process wasseverely inhibited (Fig. 5). The results are in accordance with

previous studies [17].

It is generally believed that higher temperatures result in

higher bacterial growth rate and metabolic activities [18,19].

However, in the present study it was shown that thermo-

philic digestion resulted in lower yields and a more stressed

process. Mainly ammonia were assumed to be the reason

for total process breakdown in the thermophilic batch

assays and reactor (Rt,5 and Rt,20). The VFA level in the

mesophilic CSTR reactor was lower (around 15 mM)

compared to the thermophilic CSTR reactors (around

45 mM) from the beginning of the experiments (Figs. 3 and

4), indicating that the thermophilic reactors were much

more stressed also before mixed pork waste was added to

the manure. This is also what Hansen et al. [7] concluded

when he investigated ammonia inhibition in manure at

different temperatures. Angelidaki and Ahring [17], showed

that mesophilic temperature is a better choice for processes

operating with high N loading because of a slightly lower

ammonia and lipid concentration. The VFA level is lower and

higher amounts of animal by-products could be added

0 5 10 15 20 25

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 4 1053pH and lower portion of the total ammonia-N in the free

ammonia (NH3) form.

Concerning inhibition, the same tendency was seen for

lipids as for ammonia in batch assays. When the initial

concentration of lipids reached approx. 4 g kg1 the methaneyield decreased (Fig. 6).

0

20

40

60

80

100

120

% o

f th

eo

retical C

H4 yield

0

200

400

600

800

1000

0 5 10 15 20 25 30Total-N (g kg-1)

To

tal V

FA

(m

M)

Fig. 5 Relation between the total-N (g kgL1), methane

yield (% of theoretical potential) and total VFA

concentration (mM) at 55 8C. Data are from anaerobic batch

assays of different animal by-products.The VFA end concentration increased when the initial

concentration of lipids reached 5 g kg1 (Fig. 6). The inhibi-tion is attributed to accumulation of high concentrations of

accumulated LCFA intermediates [9,10]. A balanced hydro-

lysis to LCFA and subsequent degradation of LCFA is

required in order to avoid accumulation of LCFA, which may

be obtained in continuous operated systems with careful

introduction/adaptation of lipid rich waste products. Degra-

dation of LCFA has been reported as the rate-limiting step of

lipids [9]. In the batch and CSTR experiments formation of

foam was observed and complicated the biogas process,

especially in the CSTRs. The foam was probably coming from

degradation of LCFA and resulted in reactor or the bottles

overflowing and decreased the bio-accessibility of LCFA and

other particles [4].

Pretreatment had no effect primarily because the by-

products were already easily degradable. This was indicated

by the short lag phase and the high gas yield close to theo-

retical yield for untreated mixed pork waste. This is in

agreement with previously reported in Ref. [4]. Opposite

results were reported by Edstrom et al., [14] who found that

pasteurization (70 C, 1 h) of mixed animal by-productincreased the biogas yield from 760 to 1140 dm3 kg1 VS(490740 dm3 kg1 VS). The total solids content was higher inthe pretreated waste probably due to evaporation of water

during heating and sterilization. When the amount of total

solids increased, it was clear that the methane production per

kilo treated waste increased.In summary this study showed that mesophilic tempera-

tures are preferable for digestion of animal by-products unless

sufficiently diluted by co-digestion with a substrate of lower

Lipids (g kg-1)

Fig. 6 Relation between lipids (g kgL1) and methane yield

(% of theoretical potential) and total VFA concentration

(mM) at 55 8C. Data are from anaerobic batch assays of

different animal by-products.0

20

40

60

80

100

120

% o

f th

eo

retical C

H4 yield

0

200

400

600

800

1000

To

tal V

FA

(m

M)without inhibition. Although sterilization or pasteurization

did not increase the methane yield of animal waste, it is

necessary in order to comply with the EC Byproduct regula-

tions for category 2 and category 3materials for use in biogas

reactors.

5. Conclusions

Weconclude that, animalwaste constitutes a good substrate for

biogas production with a methane potential of mixed animal

waste of 619 dm3 kg1, which is much higher than themethanepotential ofmanures (2030 dm3 kg1).Animalwaste ishowevercontaining high ammonia loads, and due to the susceptibility of

the thermophilic digestion to ammonia, itwouldbe better to use

mesophilic digestion. Co-digestion is obviously very suitable for

these resources up to a dilution level of 5%.

Pretreatment to satisfy the EU regulations No. 1774/2002

for health and safety categories 2 and 3 was demonstrated to

have no effect on the treatability or methane yield.

Acknowledgement

We acknowledge M. Eiris for her contribution to this work.

Partial fundingwas fromCOOP-CT-2005 (C.N 017641) and from

the EU-FP6 CRAFT project PIGMAN.

r e f e r e n c e s

[1] The Danish Pigmeat Industry, http://danishmeat.eu/DMA_Home/information/ pigmeat_indus.aspx; 3 March 2009.

[2] Guidance note on the disposal of animal by-products andcatering waste. Ministry of Agriculture, Fisheries, Food,http://www.defra.gov.uk/animalh/by-prods/default.htm;January 2001 [accessed 03.03.09].

[3] EC byproduct regulation. European Parliament and Council.Regulation (EC) No 1774/2002 of the European Parliament andof the Council of 3 October 2002 laying down health rulesconcerning animal by-products not intended for humanconsumption. Official Journal 10/10/2002;L273.

[4] Salminen E, Rintala J. Anaerobic digestion of organics solidpoultry slaughterhouse waste a review. BioresourceTechnology 2001;83(1):1326.

[5] Banks CJ, Wang Z. Development of a two phase anaerobicdigester for the treatment of mixed abattoir wastes. WaterScience and Technology 1999;40:6776.

[6] Angelidaki I, Ahring BK. Effects of free long-chain fatty acidson thermophilic anaerobic digestion. Applied Microbiologyand Biotechnology 1992;37:80812.

[7] Hansen HK, Angelidaki I, Ahring BK. Anaerobic digestion ofswine manure: inhibition by ammonia. Water Research 1996;32(1):512.

[8] Angelidaki I, Petersen SP, Ahring BK. Effects of lipids onthermophilic anaerobic digestion and reduction of lipidinhibition upon addition of bentonite. Applied Microbiologyand Biotechnology 1990;33:46972.

[9] Broughton MJ, Thiele JH, Birch EJ, Cohen A. Anaerobic batch

degrading long-chain fatty acids. Applied EnvironmentalMicrobiology 1995;61:24425.

[11] Cirne DG, Bjornsson L, Alves M, Mattiasson B. Effects ofbioaugmentation by an anaerobic lipolytic bacterium onanaerobic digestion of lipid-rich waste. Journal of ChemistryTechnology and Biotechnology 2005;81(11):174552.

[12] Carpentier J, Platteau W, Vanwalleghem J, Steenhoudt D,Verstraete W. Anaerobic digestion of solid slaughterhousewaste: potential of renewable energy for Belgium. In:Proceedings of the 4th symposium of anaerobic digestion ofsolid waste: Copenhagen; 2005. p. 64955.

[13] Rosenwinkel KH, Meyer H. Anaerobic treatment ofslaughterhouse residues in municipal digesters. WaterScience and Technology 1999;40:10111.

[14] Edstrom M, Nordberg A, Thyselius L. Anaerobic treatment ofanimal by-products from slaughterhouses at laboratory andpilot scale. Applied Biochemistry Biotechnology Part AEnzyme Engineering and Biotechnology 2003;109:12738.

[15] Angelidaki I, Sanders W. Assessment of the anaerobicbiodegradability of macropollutants. Reviews inEnvironmental Science and Biotechnology 2004;3(2):11729.

[16] APHA. Standard Methods for the Examination of Water andWastewater. 20th ed. Washington, DC: American PublicHealth Association; 1998.

[17] Angelidaki I, Ahring BK. Thermophilic digestion of livestockwaste: the effect of ammonia. Applied Microbiology andBiotechnology 1993;38:5604.

[18] Harmon JL, Svoronos SA, Lyberatos G, Chynoweth D.Adaptive temperature optimization of continuous digesters.Biomass and Bioenergy 1993;5:27988.

[19] Westermann P, Ahring BK, Mah RA. Temperaturecomparison in Methanosarcina barkeri by modulation of

b i om a s s an d b i o e n e r g y 3 3 ( 2 0 0 9 ) 1 0 4 6 1 0 5 41054digestion. Biotechnology Letters 1998;3:15964.[10] Angelidaki I, Ahring BK. Establishment and characterization

of an anaerobic thermophilic (55 C) enrichment culturehydrogen and acetate affinity. Applied and EnvironmentalMicrobiology 1989;55:12626.

Anaerobic digestion of slaughterhouse by-productsIntroductionMaterials and methodsSubstratePretreatmentMethane yieldContinuously stirred tank reactor experimentsAnalytical methodsCalculation on theoretical methane yield

ResultsCharacterization of by-productsBatch experimentsThe effect of pretreatmentThe effect of temperature

Continuously stirred reactor experimentsThe thermophilic reactors (55degC)The mesophilic reactor (37degC)

DiscussionConclusionsAcknowledgementReferences