ARTICLE IN PRESS

WAT E R R E S E A R C H 42 (2008) 2195 2203

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Influence of wastewater characteristics on methane

potential in food-processing industry wastewaters

L. Maya-Altamira, A. Baun, I. Angelidaki, J.E. Schmidt

Institute of Environment and Resources, Technical University of Denmark, Building 115, DK-2800 Kgs. Lyngby, Denmark

art i cle info ab st rac t

Article history: In this study, we identified the influence of wastewater characteristics on the theoretical
Received 28 June 2007 and practical methane potential using different food industrial wastewaters as substrates.
Received in revised form Ten composite wastewater samples from five industries were investigated. The ultimate
26 November 2007 practical methane yields (Bo) were compared to the theoretical methane yields (Bo,th) in
Accepted 27 November 2007 order to evaluate the biodegradability of the tested wastewaters and the influence of their
Available online 3 December 2007 physico-chemical characteristics. The analytical method applied to quantify the waste-
waters organic content proved to influence the estimation of their theoretical yields. The
Keywords:
substrate:inoculum ratio as well as the dilution factor of the wastewaters influenced the
Anaerobic digestion
ultimate practical methane yields differently in each of the wastewaters assessed.
Food-processing industry
Substrate chemical oxygen demand (COD) concentrations did not present any influence
wastewater
on ultimate practical methane yields; on the other hand, it was found that they were
Methane yield
affected positively by concentrations of total inorganic carbon when wastewaters were 25%
Wastewater characterization
and 50% diluted and affected negatively by concentrations of total acetate when
wastewaters were undiluted. Carbohydrate and protein concentrations affected negatively
the maximum achieved practical methane yields.
& 2007 Elsevier Ltd. All rights reserved.

1. Introduction high potential for methane production, due to their high

organic matter content. Thus, by anaerobic treatment of
Anaerobic digestion is a complex biological process regulated wastewaters is achieved not only removal of pollutants of the
by a bacteria consortium, which is sensitive to the type of wastewater streams but also the production of renewable
substrate and its composition. Specific substrates are utilized energy in the form of methane.
by specific types of bacteria, whilst exerting a symbiotic The theoretical methane yield, Bo,th, is widely recognized to
influence on one anothers activity forming an anaerobic food predict the methane potential of a specific waste. This can be
chain (Gavala and Lyberatos, 2001; Angelidaki and Sanders, calculated by knowing either the atomic composition of the
2004; Dearman et al., 2006). Substrates are hydrolyzed to waste or its organic fractions (Angelidaki and Sanders, 2004).
sugars, amino acids, and fatty acids which are further If the atomic composition is known, it can be estimated by
consumed by acetogens excreting acetate, hydrogen, and assuming total conversion of the organic matter based on its
carbon dioxide. These simplistic compounds are consumed composition, by a stoichiometric conversion to methane and
readily by methanogens, co-existing with the acetogens. carbon dioxide (Buswell and Neave, 1930):
Anaerobic digestion has been widely used for the treatment
     
of industrial wastewaters (Austermann-Haun et al., 1997; a b n a b n a b
Cn Ha Ob n   H2 O !  CH4  CO2 .
Lyberatos et al., 1997; del Pozo et al., 2003; Bernet and Paul, 4 2 2 8 4 2 8 4
2006). Food-processing industry wastewaters may represent (1)

Corresponding author. Tel.: +45 4525 1450; fax: +45 4593 2850.
E-mail addresses: lma@er.dtu.dk (L. Maya-Altamira), anb@er.dtu.dk (A. Baun), ria@er.dtu.dk (I. Angelidaki), jes@er.dtu.dk
(J.E. Schmidt).
0043-1354/$ - see front matter & 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2007.11.033
 ARTICLE IN PRESS
2196 WAT E R R E S E A R C H
If the organic fractions, i.e. content of carbohydrates,
proteins, lipids, acetate, and propionate, are known, it can
also be estimated using
Bo;th 0:415x carbohydrates 0:496x proteins 1:014x lipids
0:373x acetate 0:530x propionate;
where x represents the fractions that have been quantified
from the analytical measurements of total volatile solids (VS),
organic nitrogen (Norg), lipids, and volatile fatty acids (total
acetate and total propionate) in the substrate. The coefficients
in this equation are derived from the stoichiometric conver-
sion (Eq. (1)) of model compounds representing average
formulas for carbohydrates (C6H10O5), proteins (C5H7NO2),
and lipids (C57H104O6), and the formulas for acetate and
propionate. In this way the theoretical methane yield, Bo,th,
based on substrates organic fractions is in fact derived from
the first method but taking into account the analytical organic
fractions (Angelidaki and Sanders, 2004).
Although the theoretical methane potential gives a rough
idea of the amount of biogas production to be expected from
the waste composition, the practical yield obtained in a
biogas reactor will always be lower due to a number of factors,
such as the anaerobic biodegradability of the organic matter
in the wastewater, and the fact that the theoretical methane
yield does not take into account that a definite amount of the
organic matter (510%) goes to build new bacterial biomass.
Hashimoto et al. (1981) defined the ultimate practical
methane yield as Bo, which is the specific volume of methane
produced in a batch experiment per gram of substrate
added, expressed either as total VS or as chemical oxygen
demand (COD).
Increasing substrate:inoculum ratios may affect negatively
the ultimate practical methane yield (Hashimoto, 1989; Chen
and Hashimoto, 1996). However, other studies have shown no
significant influence (Gungor-Demirci and Demirer, 2004;
Raposo et al., 2006; Zhang et al., 2007). Organic or inorganic
toxicants contained in the wastewater might inhibit the
process, so dilution of the wastewater might also influence
significantly the practical methane potential. Therefore,
dilution should be done when a substrate is used for the first
time, to ensure that possible toxicity of the sample is not
overseen (Angelidaki and Sanders, 2004).
Table 1 Origin and description of wastewater samples
Sample Origin (production)
VFO Vegetable fats and oils
FBH Fish bites for human consumption
VPE Vegetable production: peas
VLO Vegetable production: leek and fried onion
FMW Fish meals
FMS Fish meals
SPR Slaughterhouse production raw
SPE Slaughterhouse production effluent
SPF Slaughterhouse flocculation concentrate
SPQ Slaughterhouse flocculation liquid
All samples were collected in triplicate.
42 (2008) 2195 2203
Many successful implementations of anaerobic digestion for
single wastewater streams coming from diverse food indus-
tries have been done during the past two decades. However, in
several cases, inhibition of the anaerobic processes has been
recorded due to high concentrations of ammonia and lipids or
2
to the presence of xenobiotic chemicals in wastewater
streams, e.g. Angelidaki and Ahring (1994), Madsen and
Rasmussen (1996), Batstone et al. (1997), Gungor-Demirci and
Demirer (2004), and Fernandez et al. (2005).
There is still no existing systematic study identifying the
coupling of wastewater characteristics on their methane
potential and anaerobic biodegradability. Wastewater physi-
co-chemical characteristics, such as its concentrations of VS,
nitrogen, lipids, and volatile fatty acids, and its oxidation and
buffer capacities expressed as COD and total alkalinity (TAL),
respectively, may exert specific influences over their methane
potential.
The aim of the present study was therefore to analyze the
influence of the wastewaters physico-chemical characteris-
tics on the predicted and achieved methane potentials.
Differences in physico-chemical characteristics were repre-
sented by seasonal and pre-treatment variations of five food-
processing industries.
2. Materials and methods
2.1. Wastewater sampling
Wastewater sampling was conducted with the focus of
covering possible variations in wastewater composition both
within the same food-processing industry and amongst
different industries (see Table 1). A number of composite
samples were taken according to sample availability. Samples
were analyzed according to the standard methods (APHA
et al., 2005) for COD, total organic carbon (TOC), total
inorganic carbon (TIC), total VS, total Kjeldahl nitrogen
(TKN), free ammonium nitrogen (NH+4 ), TAL, pH, lipids, total
acetate, and total propionate. From information provided
by the industries, total phosphorus (TP) was not determined
in samples with concentrations reported to be lower
than analytical detection limits (o1 mg/l). According to the
Sampling point Sample type
Discharge to municipality One-week composite sample
Discharge to municipality
Discharge to municipality
Discharge to municipality
Winter discharge to municipality
Summer discharge to municipality
Before pretreatment One-day composite sample
After pretreatment
Flocculation/aeration tank
Inlet to equalization tank
 ARTICLE IN PRESS
WAT E R R E S E A R C H
Table 2 Physico-chemical characteristics for inoculum used for BMP assays
pH COD (g/l) TS (g/l)
7.537.58 17.819.7 22.730.3
industries, there was no significant variation in composition
for the wastewater streams: vegetable fats and oils (VFO), and
fish bites for human consumption (FBH). For the wastewaters
from vegetable production, either of peas (VPE) or leek and
fried onion (VLO), and from production of fish meals at winter
(FMW) or fish meals at summer (FMS), the composition of the
discharged wastewater varied. For the vegetable production, it
varied according to the product in turn, i.e. peas or leek and
fried onion, which are produced in different seasons of the
year. For the fish meal production it varied according to the
product availability as during the winter season the produc-
tion is lower than during the summer season. For the
slaughterhouse (SPR, SPE, SPF, and SPQ), there was no
seasonal variation reported, but samples were collected
before, during, and after pre-treatment, at its physico-
chemical flocculation/aeration tanks and at the inlet to the
equalization tank.
2.2. Wastewater characterization and fractionation
From analytical measurements, the samples were derivated
into their atomic composition CnHaObNcPdu. This was done
by applying the method suggested by Kleerebezem and van
Loosdrecht (2005) to derive the substrates carbon, hydrogen,
oxygen, and nitrogen composition, and the substrates charge
(u). The method suggested by Speece (1996) was applied to
carry the phosphorus derivation. Substrates were also frac-
tionated into carbohydrates, proteins, lipids, total acetate,
and total propionate (organic fractions). In this way, it was
possible to estimate their theoretical methane yields, Bo,th, by
two different methods namely by atomic composition (Eq. (1))
and indirectly by the content of different groups of organic
fractions (Eq. (2)) and compare them with Bo.
The atomic derivation was done by calculating molarity
concentrations of TOC, TIC, COD, TKN, NH+4 , TP, and TAL
analytical results (see Speece, 1996; Kleerebezem and van
Loosdrecht, 2005). The organic fractionation was done by
analytical measurements of VS, lipids, TKN, NH+4 , and total
acetate and propionate.
2.3. Biological methane potential (BMP) assays for
wastewater streams
The food-processing industries wastewater composite sam-
ples were used as substrates for the BMP assays. Sludge
samples from an anaerobic continuous stirred tank reactor at
the mesophilic temperature range, i.e. 3537 1C, were used as
inoculum. The inoculum was collected at a wastewater
treatment plant treating 114,000 P.E. of domestic and indus-
trial wastewater. Inoculum was incubated for approximately
1 week at 35 1C in order to reduce the biodegradable content in
the sludge. Inoculum characteristics are shown in Table 2,
42 (2008) 2195 2203 2197
VS (g/l) TKN (g/l) NH+4 (g/l)
12.416.3 1.31.4 1.82.0
showing typical ratios for (i) COD/TKN at around 10 and
(ii) COD/VS at around 1.2. Additional nutrients were not
supplied for the assays as inoculum counted on enough
nutrients as shown by COD/TKN ratio.
The assays were performed in triplicate in glass serum vials
of 118 ml, filled with 25 ml of a suspension of inoculum and
wastewaters which were homogenized by stirring before
transfer to the vials. Before sealing the vials with rubbers
stoppers and metallic crimps, the suspensions were flushed
with 20:80 CO2:N2 for 10 min each. An inoculum volume of
5 ml was used for all batch reactors. The substrate:inoculum
ratio was set at four wastewater dilution levels of 25%, 50%,
75%, and 100% of wastewater concentration; at each dilution
level distilled water was added to conserve the same volumes.
Three controls with only inoculum were included. Batch
reactors were incubated at 35 1C during the whole experi-
ment. The duration of the experiments was different for each
wastewater assessed since different time periods were
needed to reach the point where the cumulative methane
production was constant. In order to assure that no additional
methane production occurred, samples were kept in the
incubator for a time period of about 2030 days from the time
when the last couple of measurements showed no methane
production.
Cumulative methane production from the wastewaters
assays was measured by a gas chromatrograph with an
FID detector. At the beginning of the experiments methane
was measured daily, whilst when cumulative production
was stabilized it was measured less often. Cumulative
methane production was monitored as total production
and correlated production, i.e. without and with sub-
tracting methane production from inoculum controls. pH
was measured at the beginning and at the end of the
experiments.
2.4. Calculation of theoretical methane yield (Bo,th),
ultimate practical methane yield (Bo), and methane-based
biodegradability (MBD) of the substrates
The methods relative accuracy performance in calculating
the theoretical methane yield (Bo,th) of the wastewater
samples was tested.
Angelidaki (2004) considers that all the organic nitrogen is
mineralized to ammonium and/or taken up by biomass; thus
the yields equations based on substrates atomic composition
are defined as
n=2 a=8  b=422:4
Bo;th STP lCH4 =gVS , (3)
12n a 16b
n=2 a=8  b=422:4
Bo;th STP lCH4 =gCOD . (4)
n a=4  b=232
 ARTICLE IN PRESS
2198 WAT E R R E S E A R C H
To calculate Bo,th based on the substrates organic fractions,
Eq. (2) was applied. To calculate Bo,th based on the substrates
molecular weight from its atomic composition, Eq. (3) was
applied.
To calculate Bo,th based on substrates oxidation state from its
atomic composition, Eq. (4) was applied. This can be calculated
for all types of substrates to be 0.35 STP l-CH4/gCOD.
The ultimate practical methane yield, Bo, was calculated by
dividing the last point of cumulative methane production
(correlated) from the BMP assays, named B by Chen and
Hashimoto (1996), by the initial loading of COD substrate. Bo
was only expressed in COD units for this study as COD
analytical results represented more accurately the organic
content in the wastewaters. The methane-based biodegrad-
able fraction of the organic substrate (MBD) was calculated by
dividing the maximum achieved practical methane yield in
COD units, Bo, by what was theoretically expected also in COD
units, i.e. 0.35 STP l-CH4/gCOD. This gave a fraction which
could potentially be degraded, based on methane production.
3. Results
3.1. Wastewater physico-chemical characteristics
The physico-chemical characterization of the wastewaters
showed that the main differences were in total acetate, TOC,
TIC, COD, VS, and lipids concentrations (Table 3). Some
samples were low in volume to analyze in triplicate for all
the parameters; thus confidence intervals for 95% could not be
established for some of the measurements. In seven of the 10
wastewater samples, TAL values presented confidence inter-
vals 410% of the average values. TIC also presented wide
confidence intervals, and in half of the cases they could not be
established. These two parameters reflect the charge of the
substrate. pH and nitrogen values were more uniform. For
none of the samples was it possible to establish a correlation
between TOC and COD. Total acetate concentrations were high
for VFO and the pre-treated slaughterhouse effluent (SPE).
The derivation into chemical constituents showed that fish
products for human consumption (FBH), and the raw (SPR),
flocculated concentrate (SPF), and flocculated liquid (SPQ)
slaughterhouse wastewaters contained high amounts of
hydrogen bound to organic carbon in relation to the other
elements contained in the substrate, i.e. these wastewaters
contained more reduced organic molecules. All organic
substrates possessed low organic nitrogen content. Free
ammonium nitrogen (NH+4) concentration was relatively
high only in fish meals wastewater during the winter (FMW),
and in this case an influence of NH+4 on the pH was observed.
For the rest of the wastewaters, pH did not seem to be
influenced by the chemical composition of the substrates. VFO
and VPE were the only wastewaters presenting significant TP
concentrations.
3.2. Comparison of theoretical, cumulative, and ultimate
practical methane yields (Bo,th, B, Bo)
The cumulative practical methane yield (B) was registered
over time during the BMP assays (only from correlated
42 (2008) 2195 2203
methane production). As an example of a typical methane
production graph, the pre-treated slaughterhouse effluent
(SPE) is shown in Fig. 1. Bo,th, the theoretical methane yield
based on the substrates atomic composition, is also pre-
sented for comparison with Bo. The dilution levels represent
25%, 50%, 75%, and 100% of wastewater concentration. It was
observed that the methane production for this particular
wastewater at 25% and 75% levels did not stop before the last
incubation period; however, afterwards, the last measure-
ment showed no more production. It was also observed that
as the dilution increased, the methane yields increased. This
pattern was seen throughout the experiment and at the
ultimate practical methane yields. All the rest of the waste-
waters, except fish meals effluents (FMW and FMS) diluted
and undiluted, stopped producing methane before the last
period in the incubator. FMW and FMS were kept in the
incubator for an additional period of 50 days and still an
increased methane production of about 5% was observed. At
this point, the experiments were terminated since this time
period was already unrealistic for a digester to retain a slowly
degradable wastewater as such.
Table 4 shows that when the theoretical methane yield in
VS units was estimated on substrates organic fractions (Eq.
(2)), it was higher for seven of 10 wastewater samples than
when estimated on substrates atomic compositions (Eq. (3)).
For the remaining three samples, fish meals for human
consumption (FBH), slaughterhouse production raw (SPR),
and slaughterhouse flocculated liquid (SPQ), the higher
theoretical methane yields estimated based on atomic
composition may be related to the higher hydrogen contents
in the organic molecules (cf. Eq. (3)).
By normalizing the theoretical methane yield values
expressed in VS units [Bo,th Org.Fract. (from Eq. (2)) and Bo,th
VS (from Eq. (3))] to COD units, it was possible to compare
them to what was theoretically expected as the normalized
value, i.e. 0.35 STP l-CH4/gCOD, and to the maximum practical
methane yields (see Table 5). By comparing them to the
normalized value, their relative accuracy performance was
evaluated. The normalization was done by COD/VS theore-
tical ratios (for Bo,th VS), and by COD/VS analytical ratios (for
Bo,th Org.Fract.), due to the fact that the former is calculated
on theoretical substrates VS values and the latter on
analytical substrates VS values. Theoretical yields calculated
from substrates atomic composition, i.e. Bo,th VS and Bo,th
COD (Eq. (4)), gave the closest to the maximum practical
methane yields, Bo, in nine wastewaters whilst theoretical
methane yields calculated from organic fractions, Bo,th
Org.Fract., resulted closer in only one, i.e. VFO. When Bo,th,
VS was normalized to COD units, approximate values of 0.35
STP l-CH4/gCOD were obtained, but when Bo,th Org.Fract.
yields were normalized to COD units, values were lower than
the 0.35 STP l-CH4/gCOD for nine of the 10 wastewater
samples. This indicates that analytical VS values were not
as representative of the organic content in the liquid
substrates as the theoretical VS values.
Maximum practical methane yields were achieved when
the wastewaters were diluted four times, i.e. at 25% dilution,
for eight of the 10 samples (see Table 5). The two exceptions
were samples from production of leek and fried onion (VLO)
and fish meals during the winter (FMW), which were achieved
 ARTICLE IN PRESS
WAT E R R E S E A R C H 42 (2008) 2195 2203 2199

Chemical composition

C0.0H0.3O0.1N0.0P0.01
C0.4H0.3O0.7N0.0P0.02

C0.2H0.5O0.4N0.0P0.01
C0.4H0.6O0.7N0.0P0.01

C0.2H0.2O0.3N0.0P0.01

C0.1H0.4O0.1N0.0P0.01
C0.1H1.1O0.1N0.0P0.0
0.0

C0.0H0.6O0.0N0.0P0.0
0.0
C0.0H0.1O0.1N0.0P0.0
0.0

C0.0H0.4O0.0N0.0P0.0
0.0
0.0
0.1

0.0

0.0

0.0

0.0
(mol/l)c

23.774.2

21.771.7
13.678.1
16.870.5

12.672.6
11.176.6
11.272.7
(meq/l)

5.470.0
8.270.8

6.472.1
TAL

69.4733.8
73.9724.1
9.1714.5

82.079.3
9.871.1
(mg/l)

156.3a

133.6a
65.6a

72.6a

61.6a
TIC

Fig. 1 Theoretical (Bo,th) and cumulative practical (B)

methane yields during the biological methane potential
assay for pre-treated slaughterhouse effluent (SPE) at
203.57172.2
259.3748.5
244.1727.6

600.6723.9

130.7714.8
113.8726.5
211.8724.0
202.1722.9

different wastewater dilution levels.

156.4a
12.5a
NH+4
(g/l)

at 50% dilution. By introducing high wastewater dilutions in a

Table 3 Characteristics of wastewater samples expressed in average values795% confidence intervals

VFA (TAc/TPro)

BMP assay, some of the maximum practical methane yields

and the methane-based biodegradable fractions (MBD) were
2.0/0.1
0.0/0.0
0.4/0.2
0.4/1.4
0.4/0.0
0.3/0.0
0.0/0.0
0.9/0.2
0.4/0.0
0.2/0.1
(g/l)a

overestimated. It was observed that for six of the wastewaters

Parameters

they exceeded the constant theoretical yield expected, i.e.

0.35 STP l-CH4/gCOD. Samples from production of VFO
showed the lowest maximum practical methane yields and
methane-based degradability fractions (MBD). For the waste-
0.070.0

2.070.0
1.770.2
1.070.1
0.070.0
1.370.1

0.070.0
Lipids

Calculated from the difference of total Kjeldahl nitrogen minus total free ammonium nitrogen.

waters from vegetable production no differences in max-

(g/l)

0.7a

0.0a

0.7a

imum practical yields and the MBD were observed when

comparing peas (VPE) and leek and fried onion (VLO) waste-
water samples. Wastewater from the fish meals production in
odet.lim.
odet.lim.
4.070.4

0.370.1
0.070.0

Estimated according to Speece (1996) and Kleerebezem and van Loosdrecht (2005).
n.m.

n.m.
n.m.
n.m.
n.m.

winter season (FMW) gave higher yields than for summer

(g/l)
TP

production (FMS). For the slaughterhouse production waste-

waters it was found that the effluent after pre-treatment
(SPE) had significantly lower yields and MBD than the raw
(g/l)b
Norg

wastewater (SPR). However SPR methane yield was inferior

0.1
0.1
0.2
0.0
0.0
0.1
0.4
0.0
0.2
0.1

than the yields from the flocculated solid and liquid streams
(i.e. SPF and SPQ, respectively). Thus, to increase the
3.870.0
1.970.3

4.270.3
0.770.0
1.970.3
1.870.3

maximum practical methane yield and methane-based

(g/l)

6.7a
4.7a
0.4a
0.7a
VS

biodegradable fraction, the slaughterhouse wastewater

should be flocculated and not diluted at the pre-treatment
unit.
From single or duplicates composite samples.
(g/l)a
TOC

5.2
0.9
2.9
3.2
0.5
2.1
0.5
0.5
0.8
0.3

3.3. Influence of substrate:inoculum ratio and dilution on

ultimate practical methane yields
11.370.1
17.571.7

10.371.1

10.471.0
8.370.1

4.270.3

1.870.2
6.970.9
6.970.2
COD
(g/l)

3.9a

The influence of substrate:inoculum ratio and the waste-

water dilution levels on the practical methane yields is shown
for each of the wastewaters in Fig. 2. VFO presented a
decreased methane yield with increasing ratio and lower
7.4470.55
6.9270.46
6.4271.83
6.1970.60
9.0270.21
7.8071.57
7.3170.22
7.3470.17
7.2770.57
7.0670.26

odet.lim. 0.0001 g/l.

n.m. not measured.

dilution (i.e. from 25% to 100% wastewater concentration).

pH

Fish products for human consumption (FBH) and vegetable

production for peas (VPE) presented similar yields at 25% and
50% dilution levels each, but for the former the influence was
Sample

noticed at 75% and for the latter at 100%. Vegetable

FMW
FMS
VFO

VLO
FBH

SPQ
VPE

SPR
SPE
SPF

production for leek and fried onion (VLO), differently from

b
a

peas (VPE), gave higher yields at the 50% dilution level similar
 ARTICLE IN PRESS
2200 WAT E R R E S E A R C H 42 (2008) 2195 2203

to fish meals production during the winter (FMW), and a also affected (see Fig. 3). Yields varied between 0.26 and 0.41
decrease at 75% and 100%. Fish meals production during the STP l-CH4/gCOD at the substrates total acetate concentra-
summer (FMS) presented lower yields than production during tions in the range of 8375 mg/l, whilst at concentrations
the winter (FMW) at 25%, 50%, and 75% dilution levels. in the range of 375500 mg/l yields varied between 0.16
Slaughterhouse raw (SPR), flocculated solid concentrate and 0.27 STP l-CH4/gCOD. At five of the BMP experi-
(SPF), and flocculated liquid (SPQ) gave similar yields. For the ments, the substrates total acetate concentrations exceeded
slaughterhouse wastewaters, the pre-treated effluent (SPE) 500 mg/l and yields were seriously affected. As pH values at
showed a significant decrease in practical methane yields for the beginning of the BMP assays were in the range of 6.49.0
samples diluted less than four times (25%). (results not shown) most of the total acetate measured was
assumed to be ionized.
Free ammonium nitrogen (NH+4 ) was present in very low
3.4. Influence of wastewater physico-chemical concentrations in the most tested wastewaters, except for
components on maximum practical methane yields FMW. Ultimate practical methane yields were not affected by
either NH+4 or lipid content in any of the samples (data not
A decrease in maximum practical methane yields at increas- shown).
ing total acetate concentrations in the substrate was ob- It was not possible to find a correlation between the
served. Furthermore, ultimate practical methane yields were ultimate practical methane yield and substrates COD con-
centration. This can be observed in Fig. 2, as for some
wastewaters the range of substrate:inoculum ratios for which
Table 4 Theoretical methane yields based on a satisfactory methane yield could be achieved was very
(i) theoretical VS values derived from analytical waste- narrow whilst for others it was much broader.
water characteristics (Eq. 3), and, (ii) analytical VS values
A statistical ranking analysis was carried out (Snedecor and
and organic fractions (Eq. 2)
Cochran, 1967) to attempt finding which physico-chemical
characteristics influenced the ultimate practical methane
Sample ID Theoretical methane yields yields, Bo, at their maximum achieved and at the different
Bo,th VS Bo,th Org.Fract dilution levels. The method tested the hypothesis of no
STPl-CH4/gVS STPl-CH4/gVS correlation between ultimate practical methane yield and
each of the wastewaters physico-chemical characteristics.
VFO 0.10 0.40 All the concentrations of the parameters shown in Table 3
FBH 0.89 0.66
were evaluated, plus proteins estimated from organic nitro-
VPE 0.18 0.60
gen, and carbohydrates estimated from COD, proteins, lipids,
VLO 0.15 0.66
FMW 0.39 0.81 and volatile fatty acids. These different organic components
FMS 0.11 0.44 were also evaluated as organic fractions, i.e. their relative
SPR 0.98 0.64 contributions to the total COD (%).
SPE 0.19 0.41 The total acetate in the wastewaters was found to correlate
SPF 0.41 0.45 negatively (rs 0.8167, p 0.0054) to the methane yields
SPQ 0.84 0.69

Table 5 Comparison of maximum achieved practical methane yields (Bo) versus theoretical methane yields by different
methods (Bo,th), normalized in COD units

Sample Theoretical methane yields Maximum practical Methane-based

ID methane yielda degradabilityb
Bo,th COD Bo,th VS Bo,th Org.Fract. Bo MBD (%)
STPl-CH4/gCOD STPl-CH4/gCOD

VFO 0.35 0.35 0.14 0.2370.03 66

FBH 0.35 0.35 0.07 0.2770.03 77
VPE 0.35 0.35 0.49 0.3670.05 100
VLO 0.35 0.34 0.30 0.3670.04 100
FMW 0.35 0.35 0.07 0.4170.03 100
FMS 0.35 0.36 0.09 0.3670.01 100
SPR 0.35 0.35 0.26 0.3570.07 100
SPE 0.35 0.35 0.15 0.3270.05 91
SPF 0.35 0.35 0.13 0.3670.03 100
SPQ 0.35 0.35 0.18 0.3670.05 100

Practical yields presented in average values795% confidence intervals and based on batch experiments, same as methane-based degradability.
a
From batch experiments in triplicate.
b
Introduced in Materials and methods.
 WAT E R R E S E A R C H

ARTICLE IN PRESS
42 (2008) 2195 2203
2201
Fig. 2 Influence of substrate:inoculum ratio and wastewater dilution levels on ultimate practical methane yields.
 ARTICLE IN PRESS
2202 WAT E R R E S E A R C H
Fig. 3 Effect of total acetate (TAc) concentrations on
ultimate practical methane yields at all wastewater
dilution levels.
achieved at the 100% level, i.e. at 100% wastewater concen-
tration. This indicates that only at this concentration range
did the initial acetate concentration in the wastewater affect
the ultimate practical methane yield. On the other hand, the
acetate fraction in the total organic matter also correlated
negatively to the ultimate yield at 75% (rs 0.7333, p
0.0156) and 100% (rs 0.55, p 0.0654) levels. This indicates
that the fraction of acetate contained in the total organic
matter of the substrate also exerted a negative influence on
the ultimate yields achieved for the wastewaters studied at
these dilution levels, though at a lower significance level.
Dilution showed a positive correlation only for TIC at 50%
dilution level though with weaker significancei.e. 10%
significance levelindicating a positive influence on the
ultimate practical methane yield at this dilution. However,
providing the statistical uncertainties from the wide con-
fidence intervals at the analytical measurements or the lack
of them, it is not possible to state a solid statement out of this
value. On the other hand, when evaluating the influence of all
wastewater characteristics on the maximum practical
methane yield, TIC again showed a positive correlation at
10% of significance. In this case, carbohydrate and protein
concentrations showed also a correlation but negative at 10%
and 8% of significance, respectively.
Different values within the confidence intervals for these
four parameters (total acetate, TIC, carbohydrates, and
proteins) did not increase either the correlations or the
likelihoods.
4. Discussion
Large variations in physico-chemical characteristics of the
wastewaters, even coming from the same food-processing
industry were observed, e.g. fish meals and slaughterhouse.
Results from the determination of theoretical methane
potentials based on organic fractions (Bo,th Org.Fract.) de-
viated more from the expected value than the determinations
based on substrates atomic compositions (Bo,th COD and Bo,th
VS). The quantification of organic matter by VS analysis to
wastewaters studied was less representative than by COD
42 (2008) 2195 2203
analysis. This is in accordance with Kleerebezem and van
Loosdrecht (2005) stating that COD measurements are the
most reliable indicators of wastewaters organic content
available for oxidation. So Bo,th COD and Bo,th VS are
recommended in future studies aimed at assessing predicted
methane potential of complex wastewaters (Eqs. (2) and (3)).
Furthermore, the analytical measurements carried out for the
substrates derivation to its atomic composition should be as
accurate and precise as possible.
Substrate COD did not affect the achieved methane
potential, e.g. the VFO industry presenting the lowest yield
(Bo) although its substrate concentration was comparable to
other wastewaters. The low achieved methane potential for
VFO was thus attributed to its high acetate concentrations,
which inhibited methanogenesis at these concentration
levels. The study of the influence of pre-treatment of
slaughterhouse wastewaters showed that flocculation (SPF
and SPQ) gave higher methane yields, whilst flocculation and
equalization together (SPE) did not and that the raw waste-
water (SPR) had higher Bo compared to the pre-treated (SPE).
The main difference in physico-chemical characteristics of
these samples was that SPE had acetate as the major organic
component (74%), whilst SPR contained also lipids, proteins,
and carbohydrates. After flocculation, the liquid fractions
(SPQ) showed higher acetate concentrations than the solid
fractions (SPF); however, the presence of bicarbonate alkali-
nity resulted in a similar maximum practical methane yield
for both wastewater fractions.
Concentrations of total acetate in wastewaters influenced
negatively the ultimate practical methane yields, particularly
over 375 mg/l. The ranking analysis confirmed it and also
showed that at highly diluted wastewaters, i.e. 25% and 50%,
the positive effect of inorganic carbon as bicarbonate
alkalinity may have contra rested it. When the bicarbonate
alkalinity contributes greatly to the TAL, e.g. 30% for VLO
whilst for VPE was only 10%, a coupled effect with total
acetate influences positively the practical methane yields as
described by Kyazze (2007). At undiluted wastewaters, i.e.
100% level, the bicarbonate alkalinity effect was not enough
to contra rest the toxicity effect of the total acetate, thus the
absence of a correlation. However, a lower analytical un-
certainty is necessary at the application of TIC analytical
methods for complex wastewaters as such. Considering the
maximum practical methane yield, which was achieved at
higher wastewater dilutions, i.e. 25% and 50%, TIC presented
the same significance from the ranking analysis (rs 0.6272,
p 0.0516).
5. Conclusion
The assessment of the practical methane potential of five
food-processing industries wastewaters showed that estima-
tion of theoretical methane yields based on atomic composi-
tion (Bo,th COD and Bo,th VS) were more exact than when based
on the substrates organic fractions (Bo,th Org.Fract.). This is
due to the fact that VS analytical results were less represen-
tative than COD measurements.
A negative effect of wastewaters total acetate concentra-
tions on the ultimate practical methane yield (Bo) was
 ARTICLE IN PRESS
WAT E R R E S E A R C H
observed when they were not diluted whilst their TIC
concentrations had a positive effect at 25% and 50% levels
of dilution. Substrate COD did not present any effect.
When dilution is considered in the assessment of methane
potential and methane-based degradability fraction of un-
known wastewaters, it provides valuable information on the
optimal dilution ranges for obtaining maximum practical
methane yields and identifying influencing wastewater
characteristics. This information gives the basis for further
handling alternatives such as co-digestion where the
methane production is maximized by reducing the waste-
waters inherent toxicity properties.
Acknowledgments
Hector Garcia is kindly acknowledged for his technical
support. Behrang Pazhadek is kindly acknowledged for his
experimental work regarding the biological methane poten-
tial assays for the slaughterhouse and fish bites wastewaters.
The National Council for Science and Technology of Mexico
(CONACYT) is acknowledged for its funding.
R E F E R E N C E S
APHA, AWWA, WPCF, 2005. Standard Methods for the Examina-
tion of Water and Wastewater, 21st ed. American Public
Health Association, American Water Works Association,
Washington, DC.
Angelidaki, I., Ahring, B.K., 1994. Anaerobic thermophilic
digestion of manure at different ammonia loadseffect of
temperature. Water Res. 28 (3), 727731.
Angelidaki, I., 2004. Environmental Biotechnology Notes. Irini
Angelidaki, Institute of Environment and Resources, Technical
University of Denmark, Lyngby, Denmark.
Angelidaki, I., Sanders, W., 2004. Assessment of the anaerobic
biodegradability of macro-pollutants. Rev. Environ. Sci.
Bio/Technol. 3, 141158.
Austermann-Haun, U., Seyfried, C.F., Rosenwinkel, K.H., 1997. Full
scale experiences with anaerobic pre-treatment of wastewater
in the food and beverage industry in Germany. Water Sci.
Technol. 36 (23), 321328.
Batstone, D.J., Keller, J., Newell, B., Newland, M., 1997. Model
development and full scale validation for anaerobic treatment
of protein and fat based wastewater. Water Sci. Technol. 36
(67), 423431.
Bernet, N., Paul, E., 2006. In: Cervantes, F.J., Pavlostathis, S.G., van
Haandel, A.C. (Eds.), Application of Biological Treatment
Systems for Food-Processing Wastewaters. IWA Publishing,
London, pp. 237266 (Chapter 7).
Buswell, E.G., Neave, S.L., 1930. Laboratory Studies of Sludge
Digestion. Illinois Division of State Water Survey, 30.
42 (2008) 2195 2203 2203
Chen, T.H., Hashimoto, A.G., 1996. Effects of pH and substrate:
inoculum ratio on batch methane fermentation. Bioresour.
Technol. 56, 179186.
Dearman, B., Marschner, P., Bentham, R.H., 2006. Methane
production and microbial community structure in single-
stage batch and sequential batch systems anaerobically co-
digesting food waste and biosolids. Appl. Microbiol. Biotech-
nol. 69 (5), 589596.
Del Pozo, R., Tas, D.O., Orhon, D., Diez, V., 2003. Biodegradability of
slaughterhouse wastewater with high blood content under
anaerobic and aerobic conditions. J. Chem. Technol. Biotech-
nol. 78 (4), 384391.
Fernandez, A., Sanchez, A., Font, X., 2005. Anaerobic co-digestion
of a simulated organic fraction of municipal solid wastes and
fats of animal and vegetable origin. Biochem. Eng. J. 26 (1),
2228.
Gavala, H.N., Lyberatos, G., 2001. Influence of anaerobic culture
acclimation on the degradation kinetics of various substrates.
Biotechnol. Bioeng. 74 (3), 181195.
Gungor-Demirci, G., Demirer, G.N., 2004. Effect of initial COD
concentration, nutrient addition, temperature and microbial
acclimation on anaerobic treatability of broiler and cattle
manure. Bioresour. Technol. 93 (2), 109117.
Hashimoto, A.G., 1989. Effect of inoculum/substrate ratio on
methane yield and production rate from straw. Biol. Wastes 28,
247255.
Hashimoto, A.G., Varel, V.H., Chen, Y.R., 1981. Ultimate methane
yield from beef cattle manure: effect of temperature, ration
constituents, antibiotics and manure age. Agric. Wastes 3,
241256.
Kleerebezem, R., van Loosdrecht, M.C.M., 2005. Waste character-
ization for implementation in ADM1. In: Proceedings of the
First International Workshop on the IWA Anaerobic Digestion
Model No. 1 (ADM1), pp. 193200.
Kyazze, G., 2007. Hydrogen production via dark fermentation of
carbohydrate-rich substrates. Ph.D. Thesis, submitted to
University of Flamorgan/Prifysgol Morgannwg, Pontypridd,
Wales.
Lyberatos, G., Gavala, H.N., Stamatelatou, A., 1997. An integrated
approach for management of agricultural industries waste-
waters. Non-linear Anal. Theor. Methods Appl. 30 (4),
23412351.
Madsen, T., Rasmussen, H.B., 1996. A method for screening the
potential toxicity of organic chemicals to methanogenic gas
production. Water Sci. Technol. 33 (6), 213220.
Raposo, F., Banks, C.J., Siegert, I., Heaven, S., Borja, R., 2006.
Influence of inoculum to substrate ratio on the biochemical
methane potential of maize in batch tests. Process Biochem.
41, 14441450.
Snedecor, G.W., Cochran, W.G., 1967. Correlation. In: Statistical
Methods. The Iowa State University Press, Ames, IA, USA,
pp. 172198.
Speece, R.E., 1996. Anaerobic Biotechnology for Industrial Waste-
waters. Archae Press, Nashville, TN, pp. 4345.
Zhang, R., El-Mashad, H.M., Hartman, K., Wang, F., Liu, G., Choate,
C., Gamble, P., 2007. Characterization of food waste as feed-
stock for anaerobic digestion. Bioresour. Technol. 98, 929935.