Bioresource Technology 121 (2012) 419–424

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Prediction of hydrogen sulphide production during anaerobic digestion

of organic substrates
Pascal Peu a,b,⇑, Sylvie Picard a,b, Arnaud Diara a,b, Romain Girault a,b, Fabrice Béline a,b,
Gilbert Bridoux c, Patrick Dabert a,b
a
IRSTEA, UR GERE, 17 Avenue de Cucillé, CS 64427, F-35044 Rennes, France
b
Université Européenne de Bretagne, France
c
SAUR, Société d’Aménagement Urbain et Rural, Département de Recherche et de Développement, 2 Rue de la Bresle, 78312 Maurepas, France

h i g h l i g h t s

" Feedstock used in anaerobic digesters could generate a high level of hydrogen sulphide into the biogas.
" Sulphur content of feedstock is variable with a low level for most of them.
" Hydrogen sulphide biogas content could be predicted with only feedstock analyses.
" Feedstock with carbon:sulphur ratio under 40 present a risk for their use in anaerobic digestion.

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

Article history: The main objective of this study was to develop a methodology to predict the hydrogen sulphide content
Received 19 April 2012 of raw biogas produced during anaerobic mono-digestion of a bioenergy feedstock. Detailed chemical and
Received in revised form 29 June 2012 biological analyses were made on 37 different feedstocks originating from urban wastewater treatment
Accepted 30 June 2012
plants, farms, agri-food facilities and municipal wastes. Total sulphur content ranged from 1 to 29.6 mg S/
Available online 14 July 2012
kg of total solids, and 66% of the feedstocks analysed contained less than 5 mg S/kg of total solids. The
biochemical methanogenic potential and biochemical biogas potential of each feedstock combined with
Keywords:
its S content were used to predict appearance of H2S in the raw biogas. A model to link H2S in biogas with
Anaerobic digestion
Feedstock
the carbon:sulphur ratio was established. Based on this model, a minimum carbon:sulphur ratio of 40 is
Sulphur required in feedstock to limit the concentration of hydrogen sulphide in raw biogas to less than 2% (vol-
Hydrogen sulphide ume/volume).
Biogas Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction In France, energy crops are not encouraged and the primary
vocation of agricultural production is still food. For this reason,
The French agricultural biogas sector is mainly developing in most co-substrates used for agricultural biogas plants originate
areas where livestock effluents are produced (Beline et al., 2010). from (i) agricultural waste, crop residues (residues from field crops,
In 2009, 15 facilities were operating and 90 plants were under con- market gardening, catch crops, silage, etc.), (ii) industrial waste or
struction or planned. The sector is expanding rapidly and new by-products (from slaughterhouses, food processing or wastewater
anaerobic plants are built to respond to the increasing demand treatment plants), or (iii) organic municipal wastes (green waste,
for renewable energy. Livestock wastes appear to be a substrate restaurant refusals and lawn mowing).
of interest but have one major disadvantage: a low organic content The feedstocks used in anaerobic digesters have a wide range of
coupled with low biodegradability (Vedrenne et al., 2008). Conse- chemical characteristics and some are particularly suitable for bio-
quently, using manure on its own in anaerobic digesters is rela- gas production. Nevertheless, some constituents can have adverse
tively rare and co-substrates are often added to increase biogas effects on anaerobic digestion, especially sulphur (S). Under anaer-
production (Mata-Alvarez et al., 2000). obic conditions the inorganic and organic S contained in feedstock
can be reduced (Elferink et al., 1994) or fermented (Mackie et al.,
1998) resulting in the formation of dissolved sulphides which
⇑ Corresponding author at: IRSTEA, UR GERE, 17 Avenue de Cucillé, CS 64427,
can be transferred to the biogas in the form of hydrogen sulphide
F-35044 Rennes, France. Tel.: +33 (0) 223482121; fax: +33 (0) 23482115.
(H2S). H2S smells bad, is toxic, and can damage most equipment,
E-mail address: [email protected] (P. Peu). including combined heat and power engines. For example, as rec-

0960-8524/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2012.06.112
420 P. Peu et al. / Bioresource Technology 121 (2012) 419–424

ommended by most manufacturers, for trouble free operation of cipal wastes (Table 1). Ten kilograms of each product were directly
combined heat and power installations, the H2S concentration in aliquoted and then divided into two subsamples. One part was fro-
biogas must be lower than 100–500 mg/Nm3 (65–330 ppm) zen (20 °C) to prevent degradation, before physico-chemical
depending on the equipment concerned (Wellinger and Linberg, analysis, and the other was refrigerated for measurement of bio-
2000). chemical methanogenic potential (BMP). For chemical determina-
Pollution of biogas by H2S can be prevented (or H2S removed) at tions, solid feedstock samples were thawed and ground in a meat
different levels (i) during the anaerobic process itself, (ii) by treat- grinder (Kenwood, Pro 1600) to make them homogenous. For li-
ing the biogas or (iii) by controlling feedstocks. quid samples, homogenisation was done with a lab stirrer.
In anaerobic digesters, the harmful action of sulphide in the
reactor can be controlled by adding chemical compounds such as 2.2. Biochemical methanogenic potential and biochemical biogas
specific inhibitors of sulphide producing microorganisms (Isa and potential (BBP)
Anderson, 2005) or sulphide scavengers, mainly metal ions, to pre-
cipitate the sulphide. Due to the high technicality of these manip- BMP (NL CH4/kg volatile solids (VS)) of each substrate was
ulations and the high cost of the products used, only the oil determined according to Vedrenne et al. (2008) and the standard
industry and wastewater treatment plants use these techniques. NF EN ISO 11734 (AFNOR, 1998). BMP methodology is based on
In most anaerobic digesters, sulphide is removed from the bio- measuring the pressure generated by biogas production in a closed
gas using physical–chemical or biological techniques. Different bottle at a fixed volume. This procedure is realised in mesophilic
traps are used such as dry removal processes with metal oxides (38 °C) anaerobic conditions in the presence of substrate and of
like iron and zinc oxides, alkaline solids (hydrated lime), adsor- an inoculum. The sample was diluted with a nutritive solution to
bents (molecular sieves, activated carbon) or acid washing (Cirne avoid inhibition and to maintain a fixed total liquid volume be-
et al., 2008). A simple biological technique is to create micro-aero- tween the tests. Biogas was collected and its methane (CH4) and
bic conditions in the gasometer by adding 2–6% air to the biogas. carbon dioxide (CO2) contents were determined by gas chromatog-
The presence of oxygen encourages the growth of chemoautotroph raphy (Lucas et al., 2007). BMP was performed in triplicate with
microorganisms like the genus Thiobacillus, which oxidises hydro- controls (inoculum only) until biogas production ceased. BBP
gen sulphide into elemental S and sulphate (SO42) (Cirne et al., (NL Biogas/kg feedstock (fd)) was also measured during these tests
2008; Diaz et al., 2010). Under oxygen limiting conditions, the ma- by cumulating biogas production for each test in normal
jor end product is elemental S while in fully oxygenated conditions conditions.
H2S is completely oxidised into SO42. This biological process is of- The inoculum used was obtained from a well established
ten described in the literature and on German farms, most full anaerobic pilot plant (100 L) acclimated to degrade pig slurry sup-
scale anaerobic reactors uses this desulphurisation technique plemented with horse feed as co-substrate (mixture of lignocellu-
(Cirne et al., 2008). lose materials 18%, proteins 12% and lipids 2.5%) to obtain an
However, biological desulphurisation does have some organic loading rate of 2 g VS/L/d (Peu et al., 2011). The anaerobic
limitations: sludge produced by this digester has been shown to degrade sev-
eral different organic materials and is a satisfactory medium to
(1) After desulphurisation, the concentration of residual H2S study anaerobic degradation of a wide range of feedstocks (data
may still be higher than the manufacturer’s recommenda- not shown). To maximise the biodegradation rate and ensure that
tions for combined heat and power equipment. For efficient the methane potential was achieved, a VS inoculum:substrate ratio
biological desulphurisation, the concentration of H2S in the between 0.5 and 3 was applied in accordance with the results of
raw biogas cannot exceed 1.5–2% (Cirne et al., 2008). previous studies (Chynoweth et al., 1998; Labatut et al., 2011).
(2) The amount of air and particularly of oxygen added to the
biogas should be limited to avoid explosive gas mixtures
and biogas dilution to ensure satisfactory biogas combustion 2.3. Physical–chemical characterisation
properties and a good methane number (Rahmouni et al.,
2003). Total solids (TS), VS, total Kjeldahl nitrogen (TKN) and total
(3) H2S production from the digester should be stable to ensure ammoniacal nitrogen (TAN) were measured using standard meth-
the efficient removal of H2S since oxidation is based on a ods (APHA, 1998).
biological process involving microorganisms with limited
growth efficiency. 2.3.1. Carbon and sulphur elementary analysis
Elementary analysis of carbon (C) and S was performed using
For these reasons, to design an efficient H2S removal process, specific analysers according to the manufacturer’s instructions
knowledge of the capacity of the substrates to produce H2S during (Thermo Flash 2000 and LECO SC-144DR, respectively). For C deter-
anaerobic digestion is required. mination, samples were homogenised and mixed with alumina
according to their TS content. Samples were then placed in a tin
capsule in a combustion furnace heated to 1000 °C. During com-
1.1. Purpose
bustion, an oxygen flux oxidised all forms of C into CO2. The gas
stream was then analysed by chromatography with catharometric
The main objective of this study was to predict the H2S content
detection. Calibration was done using a standard curve made of
of raw biogas (predictable H2S–S biogas content) produced by dif-
different levels of aspartic acid. Similar procedures were used for
ferent feedstocks that could be used in anaerobic farm digesters.
S determination. The samples were first dried at 105 °C and then
ground with a mortar. An aliquot was placed in a basket and cov-
2. Methods ered with vanadium pentoxide. The basket containing the mixture
was heated in the furnace at 1350 °C under a continuous flow of
2.1. Feedstocks pure oxygen causing oxidation of all S forms into sulphur dioxide.
The combustion gas and particularly sulphur dioxide content was
Thirty-seven potential feedstocks were sampled from urban analysed by an internal infrared detector to determine the S con-
wastewater treatment plants, farms, agri-food facilities and muni- tent of the samples. Specific calibration was performed using
 P. Peu et al. / Bioresource Technology 121 (2012) 419–424 421

Table 1
Origin of feedstocks.

Urban wastewater Agriculture Agri-food facilities Municipal waste

treatment plant
Cover crops Animal waste Industrial wastewater Fatty effluents Vegetable Others
treatment residues
Primary sludge Camelina crops Farm manure Biological sludge Grease trap waste Carrot pulp Rumen Restaurant refusals
(n = 3) (n = 3) content
Biological sludge White mustard Dairy cow slurry Screening refusal Onion pulp Meat Dehydrated
crops (n = 2) (n = 2) waste restaurant refusals
DAF sludge* Brown mustard Pig faeces Shallot pulp Bovine Harvested red
crops (n = 2) blood seaweeds
Radish crops Pig slurry Pig mucus Harvested green
(n = 2) seaweeds
Soybean crops Pig bristles Lawn mowing
Fish
viscera
*
Dissolved air flotation sludge.

known amounts of S from standard coal (LECO, USA). The C:S ratios content was calculated at 228 g/kg fd ± 175. This high variation is
were determined for each feedstock analysed. due to the wide range of sample types. TS values for biological
sludges, pig and cattle slurries are often low and were below
2.3.2. Calculation of crude proteins (CP) content, carbon 69 g/kg fd in this study. These low TS levels are an advantage for
biodegradability (CB) and predictable H2S–S biogas content (PSB) farms that use liquid manure as the main substrate for continuous
The contribution of each feedstock to CP (Eq. (1)) was calculated stirring co-digestion; and a disadvantage from an energy point of
using its nitrogen content (i.e. TKN and TAN) (Dintzis et al., 1988): view (Mata-Alvarez et al., 2000). The average VS level was
81% ± 16 of TS, giving an average TS:VS ratio of 1.3.
CP ðg=kg TSÞ ¼ ðTKN  TANÞ  6:25 ð1Þ

with TKN and TAN, Kjeldahl and ammoniacal nitrogen content

respectively (g N/kg TS) and 6.25, average ratio of protein N in a 3.2. Sulphur content and speciation
typical protein (g N/g pure protein).
Carbon biodegradability was calculated (Eq. (2)) with a C bal- In all feedstocks analysed, the total S content varied between 1
ance between C recovered in the biogas and cumulated during and 29.6 g S/kg TS (Table 2). Minimum values were obtained for
the BMP test and the C content of the feedstock before the anaer- grease trap waste from cattle slaughterhouse and maximum for
obic test. harvested green seaweed. Three groups could be distinguished,
one with values >20 g S/kg TS (harvested algae and pig bristles),
CO2  C þ CH4  C another group with intermediate values ranging from 5.5 to
CB ð%Þ ¼  100 ð2Þ
C  TS 8.7 g S/kg TS (some animal wastes and biological sludge), and a
third group with an S content <5 g S/kg TS (the remaining feed-
with CO2–C + CH4–C, cumulated C recovered in BBP (g C/kg fd); C,
stocks). The S content of our set of potential feedstocks was low,
feedstock total C content (g C/kg TS); TS, feedstock total solid con-
as more than 68% had a sulphur content <5 g S/kg TS.
tent (g/kg fd)
To complete this analysis, CP content was determined using
PSB is the inherent ability of a feedstock to produce H2S during
coupled NTK and TAN analyses (Table 2). Sulphur amino acid con-
anaerobic digestion. Predicted biogas H2S–S content in normal con-
tent (SAA–S) of the feedstocks was determined using CP and SAA–S
ditions was determined using two distinct calculations, one using
content of an ideal protein with a constant CP:SAA–S ratio of 79
total S analysis and biogas production (PSBmax, Eq. (3)), and the
(Sriperm et al., 2010).
other (PSBbio, Eq. (4)) by determining a molar ratio between total
Estimated S content versus CP content is presented in Fig. 1.
S and C considering that biodegradability of S was similar to carbon
Fourteen feedstocks (39%) are close to the ideal CP:SAA–S ratio un-
biodegradability (CB).
der 30% variation (see dotted line in Fig. 1) suggesting that their
ððS  TSÞ=1000 =32Þ  22:4 main S content can be explained by their SAA–S content. For all
PSBmax ð%Þ ¼  100 ð3Þ
BBP other feedstocks, different explanations are required. In the har-
S=32 vested seaweeds, the S content was as high as their nitrogen con-
PSBbio ð%Þ ¼  100 ð4Þ
C =12 tent and accounted for 2.9% and 2.3% of their dry matter
respectively. This high S content is directly linked with the pres-
with S, feedstock total S content (g S /kg TS); TS, feedstock total so- ence of residual seawater, which has a high concentration of sul-
lid content (g/kg fd); BBP, biochemical biogas potential (NL Biogas/ phate (0.9 g SO42––S/L) and to the particular chemical
kg fd); 12, 32 and 22.4, respectively denotes C and S molecular composition of algae that contain a large amount of sulphur poly-
weight (g/mol) and molar volume in normal condition for temper- saccharides (Lahaye and Robic, 2007). In animal wastes, and partic-
ature and pressure (L/mol). ularly in pig slurry, metal sulphides (chalcocite, sphalerite, pyrite,
etc.) are present in quantities that increase the total S content. This
3. Results and discussion presence is the consequence of H2S production during anaerobic
storage of animal wastes coupled to the presence of metals in fresh
3.1. Characteristics of feedstocks manure which interact together to form metal sulphides (Legros
et al., 2010). Among catch crops and crops residues, cruciferous
Table 2 lists TS and VS values of feedstocks on a wet basis. TS plants contain significant amounts of glucosinolate compounds at
content in all the samples ranged from 31 g/kg fd in dairy cow slur- concentrations that range between 6 and 200 lmol/g of fresh
ry to 975.4 g/kg fd for dehydrated restaurant refusals. Average TS material (Tripathi and Mishra, 2007). Shallot and onion pulps also
422 P. Peu et al. / Bioresource Technology 121 (2012) 419–424

Table 2
Feedstocks characteristics and predictable H2S–S biogas content.

TS VS Total S Total C CP BMP BBP CB (%) C/S PBSmax PBSbio

(g/kg fd) (g/kg fd) (g S/kg TS) (g C/kg TS) (g/kg TS) (NL CH4/kg VS) (NL Biogas/kg fd) (g/g) (%) (%)
Feedstock from urban wastewater treatment plant
Primary sludge 52.8 44.5 4.5 339.7 nd* nd nd nd 74.7 – –
Biological sludge 57.0 45.8 8.6 505.3 493.4 255.5 18.6 34.6 58.6 1.9 0.6
DAF sludge 47.6 36.3 8.4 576.5 154.0 806.1 39.4 76.9 68.7 0.7 0.5
Feedstock from agricultural resources
Cover crops
Camelina crops 213.0 178.0 4.0 394.4 132.9 234.3 84.2 53.7 98.6 0.7 0.4
White mustard crops 228.0 194.0 3.3 390.4 100.9 222.5 86.4 52.0 117.6 0.6 0.3
Brown mustard crops 173.0 154.0 3.7 514.5 98.3 264.1 87.5 52.7 137.9 0.5 0.3
Radish crops 301.0 126.0 3.4 245.8 62.3 236.7 67.6 49.0 73.2 1.0 0.5
Soybean crops 249.0 84.0 2.4 152.6 78.8 305.0 58.2 82.0 63.3 0.7 0.6
Animal wastes
Farm manure 225.2 209.9 1.2 297.3 71.3 271.5 103.0 82.4 245.3 0.2 0.2
Dairy cow slurry (1) 31.0 26.0 2.8 477.4 141.1 326.9 13.4 48.5 170.5 0.5 0.2
Dairy cow slurry (2) 106.9 79.2 3.8 429.3 140.9 238.7 28.3 33.0 112.7 1.0 0.3
Pig faeces (1) 362.0 280.0 3.1 360.0 147.8 304.0 135.1 55.5 115.4 0.6 0.3
Pig faeces (2) 390.0 320.0 3.9 398.0 124.8 328.0 166.6 57.5 102.4 0.6 0.4
Pig slurry (1) 43.2 30.4 8.0 349.3 172.2 230.0 9.8 34.8 43.6 2.5 0.9
Pig slurry (2) 38.5 25.0 8.0 411.9 204.0 140.2 5.5 18.6 51.5 3.9 0.7
Feedstock from agri-food facilities
Industrial wastewater treatments
Biological sludge PS** 60.3 47.8 8.3 553.0 130.3 259.5 18.0 28.9 66.6 1.9 0.6
Biological sludge DMI*** 69.2 40.0 7.0 319.4 445.2 282.8 16.9 41.0 45.8 2.0 0.8
Biological sludge PI**** 61.6 41.3 4.7 348.4 456.8 205.9 13.7 34.2 74.2 1.5 0.5
Screening refusal PS 328.6 310.8 5.1 500.9 338.7 545.7 247.7 80.6 97.9 0.5 0.4
*****
Screening refusal CS 176.2 163.2 2.8 531.4 227.5 490.2 119.2 68.2 189.5 0.3 0.2
Fatty effluents
Grease trap waste CS 237.7 231.6 1.0 765.9 42.6 863.4 271.7 79.9 797.9 0.1 0.0
Grease trap waste PI 361.3 350.0 1.5 702.3 67.7 1011.5 454.8 96.0 478.9 0.1 0.1
Grease trap waste PS 557.1 552.6 1.5 609.2 102.5 902.9 648.3 102.3 412.0 0.1 0.1
Vegetable residues
Carrot pulp 177.1 166.7 1.1 447.2 93.1 330.6 107.7 72.8 389.0 0.1 0.1
Onion pulp 211.5 194.6 3.0 434.5 122.5 416.7 153.1 89.2 142.6 0.3 0.3
Shallot pulp 226.7 212.7 3.4 456.0 144.6 413.3 169.9 88.0 134.8 0.3 0.3
Others
Rumen content CS 166.3 155.4 1.5 493.6 124.2 341.1 82.7 54.0 325.6 0.2 0.1
Meat waste PI 330.3 293.9 4.9 626.5 332.2 871.0 346.9 89.8 127.3 0.3 0.3
Bovine blood CS 155.8 145.1 6.8 505.1 925.0 454.9 92.3 62.8 74.5 0.8 0.5
Pig mucus PS 190.5 169.1 7.7 526.0 695.1 579.5 141.6 75.7 68.6 0.7 0.5
Pig bristles PS 285.5 278.4 26.0 492.5 904.2 277.3 105.5 40.2 18.9 4.9 2.0
Fish viscera 418.0 380.0 4.6 568.6 269.6 657.9 403.0 90.8 122.6 0.3 0.3
Feedstock from municipal waste
Canteen refusals 271.3 257.5 3.6 545.1 244.2 570.8 249.6 90.4 151.4 0.3 0.2
Dehydrated Canteen refusals 975.4 827.1 4.0 464.2 305.0 395.3 535.9 63.4 116.2 0.5 0.3
Harvested red seaweeds 200.9 120.2 23.3 276.2 136.9 133.1 25.6 24.7 11.9 12.8 3.2
Harvested green seaweeds 222.0 128.0 29.6 203.4 141.0 132.8 26.0 30.8 6.9 17.7 5.5
Lawn mowing 238.2 187.5 5.5 438.2 220.4 289.6 86.7 44.5 79.2 1.1 0.5
*
Not determined.
**
Pig slaughterhouse.
***
Dairy milk industry.
****
Pig industry.
*****
Cattle slaughterhouse.

contain allicin compounds (Asili et al., 2010) (0.7–3 g/kg of fresh 3.3. Biochemical methanogenic potential and carbon biodegradability
material), which contain S atoms in their molecular structure. Pig
bristles sampled in a piggery slaughterhouse had a high CP content Table 2 shows the results of the individual BMP assays per-
(904 g/kg TS) along with a high S content (26 g S /kg TS). This par- formed on the 37 feedstocks. The feedstocks were classified in
ticularity can be explained by the presence of keratin, a rich sul- three groups according to the methane yields obtained: one group
phur crude protein that composes hair, bristles, wool and fur with fatty substrates (grease trap waste, fish viscera, DAF sludge
(Marshall et al., 1991). Elemental S analyses of pure keratin re- and meat wastes) with values ranging between 657 and
vealed that S represented close to 5% of its TS. In the present study, 1011 NL CH4/kg VS, another group in which BMP ranged from 330
collected pig bristles waste was mixed with other residues (skin to 579 NL CH4/kg VS which included screening and restaurant
and fat) and the resulting S content was below values usually ob- refusals, vegetable residues and rumen content; and a third group,
tained for keratin. Bovine blood is mainly composed of serum comprising the remaining feedstocks, having lower BMP contents
and cells which contain high CP like albumin and globulin. Surpris- (under 300 NL CH4/kg VS).
ingly, despite the fact that this feedstock is mainly composed of CB was also estimated for all the samples analysed: in the first
proteins, no direct relationship was found between CP and SAA–S. group it represented an average of 87% of the carbon input. In the
 P. Peu et al. / Bioresource Technology 121 (2012) 419–424 423

Fig. 1. Total S versus protein content (CP) of different feedstock. Solid line represents the total sulphur amino acid content (SAA–S) for a typical protein with 30% variation
(dotted lines).

second and the third groups, the average values were 77% and 43%, the hypothesis that biodegradability of S is similar to C biodegrad-
respectively. Substrates with high lipid content and high CB, such ability (CB). The second restrictive calculation was run to deter-
as grease trap waste, had higher methane yields. In contrast, feed- mine a lower range of PSB. Values for these potentials (PSBmax
stocks with high lignocellulosic content combined with low CB, and PSBbio) are listed in Table 2. Predicted H2S-S biogas content
such as harvested algae, had a lower methane yield. These results ranged from 0.1 to more than 17% and from 0% to 5.5% for PSBmax
are in agreement with previously published findings (Labatut et al., and PSBbio respectively. In both calculations, the lowest values
2011). Results of the BMP assays reported in this study are in were obtained for fatty effluents and the highest values for har-
accordance with previous data published in the literature (El-Ma- vested seaweeds. For several feedstocks, including biological
shad and Zhang, 2010; Labatut et al., 2011; Vedrenne et al., sludge, pig slurries, pig bristles and harvested seaweeds, the pre-
2008; Ward et al., 2008; Zhu et al., 2011). dicted H2S content calculated with PSBmax ranged between 1%
and 3.9%, whereas values obtained with PSBbio were lower, espe-
3.4. Predictable H2S–S biogas content cially for biological sludge and pig slurry (<1%).
Using these data, it is possible to build predictive models directly
With S content and biogas production, a predictive hydrogen linking the C:S ratios of a feedstock to its PSB content (Fig. 2). In our
sulphide concentration (% v:v) in the biogas named ‘‘predictable dataset, an H2S content of raw biogas of <2% (i.e. threshold required
H2S–S biogas content’’ can be calculated for each feedstock used for efficient biological desulphurisation) was correlated with sub-
as a single substrate in anaerobic digestion processes. Indeed two strates with a C:S ratio >40 (PSBmax). This predictive model C:S
distinct PSB can be determined. One takes total S into account un- vs. PSBmax allows H2S content in the biogas to be predicted as a
der the hypothesis that the entire S content is transformed into dis- security indicator, by analysing the C and S contents of feedstocks
solved sulphides and totally transferred to the biogas. This without determination of BMP and biogas production.
potential is considered as the maximum value for H2S–S biogas Models relating the ratios C:S vs. PSBmax and PSBbio have been
concentration (PSBmax). The second potential (PSBbio) is deter- tested with results published in 2011 (Peu et al., 2011). In this pa-
mined using a molar ratio between total S and total C based on per, authors tested a mixture of pig slurry with stranded seaweed

Fig. 2. Predictable H2S–S biogas content of feedstock as a function of the C:S ratios. PSBmax ( ), PSBbio (dotted line).
424 P. Peu et al. / Bioresource Technology 121 (2012) 419–424

Table 3 Asili, A., Behravan, J., Reza Naghavi, M., Asili, J., 2010. Genetic diversity of persian
Biogas H2S content prediction from published data. shallot (Allium hirtifolium) ecotypes based on morphological traits, allicin
content and RAPD markers. Open Access Journal of Medicinal and Aromatic
Substrates* C:S* Predicted H2S–S biogas content Plants 1 (1), 1–6.
Beline, F., Dabert, P., Peu, P., Girault, R., 2010. Methane production from farm
PBSmax PBSbio Measured
effluents in France and Europe: principle, inventory and, prospects. Fourrages
(%) (%) (%)*
(203), 155–161.
Mixture pig slurry and stranded 13.5 9.1 2.8 3.2 Chynoweth, D.P., Wilkie, A.C., Owens, J.M., 1998. Anaerobic Processing of Piggery
seaweed Slurry: A Review. Eight World Conference on Animal Production, 28 juin–4
Pig slurry 52.2 1.5 0.7 0.2 juillet, Seoul, Korea. 39pp.
Cirne, D.G., van der Zee, F.P., Fernandez-Polanco, M., Fernandez-Polanco, F., 2008.
* Control of sulphide during anaerobic treatment of S-containing wastewaters by
Issued from Peu et al. (2011).
adding limited amounts of oxygen or nitrate. Reviews in Environmental Science
and Bio/Technology 7 (2), 93–105.
Diaz, I., Lopes, A.C., Perez, S.I., Fdz-Polanco, M., 2010. Performance evaluation of
and pig slurry as substrates in laboratory anaerobic digesters. In oxygen, air and nitrate for the microaerobic removal of hydrogen sulphide in
their tests, they measured biogas H2S production in connection biogas from sludge digestion. Bioresource Technology 101 (20), 7724–7730.
Dintzis, F.R., Cavins, J.F., Graf, E., Stahly, T., 1988. Nitrogen-to-protein conversion
with these substrates. Total C and S were also analysed for each factors in animal feed and fecal samples. Journal of Animal Science 66 (1), 5–11.
substrates. Extracted data from this article are presented in Table 3 El-Mashad, H.M., Zhang, R., 2010. Biogas production from co-digestion of dairy
versus predictions of H2S biogas content with PSBmax and PSBbio. manure and food waste. Bioresource Technology 101 (11), 4021–4028.
Elferink, S., Visser, A., Pol, L.W.H., Stams, A.J.M., 1994. Sulfate Reduction in
Based on results published by Peu et al. (2011), predicting H2S–
Methanogenic Bioreactors. Fems Microbiology Reviews 15 (2–3), 119–136.
S content of biogas with C:S vs. PSBmax clearly overestimates H2S Isa, M.H., Anderson, G.K., 2005. Molybdate inhibition of sulphate reduction in two-
concentration. This predicting model is a security indicator based phase anaerobic digestion. Process Biochemistry 40 (6), 2079–2089.
Labatut, R.A., Angenent, L.T., Scott, N.R., 2011. Biochemical methane potential and
only on total sulphur loading. With the predicting model C:S vs.
biodegradability of complex organic substrates. Bioresource Technology 102
PSBbio, theoretical and measured values are much closer and re- (3), 2255–2264.
flect a more realistic approach. However for pig slurry, the differ- Lahaye, M., Robic, A., 2007. Structure and functional properties of Ulvan, a
ence observed between theoretical and measured values suggests polysaccharide from green seaweeds. Biomacromolecules 8 (6), 1765–1774.
Legros, S., Doelsch, E., Chaurand, P., Rose, J., Masion, A., Borschneck, D., Proux, O.,
that biodegradability of the sulphur content in these waste is lim- Hazemann, J.L., Briois, V., Ferrasse, J.H., Saint Macary, H., Bottero, J.Y., 2010.
ited by the fact that S is primarily precipitated or bound (Legros Investigation of copper and zinc speciation in pig slurry by a multitechnique
et al., 2010), making it unavailable for further biochemical approach RAMIRAN, Treatment and use of organic residues in agriculture:
Challenges and opportunities towards sustainable management. FAO, Lisbonne,
transformations. pp. 130.
Lucas, T., Le Ray, D., Peu, P., Wagner, M., Picard, S., 2007. A new method for
continuous assessment of CO2 released from dough baked in ventilated ovens.
4. Conclusion
Journal of Food Engineering 81 (1), 1–11.
Mackie, R.I., Stroot, P.G., Varel, V.H., 1998. Biochemical identification and biological
This inventory enabled various feedstocks that could be used in origin of key odor components in livestock waste. Journal of Animal Science 76
agricultural anaerobic digestion to be classified according to their (5), 1331–1342.
Marshall, R.C., Orwin, D.F.G., Gillespie, J.M., 1991. Structure and biochemistry of
sulphur and carbon contents. Our results showed that the S con- mammalian hard keratin. Electron Microscopy Reviews 4 (1), 47–83.
tent of most feedstocks is low. The key achievement of this study Mata-Alvarez, J., Macé, S., Llabrés, P., 2000. Anaerobic digestion of organic solid
was to couple feedstock C:S ratio to a predictable H2S–S content wastes. An overview of research achievements and perspectives. Bioresource
Technology 74 (1), 3–16.
of biogas. The main conclusions are that the fraction of sulphur Peu, P., Sassi, J.F., Girault, R., Picard, S., Saint-Cast, P., Béline, F., Dabert, P., 2011.
available for reduction into H2S is proportional to the biodegrad- Sulphur fate and anaerobic biodegradation potential during co-digestion of
able fraction of carbon, and, once reduced, the phase transfer of seaweed biomass (Ulva sp.) with pig slurry. Bioresource Technology 102 (23),
10794–10802.
the H2S generated is nonlimiting. Rahmouni, C., Le Corre, O., Tazerout, M., 2003. Online determination of natural gas
properties. Comptes Rendus Mecanique 331 (8), 545–550.
Acknowledgement Sriperm, N., Pesti, G.M., Tillman, P.B., 2010. The distribution of crude protein and
amino acid content in maize grain and soybean meal. Animal Feed Science and
Technology 159 (3–4), 131–137.
This research was supported by a Grant from the ‘‘French Envi- Tripathi, M.K., Mishra, A.S., 2007. Glucosinolates in animal nutrition: A review.
ronment and Energy Agency, ADEME’’ (Project No. 07 06 C 0039). Animal Feed Science and Technology 132 (1–2), 1–27.
Vedrenne, F., Béline, F., Dabert, P., Bernet, N., 2008. The effect of incubation
conditions on the laboratory measurement of the methane producing capacity
References of livestock wastes. Bioresource Technology 99 (1), 146–155.
Ward, A.J., Hobbs, P.J., Holliman, P.J., Jones, D.L., 2008. Optimisation of the anaerobic
AFNOR. 1998. NF EN ISO 11734 – Qualité de l’eau – Evaluation de la digestion of agricultural resources. Bioresource Technology 99 (17), 7928–7940.
biodégradabilité anaérobie ‘‘ultime’’ des composés organiques dans les boues Wellinger, A., Linberg, A., 2000. Biogas Upgrading and Utilization. France, Paris.
de digesteurs – Méthode par mesurage de la production de biogaz. Zhu, Z., Hsueh, M.K., He, Q., 2011. Enhancing biomethanation of municipal waste
APHA, 1998. Standards Methods for the Examination of Water and Wastewater, sludge with grease trap waste as a co-substrate. Renewable Energy 36 (6),
20th ed. APHA, Washington , DC. 1802–1807.