Chemosphere 153 (2016) 48e57

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Determination of volatile organic compounds from biowaste and cofermentation biogas plants by single-sorbent adsorption

mez*, H. Lohmann, J. Krassowski
J.I. Salazar Go
Fraunhofer Institute for Environmental, Safety, and Energy Technology (UMSICHT), Osterfelder Strasse 3, 46047 Oberhausen, Germany

h i g h l i g h t s

A correlation between TVOCs and substrate composition and feeding amount was determined.

Longer residence times in the digester appear to reduce the emissions of VOCs.

Terpenes and ketones were found to be the predominant VOCs in all three biogas plants.

In situ desulphurisation reduces the amounts of hydrogen sulphide but shows little inuence on sulphides and thiols.

A simple sampling method allowed the quantication of a broad range of VOCs.

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 18 March 2015
Received in revised form
12 January 2016
Accepted 29 February 2016
Available online 21 March 2016

Characterisation of biogases is normally dedicated to the online monitoring of the major components
methane and carbon dioxide and, to a lesser extent, to the determination of ammonia and hydrogen
sulphide. For the case of Volatile Organic Compounds (VOCs), much less attention is usually paid, since
such compounds are normally removed during gas conditioning and with exception of sulphur compounds and siloxanes represent a rather low risk to conventional downstream devices but could be a
hindrance for fuel cells. However, there is very little information in the literature about the type of
substances found in biogases generated from biowaste or co-fermentation plants and their concentration
uctuations. The main aim of this study was to provide information about the time dependencies of the
VOCs in three biogas plants spread out through Germany from autumn until summer, which have
different process control, in order to assess their potential as biofuels. Additionally, this study was an
attempt to establish a correlation between the nature of the substrates used in the biogas plants and the
composition of the VOCs present in the gas phase. Signicant time-dependent variations in concentration were observed for most VOCs but only small changes in composition were observed. In general,
terpenes and ketones appeared as the predominant VOCs in biogas. Although for substances such as
esters, sulphur-organic compounds and siloxanes the average concentrations observed were rather low,
they exhibited signicant concentration peaks. The second biogas plant which operates with dry
fermentation was found to contain the highest levels of VOCs. The amount of total volatile organic
compounds (TVOCs) for the rst, second and third biogas plants ranged from 35 to 259 mg Nm3, 291
e1731 mg Nm3 and 84e528 mg Nm3, respectively.
2016 Elsevier Ltd. All rights reserved.

Handling Editor: Jun Huang

Keywords:
Biogas
Volatile organic compounds (VOCs)
Tenax
Monitoring
Thermodesorption

1. Introduction
The utilisation of biogenic gases such as, landll gas (LFG)
(Dernbach, 1985; Franzius, 1985; Aitchison, 1996), sewage gas

* Corresponding author.

mez), heiko.lohmann@
E-mail addresses: jsalazar@gmx.net (J.I. Salazar Go
umsicht.fraunhofer.de (H. Lohmann), joachim.krassowski@umsicht.fraunhofer.de
(J. Krassowski).
http://dx.doi.org/10.1016/j.chemosphere.2016.02.128
0045-6535/ 2016 Elsevier Ltd. All rights reserved.

(Friedrich et al., 2000; Rossol et al., 2003; Tsagarakis, 2006) or farm

biogas from energy crops or agricultural wastes (biowastes) (Raven
& Gregersen, 2007; Weiland, 2010; Murphy & McCarthy, 2005) as
renewable sources of energy has become a common practice in the
European context and has been in recent years increasingly supported by the introduction of incentives and subsidies (Weiland,
2000, 2003; Grabmann, 2011). The political decision taken by the
German government in June 2011 to back out completely of the
nuclear energy programme by the year 2022, decision mainly
motivated by the unfortunate natural disaster in Fukushima, Japan,


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

has enormously pushed on the utilisation of biogas in particular

from agricultural sources. This decision supports the modern trend
of a more decentralised power supply system and the search for
alternative sources of energy in order to improve the current power
and heat efciencies (Staffell et al., 2008), which will help to
mitigate the dependency to non-renewable sources of energy (e.g.
coal and oil). According to the German Biogas Association, in the
year 2013 the number of biogas plants in Germany accounted for
7850 plants, which generated around 3543 MW electricity from
farm biogas. This helped to mitigate the emission of about 16,8
million tonnes of CO2 (German Biogas Association). Based on data
of the German Ministry of Environment (BMU), in 2010 the savings
of CO2 emissions through the utilisation of sewage and landll gas
were around 1,11 million tonnes and 0,65 million tonnes of CO2,
respectively (German Ministry of Environment). Data from the
European Biogas Association (EBA) show that, in the European
scenario, Germany appears as the main biogas producer followed
by the UK (European Biogas Association).
1.1. Biogas production and composition
The term Biogas is widely understood as the gas produced by
the decomposition of organic matter under anaerobic conditions
where the organic materials undergo basically four steps: hydrolysis, acidogenesis, acetogenesis and methanogenesis (Appels et al.,
2008). Hydrolysis is generally regarded as the rate-limiting step
(Scherer, 2001) since high-molecular weight molecules e.g. polysaccharides, lipids and proteins have to be degraded to smaller
compounds. Fig. 1 shows the widely accepted sequence of anaerobic degradation processes.
Methane and carbon dioxide appear as the main constituents of
the different biogenic gases, with concentrations ranging between
40-60% and 40e55% (Rasi et al., 2010), respectively. For the case of
LFG, methane and CO2 concentrations depend strongly on the age
of the landll (anaerobic methanogenic steady phase) (Farquhar &
Rovers, 1973), the composition of the waste and partly on the type
of the collection system. Moreover, among all biogenic gases LFG
appears as the most contaminated gas due to the higher number of
harmful VOCs (e.g. halogenated compounds), followed by sewage
gas, whereas farm biogas seems to be the less contaminated one
(Rasi et al., 2007). Apart from the major components, farm biogas

49

exhibits low concentrations of ammonia and is saturated with

water vapour but may also contain relatively high amounts of
hydrogen sulphide (H2S). Depending on the substrates used it may
exhibit some similar VOCs as in landll and sewage gas such as,
ketones, alcohols, terpenes, alkanes, esters, furans, mercaptanes,
sulphides and aromatics (Smet et al., 1999). Halogenated compounds and siloxanes are in principle not expected in farm biogas.
As pointed out by Rasi et al. (2011) their concentrations in farm
biogas are apparently very low and their presence in biogas would
depend on the nature of the feedstock used for co-fermentation.
Farm biogas from exclusively energy crops may exhibit H2S concentrations as high as 2000 ppm. In contrast, farm biogas from
biowastes may exhibit H2S concentrations even over 20,000 ppm.
The presence of trace compounds in LFG and sewage gas is well
documented in the literature but for farm biogas there is little information about the type of VOCs that can be found, their concentration ranges and possible seasonal uctuations. Therefore, it is
the aim of this article to show the results gathered in a research
project. The main goal of the present study was to determine the
concentration and the seasonal variations of typical VOCs present
in the biogas of three different biogas plants spread out throughout
Germany, which are fed with very different substrates in order to
help make decisions about adequate upgrading processes.
2. Materials and methods
2.1. Biogas plant description
For the determination of VOCs in biogas, three different biogas
plants (BP) were selected and monitored for a period over six
months (from autumn until summer) in order to check for eventual
concentration peaks, since some substrates commonly used in
biogas plants are season dependent. The biogas plants are located
in different federal states throughout Germany in order to avoid
similar composition of the substrates used and differ in their
technical design and process control. The rst biogas plant is
located in the federal state of North Rhine-Westphalia. The
composition of the substrates used at this site comprises typically
45% glycerine, about 40% of fatty sludge from the biodiesel production and 15% cattle manure.
The second biogas plant is located in the federal state of Saxony-

Fig. 1. The anaerobic degradation process after (Bischoffsberger & et al., 2005).


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

50
Table 1
Key operational parameters of the three biogas plants.
Characteristics

Biogas plant 1

Biogas plant 2

Biogas plant 3

Water content
Temperature
Process control
Mixing type

wet fermentation
mesophilic
single-stage
Fermenter 1: Gas injection
Fermenter 2: Paddle mixers
Fermenter 3: Central mixer
Co-fermentation
14,000 t/a

dry fermentation
thermophilic
two-stage
Fermenter 1: Plug ow
Fermenter 2: Plug ow
e
Waste fermentation
30,000 t/a

wet fermentation
mesophilic
single-stage
Fermenter 1: Pumping, heat exchanger, gas injection
Fermenter 2: Pumping, heat exchanger, gas injection
e
Waste fermentation
8000 t/a

Feedstock
Operational capacity

Anhalt. The substrate used for the production of biogas consists

mainly of biowaste produced in residential areas. The composition
of the biowaste used as substrate at this site normally contains 87%
biowaste, 10% green and branch cuttings from gardens and ca. 3%
husk.
The third biogas plant is located in the free state of Bavaria. At
this site the substrate composition is very broad and the percentages of the different biowastes change continuously. Table 1 summarises the key characteristics of the three biogas plants.

(here 2-Bed or 3-Bed tubes) were also ordered by Markes International Ltd as an attempt to capture all VOCs in biogas, the ndings

mez et
with this type of tubes will be shown elsewhere (Salazar Go
al., submitted). The 2-Bed tubes contained apart from Tenax TA a
strong carbon molecular sieve (UniCarb, mesh 60/80), which
exhibits a surface area of ca. 1200 m2/g 3-Bed tubes contained three
sorbent beds (Tenax TA, UniCarb and Carboxen 1003). These
thermodesorption tubes were chosen due to relatively low levels of
artefacts.

2.2. Analysis of CH4, CO2, O2, N2 and H2S

Even though measurements of the major constituents in biogas
were carried out using a biogas analyser (Biogasmonitor BM2000)
of the company Ansyco GmbH, which measures selective wavelengths of CH4 and CO2 in the infrared region, the values of the
concentrations of methane and carbon dioxide were normally obtained from the biogas plant operators. For the measurement of
oxygen and hydrogen sulphide, the gas analyser was equipped with
two electrochemical cells. The determination of nitrogen was made
as the rest gas.
2.3. Sorbent tubes for sampling and thermodesorption
Commercial inert coated stainless steel TD sampling tubes (300
length, 00 internal diameter), pre-packed with ca. 200 mg Tenax TA
(mesh 35/60, ca. 35 m2/g) were purchased from Markes International Ltd (UK) and used for the determination of VOCs. Additionally, certied multibed inert coated stainless steel adsorption tubes

2.4. Sampling of biogas

In comparison to conventional sampling methods for the
determination of VOCs in indoor or outdoor air, the sampling of
biogases appears as a challenging task, since biogases normally
exhibit very high humidity, which can seriously interfere with the
determination of trace compounds due to a signicant decrease in
the breakthrough times (Shin et al., 2002). Because of the intrinsic
water problematic, a hydrophobic adsorbent like Tenax TA (2,6
diphenylene oxide) was chosen among several commercial sorbents owing to its known qualities by the adsorption of VOCs
(Maier & Fieber, 1998) such as, high hydrophobicity, thermal stability and low production of artefacts (Woolfenden, 2010). In
contrast, carbon-based sorbents have been found to show some


& Skrab

kov

reactivity towards terpenes and aldehydes (Matisova
a
a,
1995). A more detailed description of the sampling and measure
mez
ment procedures used here is found elsewhere (Salazar Go
et al., 2013).

Table 2
Measuring parameters with the TD-GC/MS.
InstrumentGCMS QP2010Plus

Parameters

Oven programm

Starting Temperature: 50  C for 5 min

1. Ramp 5  C/min until 200  C, 15 min isotherm
Rxi-5MS (Restek) or comparable column
Length: 60 m  1.00 mm  0.25 mm
Column Flow: 2.43 ml/min, Pressure: 244.2 kPaInterface Temperature: 250  C, Split-ratio: 1:1
Carrier Gas: Helium
Linear Velocity: 40 cm/s
Ion Source Temperature: 200  C,
Interface Temperature: 250  C,
Detector Voltage: 0.9 V
Mode: Scan
Mass Range: 11e500 amu
Parameters
Desorption Flow: 60 ml/min
Desorption Time: 5 min
Desorption Temperature: 300  C
Transfer Line Temperature: 250  C
Cold Trap Temperature: 15  C
Desorption Temperature of Cold Trap: 300  C
Desorption Time of Cold Trap: 5 min

Separation column

MS

Thermodesorption Unit
TD20


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

51

Table 3
Retention time, target and qualifying ions used for GC/MS analysis.
No.

Compound name

Mw (u)

RT (min)

TI

QI1

QI2

LOD (mg/Nm3)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51

Acetone
n-Pentane
Furan
Dimethyl sulphide
Carbon disulphide
2-Propanethiol
Trimethylsilanol (TMS)
2-Butanone
2-Butanol
n-Hexane
2-Methylfuran
1-Propanethiol
Methylpropionate
Tetrahydrofuran
Benzene
2-Butanethiol
Thiophene
Hexamtyldisiloxane (L2)
2-Pentanone
3-Pentanone
n-Heptane
Methyl propyl sulphide
Methylbutanoate
Dimethyl disulphide
Toluene
n-Octane
Hexamethylcyclotrisiloxane (D3)
Methylpentanoate
Ethylbenzene
Octamethyltrisiloxane (L3)
m/p-Xylene
2-Heptanone
Ethylpentanoate
n-Nonane
o-Xylene
alpha-Pinene
Camphene
Octamethylcyclotetrasiloxane (D4)
Beta-Pinene
2-Pentylfuran
n-Decane
p-Cymene
D-Limonene
Decamethyltetrasiloxane (L4)
n-Undecane
Dipropyl disulde
Decamethylcyclopentasiloxane (D5)
n-Dodecane
n-Tridecane
n-Tetradecane
n-Pentadecane

58.08
72.15
68.08
62.14
76.14
76.16
90.20
72.11
74.12
86.18
82.10
76.16
88.11
72.11
78.11
90.19
84.14
162.38
86.13
86.13
100.21
90.19
102.13
94.20
92.14
114.23
222.46
116.16
106.17
236.53
106.17
114.19
158.24
128.26
106.17
136.24
136.23
296.62
136.24
138.21
142.29
134.22
136.24
310.69
156.31
150.29
370.72
170.34
184.37
198.39
212.41

4.93
5.07
5.12
5.58
6.09
6.36
6.45
7.31
7.40
7.40
7.60
7.83
8.44
8.47
9.77
9.85
10.08
10.17
10.51
10.90
11.01
11.70
11.91
13.20
14.05
15.15
15.77
16.12
18.02
18.13
18.34
18.88
19.14
19.21
19.35
21.02
21.71
22.33
22.77
22.83
23.00
24.23
24.49
24.83
26.49
27.20
27.70
29.74
32.75
35.60
38.82

43
42
39
47
76
43
75
43
45
41
82
76
57
42
78
41
84
147
43
57
43
61
43
94
91
43
207
74
91
221
91
43
29
43
91
93
93
281
93
81
57
119
68
207
57
43
73
57
57
57
57

58
41
68
62
44
41
47
72
59
56
53
41
29
41
77
57
58
73
86
29
41
90
74
45
92
41
208
43
106
221
106
58
57
57
106
91
121
282
41
82
43
91
67
73
43
41
266
43
43
43
43

41
72
38
45
32
76
45
29
31
86
39
47
59
72
51
61
45
148
41
86
71
41
71
79
65
85
191
57
65
222
105
71
88
41
105
92
79
283
69
53
41
134
93
295
71
150
355
71
71
71
71

0.003
0.010
0.005
0.008
0.023
0.024
0.013
0.006
0.005
0.004
0.012
0.027
0.006
0.016
0.005
0.023
0.007
0.021
0.007
0.005
0.004
0.009
0.014
0.012
0.003
0.004
0.013
0.014
0.002
0.046
0.005
0.003
0.013
0.005
0.004
0.018
0.018
0.019
0.016
0.017
0.006
0.015
0.015
0.004
0.004
0.011
0.021
0.004
0.006
0.011
0.016

2.5. Instrumentation for analysis

Qualitative and quantitative analysis of VOCs was carried out using
a GC/MS instrument (QP2010Plus) from Shimadzu Deutschland
GmbH, which was coupled with a thermodesorption unit (TD20) also
from Shimadzu. The GC/MS system was a prototype, especially
designed for shimadzu for the simultaneous analysis of major and
minor compounds in biogases. The system comprises two analytical
ow lines: Flow line 1 consisted of a nonpolar capillary column (Rxi5MS 60 m  0.25 mm  1 mm; Restek, Germany) for the analysis of H2S
and VOCs. The second ow line consisted of a serial combination of
Table 4
Concentration ranges of the major components in the different biogases.

Biogas plant 1
Biogas plant 2
Biogas plant 3

CH4 (%)

CO2 (%)

O2 (%)

N2 (%)

H2S (ppm)

57e67
44e60
25e64

32e41
32e44
18e39

0.1e0.2
<0.1e2.9
0.3e11.5

0.1e0.7
0.1e19
1e45

268e3174
11e1572
21e64

three columns, two CP-Poraplot Q (12.5 m  0.53 mm  20 mm;

Agilent, Germany) for the separation of CO2 and CH4. Subsequently, a
P-Molsieve 5A PLOT (25 m  0.53 mm  50 mm; Sigma Aldrich, Germany) for the separation of H2, O2 and N2. Both gas lines used helium
as carrier gas. For the analysis of gas samples the system was equipped
with a dosing valve (10 ports). All valves are mounted in a heated oven
in order to avoid water condensation or adsorption of VOCs on metal
walls. TD-tubes in the thermodesorption unit are desorbed with a
helium ow and the desorbed substances are refocused in a cold trap,
which is then rapidly desorbed. A detailed description of the parameters used at the thermodesorption unit and at the GC/MS is
shown in Table 2.
All chemical compounds were identied by retention time and
mass spectrum. External standard calibration was used for the GC/
MS quantication. The quantication parameters such as target (TI)
and qualifying ions (QI1 and QI2), the retention time (RT), the
molecular weight and the limit of detection (LOD) are given in
Table 3.


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

52

p-Cymene

D-Limonene

20000000
-Pinene
2-Butanone
-Pinene
Propanethiol
3-Carene

15000000

Abundance

Camphene
2-Carene
-Phellandrene

Biogas Plant 1
2-Heptanone

10000000

Biogas Plant 2
Biogas Plant 3

Acetone
Dipropyl disulfide

2-Pentanone
2-Ethylfuran

5000000

n-Octane
BP 2
BP 3
BP 1

0
0

10

15

20

25

30

Retention time (min)

Fig. 2. Representative TIC chromatograms of the biogases of the three biogas plants: BP 1 (black), BP 2 (blue) and BP 3 (red) (For interpretation of the references to colour in this
gure legend, the reader is referred to the web version of this article).

2.6. Calibration with liquid-loaded standards

For the calibration of the GC/MS system, 51 substances of 8
different compound classes (silicon-organic compounds, terpenes,
alkanes, BTEX, sulphur-organic compounds, esters, furanes, and
ketones) were used. 2-Butanol was the only alcohol that could be
quantied. The different standards were purchased as kits with the
exception of trimethyl silanol (TMS), and Hexamethylcyclotrisiloxane (Siloxane D3), which were ordered separately. Chemicals were of analytical grade and certied and purchased from LGC
Standards (Germany), Supelco (Germany) and Amchro (Germany).
A set of 10 standard solutions was prepared by dilution with
methanol for every compound class with exception of TMS and
Siloxane D3, which were diluted in hexane. The limit of detection
(LOD) for each VOC was calculated based on (Armbruster & et al.,
1994; Shrivastava & Gupta, 2011). For the loading of the TD-tubes
with the different standards a Calibration Solution Loading Rig
(CSLR) from Markes International (UK) was used in order to
vaporise the liquid standards before entering the TD-tubes. Laboratory blanks were used by injecting 1 ml of solvent into the TDtubes and purging them with nitrogen for the same time as for
the standard samples.

gas production, BP 1 showed rather constant values due to a well

established process control, whereas BP 3 showed signicant
uctuations in gas production caused by failures and eventual
shutdowns. Regarding carbon dioxide, BP 1 and 2 showed concentrations variations of less than 12%, being their values in a
normal range. For the case of BP 3, the carbon dioxide content
varied up to 21% and the contents of oxygen and nitrogen appeared
to be very high, which resemble the composition of landll gases
and is presumably an indication of air intrusion into the digester.
Hydrogen sulphide showed wide variations in concentration at the
BP 1 and 2, varying from 268 ppm to 3174 ppm and from 11 to
1572 ppm, respectively. The high amounts of hydrogen sulphide in
the biogas of plant 1 are rather unexpected due to the addition of
iron-containing solutions into the digesters, which would remove
in situ the hydrogen sulphide formed. However, these additions
were carried out every three weeks, which may allow the formation
of high amounts of hydrogen sulphide at the end of the period.
Moreover, poor mixing of the iron solutions inside the fermenter
could be a factor for the high values observed. Surprisingly, the
hydrogen sulphide in the biogas of plant 3 varies only from 21 ppm
to 64 ppm, although the characterisation of VOCs of this biogas
showed the presence of many sulphur-organic compounds with
higher concentration than in the biogases of plant 1 and 2. At this
site, iron solutions were added on a daily basis.

3. Results and discussion

3.1. Main components

3.2. Concentration of VOCs

The concentration ranges of methane, carbon dioxide, oxygen,

nitrogen and hydrogen sulphide at the three biogas plants (BP) are
shown in Table 4. The highest methane concentration (67%) was
observed at BP1, whereas the lowest (25%) was measured at the BP
3. BP 2 showed middle-range methane concentrations. In terms of

Taking into account that not all compound classes like small
alcohols, aldehydes and some thiols could be quantied with the
Tenax TA TD-tubes, the high VOCs concentrations observed in this
study are in agreement with previous results (Rasi et al., 2007). 2bed and 3-bed TD-tubes were only used for the determination of


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go
Table 5
Mean concentrations of measured VOCs in the three biogas plants.
Compound name

Plant 1

Plant 2

Mean (mg/Nm3) Mean (mg/Nm3)

Plant 3
Mean (mg/Nm3)

Silicon-organic compounds
L2
b.d.l.
L3
b.d.l.
L4
0.01
D3
0.34
D4
0.43
D5
0.68
TMS
0.07
Terpenes
alpha-Pinene
6.01
Camphene
0.75
Beta-Pinene
1.82
D-Limonene
38.09
BTEX
Benzene
0.24
Toluene
3.18
Ethylbenzene
0.56
m/p-Xylene
0.31
o-Xylene
0.17
p-Cymene
2.11
Furanes
Furan
b.d.l.
2-Methylfuran
1.43
Tetrahydrofuran
0.12
2-Pentylfuran
3.15
Ketones and Alcohols
Acetone
26.96
2-Butanone
21.88
2-Butanol*
1.96
2-Pentanone
2.32
3-Pentanone
1.34
2-Heptanone
1.81
Alkanes
n-Pentane
2.22
n-Hexane
3.03
n-Heptane
0.83
n-Octane
1.23
n-Nonane
0.43
n-Decane
1.72
n-Undecane
3.62
n-Dodecane
2.36
n-Tridecane
0.44
n-Tetradecane
0.05
n-Pentadecane
b.d.l.
Esters
Methyl propionate
0.88
Methyl butanoate
0.73
Methyl pentanoate
1.10
Ethyl pentanoate
0.46
Sulphur-Organic Compounds
Dimethyl sulphide
2.02
Carbon disulphide
1.62
2-Propanethiol
1.19
1-Propanethiol
3.77
2-Butanethiol
0.79
Thiophene
0.16
Methyl propyl sulphide 0.35
Dimethyl disulphide
1.14
Dipropyl disulphide
0.76

b.d.l.
b.d.l.
b.d.l.
0.83
1.36
6.70
0.04

0.04
b.d.l.
b.d.l.
0.93
1.85
0.70
0.17

71.84
23.15
42.89
158.40

9.26
0.53
7.03
26.78

2.13
3.59
1.57
0.84
0.58
359.84

1.58
8.15
1.24
1.26
0.60
6.89

1.49
3.06
0.11
6.52

b.d.l.
4.60
0.14
0.50

79.04
111.67
31.10
18.30
9.22
14.87

47.07
60.97
5.22
1.75
0.97
0.50

3.51
0.37
0.33
1.86
0.98
0.75
2.44
1.12
0.52
0.30
0.07

2.62
1.37
0.97
2.89
0.34
0.23
0.15
0.09
0.04
0.10
0.03

0.04
0.15
0.06
0.22

b.d.l.
b.d.l.
b.d.l.
0.10

2.76
5.91
0.50
2.87
4.65
0.74
1.76
0.68
1.65

1.25
4.69
0.52
19.51
0.79
1.16
0.50
0.39
3.16

TMS and siloxane L2, since for other VOCs like terpenes a very
strong adsorption was observed. Therefore, due to the unexpected
high VOC concentrations found in the biogases, the loading times
had to be reduced in order to avoid breakthrough of highly volatile
compounds. Fig. 2 shows a comparison of example chromatograms
of biogas samples taken in the late winter. The gas ows during
sampling were the same for all samples (100 ml/min); the only
difference is that for the biogases of plants 1 and 3 the loading time
was 1 min but for the biogas of plant 2, the sampling time has to be

53

reduced to 30 s due to the high content of VOCs found at this site.

These sampling parameters were found to be optimal in order to
avoid breakthrough of the 51 substances chosen for quantication
in this study. From Fig. 2 it is clear that terpenes appear as the
predominant substance group in the biogases of plants 1 and 2,
whereas for plant 3, ketones show the highest concentration followed by terpenes. Terpenes are constituents of essential oils and
can be obtained from owers, leaves, fruits, barks and roots. The
most relevant terpenes are limonene, pinene and camphene. According to Rasi et al. (2007), the terpenoid p-cymene appears as the
main VOC in biogas, which in this case correlates only for the biogas
of plant 2. One of the plausible reasons for these dissimilarities
between the biogas plants studied may lie in the type of operation,
since for instance BP 2, is the only one carrying out digestion at
thermophilic temperatures. This means that at higher temperatures more volatile compounds like terpenes are expected to
evaporate and migrate into the gas phase. Moreover, the composition of the feeding material at this site consists mainly of biowastes (ca. 87%), which have a high content of citrus fruits and
green wastes.
The presence of signicant amounts of sulphur-organic compounds in the biogas of plants 2 and 3 (Table 5) suggests high levels
of proteins in the substrates used as feedstock. Nevertheless, due to
regular additions of iron solutions into the digesters at BP 3, the
concentration of hydrogen sulphide was kept low. Acetone, 2Butanone and sometimes 2-Butanol were observed in all three
biogases. The highest concentrations of acetone and 2-butanone
where observed at the BP 2 (thermophile) with mean values of
79,04 and 111.67 mg/Nm3, respectively. For hardly degradable
chemical compounds, hydrolysis appears as the rate-determining
step. In contrast, for easily degradable substances is the acetate
production during fermentation the rate-limiting step. This may
result in accumulation in the fermenter of C3eC4 substances such
as, acetone or 2-butanone. Their accumulation in the digester and
their subsequent high levels in the biogas may be assigned to the
higher temperatures at this biogas plant, since according to the
Arrhenius equation, the rate of chemical reactions increases with
growing temperature.
Volatile organic silicon compounds measured in this study at
the 3 biogas plants were found to exhibit rather low concentrations, in accordance to (Rasi et al., 2010) but higher than those
found by Rasi et al. (Smet et al., 1999). Siloxane D6 was no
determined because due to its high molecular weight and low
vapour pressure was not expected to signicantly contribute to
the VOCs. Table 5 shows that the amounts of linear siloxanes
were mainly below the detection limits in all 3 biogas plants. On
the other hand, cyclic siloxanes exhibited signicant concentrations, being siloxane D5 the predominant siloxane in the biogas
of plants 1 and 2, whereas siloxane D4 showed higher values in
the biogas of plant 3. Although the concentrations of siloxane D4
and D5 appear to be low, they are in the range of typical values
found at landlls and sewage plants (Dewil et al., 2006;
Schweigkoer & Niessner, 1999). Siloxane D3 and TMS have
been found in previous studies (Rasi et al., 2010) to be the main
silicon organic compounds at German biogas plants, which process only manure and biowaste or energy crops.
Alkanes, furans and esters exhibited no signicant concentration levels in all biogases. Among aromatics, only toluene
exhibited a major contribution with mean concentrations at BP 1,
2 and 3 of 3.18, 3,59 and 8.15 mg/Nm3, respectively. Toluene is a
common component of landll gas or sewage gas but for biogas it
source can be presume to be due to the use of solvents normally
employed for cleaning purposes. Since some fractions of the
substrates used in the biogas plants contain fats and oils from
industrial or residential sources, some traces of toluene or even


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

54
1800
1600
1400

TVOCs (mg/Nm3)

1200
1000
800
600
400
200

B i og as P l an t 1

Aug-11

July-11

June-11

Apr-11

May-11

March-11

Jan-11

Feb-11

Dec-10

Oct-10

Nov-10

Aug-11

Sept-10

July-11

June-11

Apr-11

B i o g a s P l an t 2

May-11

March-11

Jan-11

Feb-11

Dec-10

Oct-10

Nov-10

Sept-10

July-11

Aug-11

June-11

Apr-11

May-11

March-11

Jan-11

Feb-11

Dec-10

Oct-10

Nov-10

Sept-10

B i o ga s P l an t 3

Fig. 3. Temporal variations of TVOCs concentrations in the biogases of the three biogas plants.

benzene can be expected. However, Rasi et al. (Farquhar &

Rovers, 1973) showed that aromatic compounds can be produced by the degradation of lignin under anaerobic conditions.
3.3. Time variations of VOCs
Regarding the amounts of total volatile organic compounds
(TVOCs), BP 1 showed a rather stable trend from autumn until late
summer with an average concentration of 146 mg/Nm3 (Fig. 3). In
contrast, BP 2 exhibited substantial concentration uctuations
ranging from 291 to 1731 mg/Nm3. BP 3 exhibited smaller variations for the period tested, ranging only from 84 to 528 mg/Nm3.
These ndings indicate a more homogeneous and constant substrate feeding at BP 1 and a high seasonal dependency of the substrate composition at plants 2 and 3. Although it was not possible to
sample all three biogas plants every month due to adverse factors
like failures or maintenance of gas engines, it is noticeable from
Fig. 3 that between February and April an increment in concentration takes place in all biogas plants. According to data from the
biogas operator of BP 2, substrate input of biowastes in November
accounted for about 1700 tons, whereas in December it was only
nearly half. A similar situation occurred between March and April,
the input of feedstock increased from 1200 tons in March to more
than 2000 tons in April. Surprisingly, the gas output in April was
smaller than in March at this site. In comparison to other biogas
plants, it seems plausible that due to the short residence times at
this site, the impact of the feeding rates is more noticeable on the
emission of VOCs, since not all the substrates are converted into
methane. Analyses of the fermentation residues conrm this idea
and showed that at BP1 the conversion of the dry matters is in the
range between 75 and 78%, whereas at BP 2 the conversion
accounted to 36e44% and at BP 3 it was in the range 82e89%.
It was found that the different compound classes analysed in
this study contribute in various ways to the TVOCs under the same
conditions and even within a compound class the trends observed
can vary considerably. Fig. 4 shows the trend developments in
concentration for terpenes and ketones, which are the most
important VOCs in all three biogas plants. At BP 1 D-limonene

appears as the predominant terpene with signicant uctuations in

concentration, showing a maximum in April (77 mg/Nm3). Only pcymene exhibits a small rise in concentration in January similarly to
D-limonene. In contrast, p-cymene emerges as the major terpene at
BP 2, showing a gradual increase in concentration and also reaching
a maximum in April with an extremely high level (751 mg/Nm3). At
this site D-limonene follows in concentration after p-cymene but
shows a different trend, although in April it also exhibits a raise in
concentration. At BP 3 the situation is a bit mixed, since at the
beginning of the sampling a-pinene and p-cymene predominate in
the biogas but from March on, D-limonene takes over. It is striking
that regardless of the terpene all biogas plants exhibit an increase
in concentration around April. Although there is a lack of information about the exact composition of the substrates fed in April at
the biogas plants, it is presumed that the fraction of fruits and
vegetables grows in comparison to the winter time.
For substances like ketones, which are in principle not present
in the substrates fed into the digesters but rather formed after
fermentation of the organic matter, the trends observed with
exception of BP 3, correlate to a certain degree with those of terpenes. For BP 1 acetone and 2-butanone appear as the most
important ketones but being acetone the predominant one with a
concentration peak in April. Similarly, acetone and 2-butanone
dominate at BP 2 but show slightly different trends to those of
terpenes. It is apparent that acetone accumulates in the biogas
quicker than 2-butanol, being its maximum already in March. 2Butanol shows a peak in April going through a decrease in May
and exhibiting a remarkable concentration peak in July (483 mg/
Nm3). At BP 3 once again are acetone and 2-butanone the main
ketones but they show a kind of alternating behaviour. The terpene
concentration peak observed in December is only resembled by
acetone. A correlation between the substrate input rates and the
concentration of the ketones could not be established. The biowaste
input in December accounted for 859 tons and in March it was over
2000 tons. Thus, an increase in acetone concentration would be
expected in March. It means that some unknown factors like the
composition of the substrates are determining the selectivity towards the production of acetone and 2-butanone.


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

55
Ketones Biogas Plant 1

Terpenes Biogas Plant 1

80

70

70

a)

d)

60

50
50
alpha-Pinene
Camphene

40

beta-Pinene
D-Limonene
p-Cymene

30

Concentration [mg/Nm ]

Concentration [mg/Nm ]

60

Acetone

40

2-Butanone
2-Butanol
2-Pentanone

30

3-Pentanone
2-Heptanone

20
20
10

10

0
Sept-10

Oct-10

Nov-10

Dec-10

Jan-11

Feb-11

Apr-11

June-11

July-11

Aug-11

Sept-10

Oct-10

Nov-10

Dec-10

Terpenes Biogas Plant 2

800

700

Jan-11

Feb-11

Apr-11

June-11

July-11

Aug-11

Ketones Biogas Plant 2

500

b)

450

e)

400
600

alpha-Pinene
Camphene

400

Beta-Pinene
D-Limonene
p-Cymene

300

Concentration [mg/Nm ]

Concentration [mg/Nm ]

350
500

300

Acetone
2-Butanone
2-Butanol

250

2-Pentanone
3-Pentanone

200

2-Heptanone
150

200
100
100

50

0
Nov-10

Dec-10

Feb-11

March-11

Apr-11

May-11

Nov-10

July-11

Dec-10

Feb-11

Terpenes Biogas Plant 3

120

March-11

Apr-11

May-11

July-11

Ketones Biogas Plant 3

180

c)

160

f)

100

80
alpha-Pinene
Camphene

60

Beta-Pinene
D-Limonene
p-Cymene

40

Concentration [mg/Nm ]

Concentration [mg/Nm ]

140

120
Acetone
100

2-Butanone
2-Butanol
2-Pentanone

80

3-Pentanone
2-Heptanone

60

40
20
20

0
Nov-10

Dec-10

Feb-11

March-11

Apr-11

May-11

June-11

July-11

Nov-10

Dec-10

Feb-11

March-11

Apr-11

May-11

June-11

July-11

Fig. 4. Temporal variations of the concentration of terpenes in the biogases of plant 1 (a), plant 2 (b) and plant 3 (c) and of ketones in plant 1 (d), plant 2 (e) and plant 3 (f).

Since the monthly sampling carried out at the biogas plants

shows only punctual measurements of a continuous process, the
biogases of BP 1 and BP 2 were chosen for an intensive sampling
campaign in order to gain some additional information about
short-term variations, which may reect any change in the substrates fed into the digesters or provide a correlation with other
process parameters. BP 1 was chosen due its apparent process
stability and at BP 2 due its high VOCs concentrations. The sampling
at BP 1 comprised a week whereas at BP 2 the sampling was carried
out for two weeks.
Fig. 5a shows that at BP 1 during the intensive sampling the
predominant terpene was D-Limonene, which exhibited much
narrower concentration variations (36e56 mg/Nm3) in comparison

to the monthly sampling. All other terpenes showed no signicant

contribution to the VOCs. When compared with the concentration
prole of ketones (Fig. 5c), it seem apparent that 2-butanone follows a similar trend as D-limonene, whereas acetone exhibits from
one day to the next a remarkable increase in concentration of about
230%. The other ketones analysed showed a very small contribution
in this period.
At BP 2 (Fig. 5b,d) the concentration trends of terpenes and
ketones show clear similarities, which is in contrast to the monthly
sampling. A concentration peak at the second day of sampling is
observed for all substances but in general the concentrations of
both terpenes and ketones appear to be extremely high during the
intensive sampling. The similarities of the concentration proles of


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

56

Ketones Biogas Plant 1

Terpenes Biogas Plant 1

80

120

a)

c)

70

100

50
alpha-Pinene
Camphene

40

beta-Pinene
D-Limonene
p-Cymene

30

Concentration [mg/Nm ]

Concentration [mg/Nm ]

60

80
Acetone
2-Butanone
2-Butanol

60

2-Pentanone
3-Pentanone
2-Heptanone

40

20

20
10

0
01.08.2011

02.08.2011

03.08.2011

04.08.2011

05.08.2011

01.08.2011

02.08.2011

04.08.2011

05.08.2011

Ketones Biogas Plant 2

Terpenes Biogas Plant 2

700

03.08.2011

900

b)

800

600

d)

700
600
400
alpha-Pinene
Camphene
300

Beta-Pinene
D-Limonene
p-Cymene

200

Concentration [mg/Nm ]

Concentration [mg/Nm ]

500

500

Acetone
2-Butanone

400

2-Butanol
2-Pentanone

300

3-Pentanone
2-Heptanone

200
100

100

Fig. 5. Temporal variations of the concentration of terpenes in the biogases of plant 1 (a) and plant 2 (b) and of ketones in plant 1 (c) and plant 2 (d).

terpenes and ketones may indicate a better process control at this

site. According to data from the biogas operator, the amounts of
input material in July increased by about 2.5 tons of biowaste in
comparison to June. This explains partly the high concentrations
observed, since due to the rather short residence times at this site,
the volatile decomposition products can migrate accordingly into
the gas phase. Among terpenes, p-cymene and D-limonene are the
predominant substances with maximum concentrations of 685 and
626 mg/Nm3, respectively. In terms of ketones, 2-butanone appears
as the predominant substance (Fig. 5d) with a maximum concentration of 878 mg/Nm3.
4. Conclusions
The amount and composition of the substrates used as feedstock
at biogas plants have not only an inuence on biogas production
but also a direct effect on the amounts of TVOCs. The concentrations of VOCs were found to be higher at the thermophilic biogas
plant than at the two mesophilic plants and their variations in
concentration showed a strong seasonal dependency, being the
concentrations in spring and summer considerably high. Longer
residence times seem to favour methane production and reduce
VOCs emissions. In terms of VOC composition, all three biogases
studied exhibited a similar compound spectrum but only the concentration of specic substances was found to be quite different,
although the nature of the substrates used at the biogas plants were
remarkably different. It was conrmed that terpenes and ketones
are the predominant VOCs in biogas and may exhibit signicant
concentration uctuations. The only halogenated compound found

during previous screening experiments was Dichloromethane and

among silicon-organic compounds, cyclic siloxanes appeared as the
main contributors and although their concentration was found to
be low, they could exhibit signicant concentration peaks.
Regarding analytics, it could be stated that the utilisation of a hydrophobic weak adsorbent, such as Tenax TA enables the quantitative measurement of the most important VOCs in biogas,
making a TD-GCMS a powerful technique due to its simplicity
during sampling and measurement. The in situ desulphurisation
through the addition of iron solutions into digesters reduces the
amounts of hydrogen sulphide but has little inuence on sulphides
and thiols. In terms of gas utilisation as energy source (for combustion engines or as biomethane), das biogas at BP 1 seems more
adequate due to the lower costs associated with its upgrading, since
the amounts of TVOCs are low. Only a better control of hydrogen
sulphide through more regular additions of iron solutions is
suggested.

Acknowledgements
The authors would like to thank all the biogas operators for their
cooperation during the measurement campaign and the information provided from the operational log books. The authors would
also like to thank the German Federal Ministry of Environment,
Nature Conservation, Building and Nuclear Safety for funding this
rung e Beitr
project (In German: Bioabfallverga
age zur Prozessoptimierung und zur verbesserten Gasnutzung. Grant No.
03KB032A and 03KB032B).


mez et al. / Chemosphere 153 (2016) 48e57

J.I. Salazar Go

References
Aitchison, E., 1996. Methane generation from UK landll sites and its use as an
energy source. Energy Convers. Mgmt. 37, 1111e1116.
ve, J., Dewil, R., 2008. Principles and potential of the
Appels, L., Baeyens, J., Degre
anaerobic digestion of waste-activated sludge. Prog. Energ. Combust. Sci. 34,
755e781.
Armbruster, D.A., et al., 1994. Limit of detection (LOD)/Limit of quantitation (LOQ):
comparison of the empirical and the statistical methods exemplied with GCMS assays of abused drugs. Clin. Chem. 40 (7), 1233e1238.
Bischoffsberger, W., et al., 2005. In: Anaerobtechnik, 2. Springer Verlag, Berlin
vollst
andig berarbeitete Auage, ed.
Dernbach, H., 1985. Landll gas utilization in Braunschweig e quality of gas and
damages due to corrosion. Waste Manag. Res. 3, 149e159.
Dewil, R., Appels, L., Baeyens, J., 2006. Energy use of biogas hampered by the
presence of siloxanes. Energy Convers. Manage 47, 1711e1722.
European Biogas Association, http://www.european-biogas.eu/eba/images/stories/
biogasbarometer.pdf.
Farquhar, G.F., Rovers, F.A., 1973. Gas production during refuse decomposition.
Water, Air Solid Pollut. 2, 483e495.
Franzius, V., 1985. State and trends regarding the utilization of landll gas in the
Federal Republic Of Germany. Waste Manag. Res. 3, 245e250.
rgasreinigung mittels Aktivkohle:
Friedrich, V., Beckler, J., Schneider, M., 2000. Kla
sung fr den Einsatz eines Oxidationskatalysators bei
eine wirtschaftliche Lo
gleichzeitiger Vermeidung silicatischer Ablagerungen, KA-Wasserwirtschaft,
Abwasser. Abfall 47, 1637e1640 (in German).
German Biogas Association, http://www.biogas.org/edcom/webfvb.nsf/id/DE_
Branchenzahlen/$le/14-11-25_Biogas%20Branchenzahlen_Prognose_20142015.pdf.
German Ministry of Environment, http://www.bmu.de/les/pdfs/allgemein/
application/pdf/broschuere_ee_zahlen_bf.pdf.
Gramann, N., 2011. Recht der Biogaseinspeisung. In: Graf, F., Bajohr, S. (Eds.),
Biogas: Erzeugung, Aufbereitung, Einspeisung. Oldenbourg Industrieverlag
GmbH, Mnchen, pp. 41e56 (in German).
Maier, I., Fieber, M., 1998. Retention characteristics of volatile compounds on Tenax
TA. J. High Res. Chromatogr. Chromatogr. Comm. 11, 566e576.


kova

, S., 1995. Carbon sorbents and their utilization for the
Matisov

a, E., Skrab
a
preconcentration of organic pollutants in environmental samples.
J. Chromatogr. A 707, 145e179.
Murphy, J.D., McCarthy, K., 2005. The optimal production of biogas for use as a
transport fuel in Ireland. Renew. Energy 30, 2111e2127.
Rasi, S., Veijanen, A., Rintala, J., 2007. Trace compounds of biogas from different
biogas production plants. Energy 32, 1375e1380.
Rasi, S., Lehtinen, J., Rintala, J., 2010. Determination of organic silicon compounds in
biogas from wastewater treatments plants, landlls, and co-digestion plants.

57

Renew. Energy 35, 2666e2673.

, J., Rintala, J., 2011. Trace compounds affecting biogas energy utiRasi, S., L
antela
lisation e a review. Energy Convers. Manage 52, 3369e3375.
Raven, R.P.J.M., Gregersen, K.H., 2007. Biogas plants in Denmark: successes and
setbacks. Renew. Sustain. Energy Rev. 11, 116e132.
Rossol, D., Schmelz, K.-G., Hohmann, R., 2003. Siloxane im Faulgas e Erfahrungen
mit verschiedenen Gasreinigungsverfahren bei Emschergenossenschaft und
Lippeverband, KA-Abwasser. Abfall 50, 1043e1051 (in German).

mez J.I., Lohmann H., Krassowski J., Time-dependent concentration of
Salazar Go
siloxanes in biogases from biowaste and co-feeding plants by single-sorbent
adsorption, Energy, (submitted).

mez, J.I., Gerstner, A., Jovic, A., 2013. Messung von organischen SpurSalazar Go
n, D. (Eds.), Messmethoenverbindungen. In: Liebetrau, J., Pfeiffer, D., Thra
densammlung Biogas: Methoden zur Bestimmung von analytischen und
prozessbeschreibenden Parametern im Biogasbereich, vol. 7. Fischer Druck,
Leipzig, pp. 52e59.
Scherer, P.A., 2001. Mikrobiologie der Verg
arung von festen Abfallstoffen. In:
mpfer, P., Weienfels, W.D. (Eds.), Biologische Behandlung Organische
Ka
lle. Springer-Verlag Berlin-Heidelberg, p. 48 (in German).
Abfa
Schweigkoer, M., Niessner, R., 1999. Determination of siloxanes and VOC in landll
Gas and sewage Gas by Canister samplingand GC-MS/AES analysis. Environ. Sci.
Technol. 33, 3680e3685.
Shin, H.-C., Park, J.-W., Park, K., Song, H.-C., 2002. Removal characteristics of trace
compounds of landll gas by activated carbon adsorption. Environ. Pollut. 119,
227e236.
Shrivastava, A., Gupta, V.B., 2011. Methods for the determination of limit of detection and limit of quantitation of the analytical methods. Chron. Young Sci. 2,
21e25.
Smet, E., van Langenhove, H., De Bo, I., 1999. The emission of volatile compounds
during the aerobic and the combined anaerobic/aerobic composting of biowaste. Atmos. Environ. 33, 1295e1303.
Staffell, I., Green, R., Kendall, K., 2008. Cost targets for domestic fuel cell CHP.
J. Power Sources 181, 339e349.
Tsagarakis, K.P., 2006. Ch. Papadogiannis, Technical and economic evaluation of the
biogas utilization for energy production at Iraklio Municipality, Greece. Energy
Convers. Manage 47, 844e857.
Weiland, P., 2000. Anaerobic waste digestion in Germany e status and recent developments. Biodegradation 11, 415e421.
Weiland, P., 2003. Production and energetic use of biogas from energy crops and
wastes in Germany. Appl. Biochem. Biotech. 109, 263e274.
Weiland, P., 2010. Biogas production: current state and perspectives. Appl. Microbiol. Biotechnol. 85, 849e860.
Woolfenden, E., 2010. Sorbent-based sampling methods for volatile and semivolatile organic compounds in air. Part 2. Sorbent selection and other aspects
of optimizing air monitoring methods. J. Chromatog. A 1217, 2685.