Accepted Manuscript Title: Comprehensive two-dimensional gas chromatography for biogas and biomethane analysis Authors: F. Hilaire, E. Basset, R. Bayard, M. Gallardo, D. Thiebaut, J. Vial PII: DOI: Reference:

S0021-9673(17)31452-8 https://doi.org/10.1016/j.chroma.2017.09.071 CHROMA 358900

To appear in:

Journal of Chromatography A

Received date: Revised date: Accepted date:

30-11-2016 8-8-2017 28-9-2017

Please cite this article as: F.Hilaire, E.Basset, R.Bayard, M.Gallardo, D.Thiebaut, J.Vial, Comprehensive two-dimensional gas chromatography for biogas and biomethane analysis, Journal of Chromatography A https://doi.org/10.1016/j.chroma.2017.09.071 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Comprehensive two-dimensional gas chromatography for biogas and biomethane analysis F. Hilaire 1, E. Basset2 , R.Bayard3-4, M. Gallardo 2, D. Thiebaut 1,*, J. Vial 1 1UMR

CBI, Laboratoire des Sciences Analytiques, Bioanalytiques et Miniaturisation, ESPCI Paris,PSL Research University, 10 rue Vauquelin, 75231 Paris Cedex 05, France 2ENGIE, Research and Technologies Division, CRIGEN, 361 av. du Président Wilson, BP 33, 93211 St-Denis-la-Plaine Cedex, France 3Univ Lyon, INSA-Lyon, DEEP - EA 7429, 9 rue de la Physique, F69621 Villeurbanne Cedex, France 4 RECORD – Campus LyonTech, 66 bld. Niels Bohr CEI 1 – CS52132, Villeurbanne Cedex, France *Corrresponding

author :Tel.: +331 40 79 46 48 ; fax: +331 40 79 47 76

E-mail address: [email protected]

  

Highlights for submission of “Comprehensive two-dimensional gas chromatography for biogas and biomethane analysis” First report of GCxGC applied to biogas and biomethane analysis Access to detailed composition of biogas versus biomass inputs Qualify the performances of biogas purification processes

ABSTRACT The gas Industry is going to be revolutionized by being able to generate bioenergy from biomass. The production of biomethane – a green substitute of natural gas – is growing in Europe and the UnitedStates of America. Biomethane can be injected into the gas grid or used as fuel for vehicles after compression. Due to various biomass inputs (e.g. agricultural wastes, sludges from sewage treatment plants, etc.), production processes (e.g. anaerobic digestion, municipal solid waste (MSW) landfills), seasonal effects and purification processes (e.g. gas scrubbers, pressure swing adsorption, membranes for biogas upgrading), the composition and quality of biogas and biomethane produced is difficult to assess. All previous publications dealing with biogas analysis reported that hundreds of chemicals from ten chemical families do exist in trace amounts in biogas. However, to the best of our knowledge, no study reported a detailed analysis or the implementation of comprehensive two-dimensional gas chromatography (GC x GC) for biogas matrices. This is the reason why the benefit of implementing two-dimensional gas chromatography for the characterization of biogas and biomethane samples was evaluated. In a first step, a standard mixture of 89 compounds belonging to 10 chemical families, representative of those likely to be found, was used to optimize the analytical method. A set consisting of a non-polar and a polar columns, respectively in the first and the second dimension, was used with a modulation period of six seconds. Applied to ten samples of raw biogas, treated biogas and biomethane collected on 4 industrial sites (two MSW landfills, one anaerobic digester on a wastewater treatment plant and one agricultural biogas plant), this analytical method provided a “fingerprint” of the gases composition at the molecular level in all biogas and biomethane samples. Estimated limits of detection (far below the µg.Nm-3) coupled with the resolution of GC x GC allowed the comparison of the real samples considered. This first implementation of GC x GC for the analysis of biogas and biomethane demonstrated unambiguously that it is a promising tool to provide a “fingerprint” of samples, and to monitor trace compounds by families. Keywords: biogas; biomethane; biomass; comprehensive two-dimensional gas chromatography x GC); landfill; anaerobic digestion

(GC

1. INTRODUCTION Due to the increasing fossil fuel prices and their harmful effect on the environment, current waste management policies favor the development of renewable energies. The great worldwide challenge

relies on biomass conversion to energy. Gas generation from biomass is considered as a promising option to produce an energy carrier for natural gas substitution. Consequently, the gas industry will probably be revolutionized by the bioenergy industry in the near future. Indeed, new categories of gases such as syngas, biogas and biomethane will be produced from various biomass-transformation pathways (anaerobic digestion, gasification and methanation) [1]. In the European policy context to ensure energy supply and to develop renewable energies, European Union (EU) countries are developing strategies to produce biofuels. The production of gas from biomass can be performed with thermochemical processes (gasification and methanation), and biological processes [2,3]. Among them, methane production by anaerobic digestion of organic residues such as municipal solid wastes (MSW), agricultural residues and others biowastes is becoming a major option for reaching the European biofuel target of 10% of total fuel consumption by 2020. The primary energy production in Europe from biogas was 13.4 Million Tons of Oil Equivalent (MTOE) in 2013, with an annual growth of over 10% during the past 5 years [4]. For France, the French environmental protection agency (ADEME) estimates a potential of between 12 and 30 terawatt hours (TWh) of biomethane from biogas production by 2030, which could represent more than 5 % of the overall gas consumption in the country [5]. Coming from MSW landfills and anaerobic digesters, biogas and biomethane are produced by the biological pathway of methanogenesis. Without oxygen, organic matter is hydrolysed and monomers are converted to water, carbon dioxide and methane. Biogas (digester gas (DG) and landfill gas (LFG)) can be used either in combined heat and power engines (CHP) to produce electricity and heat by cogeneration ; or when purified and upgraded as biomethane, they can be injected into the gas grid or to be used as fuel for vehicles. In these two latter cases, the biomethane quality must comply with the specifications of distribution systems operators (DSO) and transportation systems operators (TSO) [6]. Digester gas and landfill gas stemming from anaerobic biodegradation of sewage sludges, MSW and biomass residues from agriculture, are composed of three main families of compounds. Firstly, main gases such as methane (CH4), carbon dioxide (CO2) and nitrogen (N2) are incondensable. Secondly, inorganic compounds (e.g. hydrogen sulfide (H2S), carbonyl sulfide (COS) or ammonia (NH3)) can be present from part per million level (ppm) to hundreds ppm. Finally, a biogas contains volatile organic compounds (VOC) such as terpenes, ketones or siloxanes, depending on the biomass origin. Usually, one part of trace components is collected during water condensation. This can lead to variable compositions according to wastes, process parameters, seasonal effects, cleaning steps, etc. [7]. Generally, the gas characterization is defined by the determination of total VOCs content. In fact, VOCs can impact the valorization process by blocking gas coolers or filter elements, and engine suction channels [8]. For instance, siloxanes can destroy engines used for cogeneration and should be treated before gas valorization. Thus, analytical monitoring of biogas and biomethane is needed to control the gas quality and to design cleaning processes at industrial scale [9]. All along the biomethane production process, gas sampling and analysis is a major challenge in order to control and adjust the cleaning process parameters, optimize them and assess their performance to reach the biomethane specifications as recommended by the European Committee of Standardization - CEN TC 408 - Natural gas and biomethane for use in transport and biomethane for injection in the natural gas grid NF EN 16723. To optimize gas treatment processes, dedicated sampling and analytical methods have to be developed for an enhanced understanding of the biogas composition from different biomasses (landfill, agricultural waste, sewage sludge, etc.). Few articles deal with composition of landfill gases, or digester gases from sewage sludge, MSW or agricultural waste [10–30], and only one is devoted to a detailed biomethane characterization [10].

Gas chromatography coupled with mass spectrometry (GC-MS) is the main technique used for biogas analysis [13–30]. Up to 200 compounds have been detected in some biogases [19,20,24]. 10 chemical families are often reported: alkanes, cycloalkanes, alkenes, terpenes, mono- and poly- aromatic hydrocarbons (PAHs), oxygen-organic compounds (e.g. ketones, esters, furans, aldehydes and alcohols), sulfur-organic compounds (e.g. thiophene and derivatives), nitrogen-organic compounds (e.g. pyridin, quinolin) and siloxanes. Rasi et al. [20,21] evidenced a variation of biogas composition as a function of initial biomass, by comparing three biogas obtained from a MSW landfill site, sewage sludge and agricultural waste. More aromatics and nitrogen compounds could be identified in a LFG compared to a sewage sludge or an agricultural waste DG. This can be explained by the age of the landfill, and the waste degradation state [16–18]. Aliphatic and aromatic compounds contained in sewage sludge biogas come from paints, solvents or industry products. Siloxanes are also present in larger quantities in biogas from landfill and sewage sludge than from agricultural wastes. Siloxanes mainly come from cosmetics (creams, shampoos, etc.) or chemical products (glue, seals, additives, etc.). Finally, Rasi et al. [20,21] noticed that the VOCs content is lower in agricultural waste biogas because of the lower chemical diversity of agricultural feedstocks. Rasi et al. [20,21] also reported that sampling conditions have a strong influence on biogas analyses. Moisture, sampling temperature or site location (urban area or not) imply different VOCs concentrations. Total VOCs varied between 46 and 173 mg.m-3, 13 and 268 mg.m-3 and 5 to 8 mg.m-3 respectively for the LFGs, sewage sludge DGs and agricultural waste DGs. Piechota et al. [22] confirmed similar ranges of concentrations (250-500 mg.m-3 for landfill biogas, 70-150 mg.m-3 for biogas from wastewater sludges and 7 to 20 mg.m-3 for biogas generated from agricultural feedstocks). Zhou et al. [23] identified 48 compounds during winter against 60 during summer. The concentration of VOCs contained in LFGs is depending on weather [24,25] and depth of sampling [26]. Temperature is also an important factor concerning the variability of total VOCs concentration because the boiling point of several compounds is low (near 40 °C). Marine et al. [27] studied sampling methods for characterization of pollutants in biogas. Tedlar bags [24,27] and sorbents tubes are very common. However, some compounds can be adsorbed on the wall of Tedlar bags and the sampling stability is not guaranteed after 48 hours [20,23]. Montiu et al. [28] reported that sorbent tubes were a more convenient sampling technique for quantifying low concentrations compared to impingers and Tedlar bags. Allen [18], Chiriac [19], Gallego [24,25], Arrhenius [29] and LaRegina [30] analyzed biogas using sorbent tubes associated to thermaldesorption and Gas Chromatography. This injection technique involved cold trapping and flash heating. Gallego identified 68 compounds in a model standard mixture despite some compounds were coeluted. Therefore, in the present study, the more resolutive GC x GC [31] has been used to improve separation capabilities. To our knowledge, this is the first example of GC x GC implementation for biogas analysis. Indeed, GC x GC has been used in petrochemical and bio petrochemical analysis [32–38]. In most of the cases, a non-polar column was used in the first dimension while a more polar one was implemented in the second dimension [39–43]. In this paper, our objective is to demonstrate the interest of the implementation of GC x GC for the analysis of biogas and biomethane. Gas sampling, method development using a test mixture composed of 89 relevant compounds, and comparison of preliminary semi-quantitative results obtained on 10 different real samples are discussed.

2. EXPERIMENTAL 2.1. Standards

All standard compounds and solvents, of GC-grade (purity > 98%), were obtained from Restek (USA) and Sigma–Aldrich (USA). In order to optimize the analytical method, commercial standard mixtures and pure compounds were mixed to obtain a diluted standard mixture (10 µg.mL-1 in 1:1 acetone/dichloromethane (v/v)). The standard mixture was composed of 89 volatile organic components that can be gathered into ten main families described in Table 1. The 21 compounds in bold characters were selected for semi-quantitative results and referred as target compounds. Nitrogen (99,999% purity) and Helium (99,9999% purity) were used, respectively, for accelerated solvent extraction (ASE) and as carrier gas in GC and GC x GC. They were obtained from l’Air Liquide (France) and Messer (Germany).

2.2. Samples Samples description Biogas and biomethane from 4 different sites were sampled (Table 2). On each site, raw biogas was sampled. For landfill sites A and B, the treated biogas was collected after a treatment step by adsorption on activated charcoal. On site B, a pre-treated biogas was collected after a pre-treatment step on oxides. For site C, corresponding to an anaerobic digester using sludges of wastewater treatment plant, biogas was treated using adsorption on activated charcoal and biomethane was produced after an upgrading step using a membrane system to remove carbon dioxide. Finally, for site D, raw biogas and biomethane were collected from the anaerobic digester of an agricultural biogas plant.

Sample preparation procedure Raw biogas was sampled at a flow of 100 mL.min-1 through an ORBOTM 609 Amberlite® XAD®-2 and ORBOTM Charcoal tubes (Supelco®, Sigma–Aldrich, USA). Treated biogases and biomethanes were sampled at 500 mL.min-1 on the same type of sorbent tubes. Flow rates were adjusted with a 1/8” micrometric valve (Swagelok®, USA) and the volume was measured with a volumetric gas flow meter (Ritter®, Germany). A mass flow controller from Bronkhorst (France) was also used in order to control the flow rate and to measure the total gas volume collected. Between 1 and 20 liters of gas were sampled depending on the source. Each tube contained two beds of sorbent. After sampling, compounds collected on first sorbent bed were extracted by Pressurized Liquid Extraction (PLE) using an ASE 200 system (DionexTM, USA). Solvents used for this extraction were acetone/dichloromethane in 1:1 (v/v) ratio. An aliquot of 50 mL of the extracted volume was gently evaporated using nitrogen until a final volume of 150 µL. The breakthrough volume was checked by analyzing the second bed of sorbent tube.

2.3. Instrumentation and chromatographic conditions Standard and samples were analyzed using a TRACE GC x GCTM model gas chromatograph (ThermoFischer Scientific, USA) equipped with a TriPlus RSHTM Autosampler and a split/splitless injector; it was coupled to a ISQTM quadrupole mass spectrometer (MS). To avoid contamination by septum, the injector was equipped with a Microseal system (Merlin Instrument Company, USA). The injected volume was 1 µL with a split ratio of 10:1 and the injector temperature was 275°C. A two-jet cryogenic CO2 modulator was used with a modulation period of 6 seconds. Helium at a constant flow of 1 mL.min1 was used as carrier gas. The column set was composed of a nonpolar DB-5MS column (5% phenyl/95% dimethylpolysiloxane, 30 m x 0.25 µm; 0.25 µm) in the first dimension and a semi-polar DB-17 column (50% phenyl/50% dimethylpolysiloxane, 1.35m x 0.1 µm; 0.1 µm) in the second dimension, both obtained from Agilent, USA. The initial temperature was 45°C and was maintained during 5 min. Then, the slope of the temperature gradient was 1°C.min-1 until 53°C, 3°C.min-1 until 125°C and 20°C.min-1

until 275°C. The MS operating conditions were: ion source temperature at 230°C, interface temperature at 280°C, the electron ionization mode was set at 70 eV with a mass range of m/z 10-350. The solvent delay was 5 min. For comparison with GC-MS, the same conditions were applied but the modulator was off. XcaliburTM and ChromCardTM (Thermo-Fischer Scientific, USA) software was used for data acquisition and processing. The compounds identification was performed using NIST Mass Spectral Library software (NIST 02, Software Version 2.0, USA). The identification used both the mass spectra comparison with the library and the comparison with retention time of the compounds of the standard mixture. Identification was performed only for the peaks having a similarity with the library data higher than 80%.

2.4. Semi-quantitative analysis A semi-quantitative analysis, based on a calibration forced through zero, was carried out to estimate the content of target compounds in biogas. By semi quantitative analysis, it is meant that the results provided here were only given to obtain an idea of the concentrations. A more accurate determination of the concentrations would have required further developments which would have been far beyond the scope of the present feasibility study. The standard mixture was injected twice and served as calibration to estimate the content of the 21 target compounds in real samples. The concentration (C1) of each compound was estimated by comparing peaks area of the compounds in sample and standard mixture. C1 corresponds to target compound content in the concentrate volume (after evaporation). The equation (1) was used to calculate the concentration of target compounds in each sample (Cs): 𝐶𝑠 = 𝐶1 ∗

𝑉1 ∗ 𝑉0 (1) 𝑉𝑠 ∗ 𝑉𝑖

Where: V1 is the concentrated volume after evaporation V0 is the extracted volume (about 15 milliliters) VS is the collected gas volume (Nm3, in normative conditions of temperature (273.15K) and at atmospheric pressure) Vi is the initial volume after extraction (50 milliliters) Equation (2) was used in order to assess the performance of purification treatments (E). Cf is the concentration of target compound in biomethane or treated biogas and Ci the concentration of target compound in raw biogas for the same site. Thus, E represents the elimination rate of the treatment process for the component. 𝐸 = 100 −

𝐶𝑓 × 100 (%) (2) 𝐶𝑖

3. RESULTS AND DISCUSSION 3.1. Comparison between GC and GC x GC The GC x GC chromatogram of the standard mix is presented on the Figure 1. As expected using our column set, the usual organization by families was observed on the chromatogram [41,42]; the 89 compounds were organized in two groups: in the second dimension, non-polar compounds such as alkanes or siloxanes eluted first. More polar compounds such as sulfur- and halogen-compounds eluted above. In the first dimension, compounds eluted according to their carbon number. In order to compare GC x GC-MS and GC-MS, same conditions (except for specific GC x GC conditions) and same detection mode were used in GC-MS. The Figure 2a shows a chromatogram of the standard

mixture analyzed by GC-MS. Among the 89 compounds, only 79 were identified by GC-MS. Some compounds could not be identified because of the lack of sensitivity of GC-MS. On the standard mixture chromatogram, 18 co-elutions were identified in GC-MS for compounds of different chemical families. 15 co-elutions were resolved with GC x GC-MS thanks to the second dimension (vertical axis). Figures 2b and 2c illustrate examples of separation with GC x GC-MS for compounds that were partly or totally co-eluted with other compounds in GC-MS. In fact, as illustrated in Figure 2b, the peak at 23.7 min in GC-MS could be separated into three spots in GC x GC-MS: phenol, more polar on the top, 1,2,3-trimethylbenzene and beta-pinene. These compounds have similar boiling points (respectively 182°C, 176°C and 167°C). Hence, they were almost eluted on the vertical axis in the two-dimensional chromatogram. With GC-MS, 1,2-dichlorobenzene was partially coeluted with indane, which was also totally merged with butylcyclohexane. Boiling temperatures are quite similar for these three compounds: 180°C for the 1,2-dichlorobenzene, 177°C for indane and 180°C for butylcyclohexane. Others pairs of compounds partly unresolved in GC-MS were identified respectively as hexanoic acid butyl ester (BP: 204°C) and benzothiophene (221°C), and as dodecamethylhexasiloxane (230 °C) and benzothiazole (224°C). For the second pair, the GC-MS resolution was below 1. Some compounds were still not separated with GC x GC, essentially compounds having similar polarity like m-xylene and p-xylene, or styrene and o-xylene or butanoic acid ethyl ester and dibromochloromethane. Nevertheless, 80% of co-elution occurring in GC-MS could be avoided with GC x GC-MS.

3.2. Method assessment First, for sampling, two sorbent tubes were tested: XAD-2 and charcoal. Blank extracts were carried out using the protocol described in the experimental section for real samples analysis. Figure S1 shows the two chromatograms obtained using a XAD-2 tube (Figure S1.a) and a charcoal tube (Figure S1.b). For XAD-2 tube, 33 compounds were detected against 47 for charcoal. Moreover, 4 compounds detected in XAD-2 tube were present in the standard mixture (p-xylene, n-heptane, limonene and dodecamethylcyclohexasiloxane) against 12 for charcoal. Thus, XAD-2 was preferred because it gave less “artifacts”. For the analysis of a real sample, the biogas of site A, the difference between the two sorbents was much higher: using the XAD-2 matrix (Figure S2) 216 peaks could be trapped and detected while charcoal allowed to trap and detect only 98 compounds. To give an idea of the method repeatability, two injections of the standard mixture at 10 µg.mL-1 of each component were used. Only the 21 target compounds were considered. Relative standard deviation (RSD) was estimated for retention time on the first and on the second dimensions, and for peak area. Relative standard deviation (RSD) for retention time on the first dimension, on the second dimension and for peak area was respectively between 0 and 0.5%, 0 and 7%, 2 and 30%. The pooled RSD (quadratic mean of the RSDs), corresponding to a kind of “average” estimation of the RSD calculated for all the compounds, was also estimated for these three parameters. It was 0.2% for retention time on the first dimension, 4.3 % for the second dimension and 11.8 % for peak area. If a RSD calculated from only two repetitions is not really informative, the pooled value combines the degrees of freedom and presents, from a statistical point of view, the same quality as what we would have obtained for a single compound injected more than 20 times. These values were consistent with our objective of semi-quantitative analysis. In order to provide a first estimation of the detection limit (LOD) and of semi-quantification limit, the standard mixture was used for which the 21 target compounds were all at the same concentration of

1 µg.mL-1 in acetone/dichloromethane 1:1. For each analysis, a peak was considered when the signal to noise ratio was higher to 3. This corresponds to the limit of detection (LOD). Limit of quantification (LOQ) is generally calculated when signal/noise (S/N) is equivalent to 10 [44]. In the present study where only semi-quantitation was considered, LOQ was taken as the concentration corresponding to a S/N of 5. For liquid extracts, the LOD was between 20 ng.mL-1 to 350 ng.mL-1. Such values, determined for the synthetic mixture prepared in a liquid, represents 1 ng.Nm-3 for octamethylcyclotetrasiloxane, and 9 ng.Nm-3 for toluene in the gas sample. LOQ ranged between 30 ng.mL-1 to 600 ng.mL-1 for liquid samples, and 1.5 ng.Nm-3 for octamethylcyclotetrasiloxane and 14 ng.Nm-3 for toluene in gas samples. 21 target compounds of standard mixture have also been injected in GC-MS in order to compare S/N ratio of the two techniques. A mean ratio of 5 was found on the S/N ratio between GC x GC-MS and GC-MS. A similar ratio was obtained by Vendeuvre [32].

3.3. Sample analysis Figure 3 shows a comparison of GC x GC-MS and GC-MS chromatograms obtained for a real sample of biogas from site A. Conclusions given above for standard mixture can be applied to real samples because co-elutions observed in GC-MS (cf. Figure 3b) can be resolved in GC x GC-MS thanks to the selectivity provided by the second dimension (cf. Figure 3c). Four biogases were analysed (figures 4-5 and table 3). They were primarily composed of oxygenatedorganic compounds, aromatics hydrocarbons, alkanes, cycloalkanes, alkenes, terpenes and siloxanes. In fact, several target compounds of the standard mixture were found in these biogases: for example, 38 compounds of the standard mixture were found in the landfill biogas of site A (Figure 4a), including mono-aromatic hydrocarbons such as toluene, ethylbenzene, p-cymene or terpenes such as alphapinene and d-limonene. The Figures 4b and 4c also show GC x GC chromatograms of biogas from sewage sludge of site C and biogas from agricultural residues (site D) using the same intensity scale. More than 100 and 50 compounds were detected, respectively, in these biogases and around 50% of the peaks detected could be identified in each sample thanks to mass spectrum library. Figure 5 provides the number of compounds by family for each type of biogas. Table 3 shows the the range of total VOC by family for the the biogases investigated in this study. This range was estimated from the semi-quantitative results measured on target compounds identified in the sample. Comprehensive two-dimensional gas chromatography can give access to a sample cartography by chemical families in one analysis and chemical families were confirmed to be different from one biogas to another. As an example, the landfill biogas of site A is not the same as the one of site B even if the same production process was used for both. Site A sample contains more alkanes than site B sample. It is possible to explain this difference by the degradation state of wastes, which is very influent on the biogas composition, as already reported in the literature [16–18]. Moreover, there are more alkanes and cycloalkanes in landfill biogas and biogas from wastewater treatment plant than in agricultural biogas. In our samples, sewage sludge biogas contained less oxygen compounds (phenols, ketones, alcohols) than landfill biogas and no halogenated compound. The biogas from agricultural residues was mainly composed of oxygen and sulfur compounds. Mono-aromatic hydrocarbons seemed to be more present in landfill biogas than in agricultural biogas and sewage sludge, except toluene which was more abundant in sewage sludge biogas than in agricultural biogas. Siloxanes, mainly coming from shampoo, lotion and cream were more abundant in landfills and sewage sludges (site C). Concerning terpenes, highest concentrations were found in biogas from landfill and sewage sludge (Table 3). These differences could be related to the waste characteristics (degradation state, origins and compositions). Differences in biogas composition according to the different origins of biomass used for its production -even with the same process- were easily exhibited with the cartography obtained with GC x GC.

However, for these reasons, these results obtained from a limited number of samples for each site must be considered as merely indicative. 3.4. Estimation of elimination rate of trace compounds The effect of treatments involved in the conversion of biogas into biomethane was investigated for each production of biomethane or treated biogas (Table 4). The effect of the treatment process (E) can be assessed by the decrease of the VOCs amount between the raw biogas (input) and the treated biogas or biomethane (output) as described in Equation 2. For all sites, at least a 94% decrease in VOCs concentration was found between raw biogas and treated biogas or biomethane. For the site A, the treatment by adsorption on activated charcoal could decrease the total VOCs by 95 % between raw and treated biogas. The treatment by oxides, used principally for H2S removal, implied the elimination of 86% of VOCs. For all treated biogas considered, most of the compounds were eliminated by activated charcoal. Owing to the efficiency of treatments, for all the biomethanes and treated biogases investigated during this work, the VOCs concentration were dramatically decreased making their injection into the gas grid quite realistic. Used after adsorption on activated charcoal in order to concentrate methane and eliminate carbon dioxide, the membrane process had no significant effect on the VOCS concentration as it was expected. To illustrate the elimination of COVs, Figure 6 shows the chromatograms of agricultural biogas and biomethane obtained from the same site, at the same intensity scale. 5 liters of raw biogas and 20 liters of biomethane were sampled. The visual comparison of the two chromatograms confirmed that the compounds contained in the biogas have almost been totally removed. Indeed, results indicated that 99% of by-products were removed. The treatment by filtration on charcoal was particularly efficient on alkanes (located at the bottom of the GC x GC chromatogram), terpenes, siloxanes, hydrocarbons and sulfur compounds.

4. CONCLUSION This paper proposes an analytical method including sampling of biogas and biomethane and analysis by comprehensive two-dimensional gas chromatography. It shows that GC x GC-MS is a powerful analytical tool for the detailed characterization of complex biogas and biomethane samples, and to estimate the efficiency of purification processes. The analytical procedure is sensitive and resolutive: more compounds can be detected than in GC-MS, and GC x GC-MS provides easier identification of target compounds. The sensitivity is 5 times better compared to GC-MS. A cartography of biogas and biomethane could be obtained to improve the knowledge on sample composition versus their origin. The efficiency of treatment by charcoal could also be assessed using a raw semi quantitative approach. Because of the diversity of the inputs and the variability inherent to the processes and production sites, the results provided here must only be taken as representative examples of what can be observed. Thus, the results that could be obtained from this preliminary study, cannot be considered to be representative of a given process type or site. As a conclusion, in the current context of the new gas and bioenergy industry, GC x GC-MS may allow enhanced diagnostics for process performances or risk management. For biogas characterization, the benefits of GC x GC-MS are significant thanks to higher sensitivity and resolution. Thus, GC x GC-MS offers new perspectives to handle and optimize the biogas and biomethane production. Further developments can be suggested such as monitoring biogas characteristics in order to manage efficiently the biological process and prevent any dysfunction. Purification processes could also be assessed more accurately through an extended characterization (downstream and upstream) to

optimize the operational costs. Full biogas characterization could also help to adapt the purification processes according to new biomass or wastes used to produce biogas.

ACKNOWLEDGEMENTS Authors thanks RECORD for its support and ENGIE and Sita BioEnergies for granting the permission to access the biogas production sites.

REFERENCES [1] [2] [3]

[4] [5]

[6]

[7]

[8]

[9] [10]

[11] [12] [13]

[14] [15] [16]

P. McKendry, Energy production from biomass (part 2): conversion technologies, Bioresour. Technol. 83 (2002) 47–54. A. Demirbas, Biomass resource facilities and biomass conversion processing for fuels and chemicals, Energy Convers. Manag. 42 (2001) 1357–1378. A. Bauer, C. Leonhartsberger, P. Bösch, B. Amon, A. Friedl, T. Amon, Analysis of methane yields from energy crops and agricultural by-products and estimation of energy potential from sustainable crop rotation systems in EU-27, Clean Technol. Environ. Policy. 12 (2010) 153–161. EUROBSERV’ER, Biogas Barometer, Intelligent Energy Europe, 2014. http://www.energiesrenouvelables.org/observ-er/stat_baro/observ/baro224_Biogas_en.pdf. G. Brandel, N. Canfrene, J. Le Goff, Green Gas Grids : Une vision pour le biométhane en France pour 2030, ADEME (Agence de l’Environnement et de la Maîtrise de l’Energie) et GrDF, 2014. http://www.ademe.fr/sites/default/files/assets/documents/d4.1_roadmap_france_french.pdf. GrDF, Prescriptions techniques du distributeur GrDF, 2009. http://www.grdf.fr/documents/10184/1291504/Prescriptions_techniques_du_distributeur.pdf /7d6000a8-2e32-4491-b0e0-30f612f0ccde. R. Chiriac, J. De Araujos Morais, J. Carre, R. Bayard, J.M. Chovelon, R. Gourdon, Study of the VOC emissions from a municipal solid waste storage pilot-scale cell: Comparison with biogases from municipal waste landfill site, Waste Manag. 31 (2011) 2294–2301. M. Meres, P. Formisyn, P. Thesis : Analyse de la composition du biogaz en vue de l’optimisation de sa production et de son exploitation dans des centres de stockage des déchets ménagers, Science de l’environnement, École nationale supérieure des mines (Saint-Etienne) Uniwersytet Jagielloński (Cracovie), 2005. https://tel.archives-ouvertes.fr/tel-00793654/document. F. Defoort, S. Thiery, S. Ravel, A promising new on-line method of tar quantification by mass spectrometry during steam gasification of biomass, Biomass Bioenergy. 65 (2014) 64–71. I. Zdanevitch, J.-P. Jaeg, L’injection de biométhane issu de boues de STEP dans le réseau de distribution de gaz naturel en France, ANSES (Agence nationale de sécurité sanitaire alimenttion, environnement, travail), 2013. T. Assmuth, K. Kalevi, Concentrations and toxicological significance of trace organic compounds in municipal solid waste landfill gas, Chemosphere. 24 (1992) 1207–1216. J. Brosseau, M. Heitz, Trace gas compound emissions from municipal landfill sanitary sites, Atmos. Environ. 28 (1994) 285–293. B. Eklund, E.P. Anderson, B.L. Walker, D.B. Burrows, Characterization of Landfill Gas Composition at the Fresh Kills Municipal Solid-Waste Landfill, Environ. Sci. Technol. 32 (1998) 2233–2237. P.J. Young, A. Parker, The Identification and Possible Environmental Impact of Trace Gases and Vapours in Landfill Gas, Waste Manag. Res. 1 (1983) 213–226. T. Parker, J. Dottridge, S. Kelly, Investigation of the Composition and Emissions of Trace Components in Landfill Gas, Environment Agency, 2002. E. Bietlot, S. Garzaniti, C. Collart, Reseau de contrôle des C.E.T. en region wallonne - C.E.T. de Chapois Cinquième campagne de contrôle - Partie air (2012), Institut scientifique de service public - Département de la Police et des Contrôles (DG03), 2013.

[17] J. Poulleau, Caractérisation des BIOGAZ - Bibliographie - Mesures sur sites, INERIS (Institut National de l’environnement industriel et des risques), 2002. http://www.ineris.fr/centredoc/biogaz.pdf. [18] M.R. Allen, A. Braithwaite, C.C. Hills, Trace Organic Compounds in Landfill Gas at Seven U.K. Waste Disposal Sites, Environ. Sci. Technol. 31 (1997) 1054–1061. [19] R. Chiriac, J. Carre, Y. Perrodin, L. Fine, J.-M. Letoffe, Characterisation of VOCs emitted by open cells receiving municipal solid waste, J. Hazard. Mater. 149 (2007) 249–263. [20] S. Rasi, A. Veijanen, J. Rintala, Trace compounds of biogas from different biogas production plants, Energy. 32 (2007) 1375–1380. [21] S. Rasi, Thesis : Biogas composition and upgrading biomethane, Biological and environmental science, University of Jyväskylä, 2009. [22] G. Piechota, B. Igliński, R. Buczkowski, Development of measurement techniques for determination main and hazardous components in biogas utilised for energy purposes, Energy Convers. Manag. 68 (2013) 219–226. [23] H. Zhou, C. Wu, J.A. Onwudili, A. Meng, Y. Zhang, P.T. Williams, Polycyclic aromatic hydrocarbons (PAH) formation from the pyrolysis of different municipal solid waste fractions, Waste Manag. 36 (2015) 136–146. [24] E. Gallego, F.J. Roca, J.F. Perales, X. Guardino, E. Gadea, Development of a method for determination of VOCs (including methylsiloxanes) in biogas by TD-GC/MS analysis using SupelTM Inert Film bags and multisorbent bed tubes, Int. J. Environ. Anal. Chem. 95 (2015) 291–311. [25] E. Gallego, F.J. Roca, J.F. Perales, X. Guardino, Comparative study of the adsorption performance of a multi-sorbent bed (Carbotrap, Carbopack X, Carboxen 569) and a Tenax TA adsorbent tube for the analysis of volatile organic compounds (VOCs), Talanta. 81 (2010) 916– 924. [26] F. Tassi, G. Montegrossi, O. Vaselli, C. Liccioli, S. Moretti, B. Nisi, Degradation of C2–C15 volatile organic compounds in a landfill cover soil, Sci. Total Environ. 407 (2009) 4513–4525. [27] S. Mariné, M. Pedrouzo, R. Maria Marcé, I. Fonseca, F. Borrull, Comparison between sampling and analytical methods in characterization of pollutants in biogas, Talanta. 100 (2012) 145–152. [28] J. Raich-Montiu, C. Ribas-Font, N. de Arespacochaga, E. Roig-Torres, F. Broto-Puig, M. Crest, L. Bouchy, J.L. Cortina, Analytical methodology for sampling and analysing eight siloxanes and trimethylsilanol in biogas from different wastewater treatment plants in Europe, Anal. Chim. Acta. 812 (2014) 83–91. [29] K. Arrhenius, U. Johansson, Characterisation of contaminants in biogas before and after upgrading to vehicle gas, Svenskt Gastekniskt Center, Report SGC 246, 2012. http://www.sgc.se/ckfinder/userfiles/files/SGC246_eng.pdf. [30] J. LaRegina, J.W. Bozzelli, R. Harkov, S. Gianti, Volatile organic compounds at hazardous waste sites and a sanitary landfill in New Jersey, Environ. Prog. 5 (1986) 18–27. [31] M. Adahchour, J. Beens, R.J.J. Vreuls, U.A.T. Brinkman, Recent developments in comprehensive two-dimensional gas chromatography (GC×GC) IV. Further applications, conclusions and perspectives, TrAC Trends Anal. Chem. 25 (2006) 821–840. [32] C. Vendeuvre, F. Bertoncini, L. Duval, J.-L. Duplan, D. Thiébaut, M.-C. Hennion, Comparison of conventional gas chromatography and comprehensive two-dimensional gas chromatography for the detailed analysis of petrochemical samples, J. Chromatogr. A. 1056 (2004) 155–162. [33] J. Beens, J. Blomberg, P.J. Schoenmakers, Proper Tuning of Comprehensive Two-Dimensional Gas Chromatography (GC×GC) to Optimize the Separation of Complex Oil Fractions, J. High Resolut. Chromatogr. 23 (2000) 182–188. [34] L.L.. van Stee, J. Beens, R.J.. Vreuls, U.A.T. Brinkman, Comprehensive two-dimensional gas chromatography with atomic emission detection and correlation with mass spectrometric detection: principles and application in petrochemical analysis, J. Chromatogr. A. 1019 (2003) 89–99.

[35] F. Bertoncini, M. Courtiade-Tholance, D. Thiébaut, Gas chromatography and 2D-gas chromatography for petroleum industry: the race for selectivity, Editions Technip, 2013. [36] B. Omais, N. Charon, M. Courtiade, J. Ponthus, D. Thiébaut, A novel analytical approach for oxygen speciation in coal-derived liquids, Fuel. 104 (2013) 805–812. [37] B. Omais, T. Dutriez, M. Courtiade, N. Charon, J. Ponthus, H. Dulot, D. Thiébaut, SFC-GCxGC to Analyse Matrices from Petroleum and Coal, LCGC Eur. 24 (2011) 352–365. [38] J. Vial, D. Thiébaut, P. Sassiat, M.S. Beldean-Galea, M.J. Gomez Ramos, G. Cognon, S. Mallipattu, B. Teillet, M. Bouzige, Comprehensive two-dimensional GC of supercritical fluid and pressurized liquid extracts of tobaccos, J. Chromatogr. Sci. 48 (2010) 267–273. [39] G.S. Frysinger, R.B. Gaines, Comprehensive Two-Dimensional Gas Chromatography with Mass Spectrometric Detection (GC × GC/MS) Applied to the Analysis of Petroleum, J. High Resolut. Chromatogr. 22 (1999) 251–255. [40] P.J. Schoenmakers, J.L.M.M. Oomen, J. Blomberg, W. Genuit, G. van Velzen, Comparison of comprehensive two-dimensional gas chromatography and gas chromatography – mass spectrometry for the characterization of complex hydrocarbon mixtures, J. Chromatogr. A. 892 (2000) 29–46. [41] N.S. Tessarolo, L.R.M. dos Santos, R.S.F. Silva, D.A. Azevedo, Chemical characterization of biooils using comprehensive two-dimensional gas chromatography with time-of-flight mass spectrometry, J. Chromatogr. A. 1279 (2013) 68–75. [42] M. van Deursen, J. Beens, J. Reijenga, P. Lipman, C. Cramers, J. Blomberg, Group-Type Identification of Oil Samples Using Comprehensive Two- Dimensional Gas Chromatography Coupled to a Time-of-Flight Mass Spectrometer (GC×GC-TOF), J. High Resolut. Chromatogr. 23 (2000) 507–510. [43] M.E. Machado, E.B. Caramão, C.A. Zini, Investigation of sulphur compounds in coal tar using monodimensional and comprehensive two-dimensional gas chromatography, J. Chromatogr. A. 1218 (2011) 3200–3207. [44] J. Vial, A. Jardy, Experimental Comparison of the Different Approaches To Estimate LOD and LOQ of an HPLC Method, Anal. Chem. 71 (1999) 2672–2677.

Figure 1: GC x GC Chromatogram of standard mixture containing 89compounds: Mono-aromatics hydrocarbons (1-17); Poly-aromatics hydrocarbons (18-22); Alkanes (23-32); Cycloalkanes (33-38); Alcohols (39-41); Ketones (42-44); Esters (4548); Furans (49-51); Aldehydes (52-53); Sulfur organo-compounds (54-59); Halogen organo-compounds (60-74); Alkenes (75-76) ; Terpenes (77-81) and Siloxanes (82-89). Conditions: see experimental section. Figure 2: a) GC-MS chromatogram of standard mixture. b) Zoom between 21.9 and 38 min. c) GC x GC-MS Chromatogram of standard mixture between 21.4 and 38.6 min. Compounds: 1,3,5-trimethylbenzene (12); indane (18); naphthalene (19); butylcylohexane (36); phenol (41); hexanoic acid, butyl ester (48); benzothiophene (57); benzothiazole (59); 1,2dichlorobenzene (67); beta-pinene (79); limonene (80); dodecamethylpentasiloxane (87). Conditions: see experimental section. Figure 3: a) GC x GC Chromatogram of a biogas (site A) obtained in b) GC-MS chromatogram for the same biogas between 21 and 28 min and c) zoom of figure 3a. Compounds: p-ethyltoluene (10); o-ethyltoluene (11). Conditions: see experimental section. Figure 4: GC x GC-MS Chromatograms obtained for biogas from a) landfill, b) sewage sludge and c) agricultural residues – Same scale of intensity: Mono-aromatics hydrocarbons (1-17); Poly-aromatics hydrocarbons (18-22); Alkanes (23-32); Cycloalkanes (33-38); Alcohols (39-41); Ketones (42-44); Esters (45-48); Furans (49-51); Aldehydes (52-53); Sulfur organocompounds (54-59); Halogen organo-compounds (60-74); Alkenes (75-76) ; Terpenes (77-81) and Siloxanes (82-89). Conditions: see experimental section. Figure 5: Number of compounds by chemical families for each type of biogas analyzed Figure 6: GC x GC-MS Chromatograms obtained for a) biogas and b) biomethane from agricultural residues (site D) – Same scale of intensity: Mono-aromatics hydrocarbons (1-17); Poly-aromatics hydrocarbons (18-22); Alkanes (23-32); Cycloalkanes (33-38); Alcohols (39-41); Ketones (42-44); Esters (45-48); Furans (49-51); Aldehydes (52-53); Sulfur organocompounds (54-59); Halogen organo-compounds (60-74); Alkenes (75-76) ; Terpenes (77-81) and Siloxanes (82-89) Conditions: see experimental section.

Figr-1 Relative abundance

a

Time [min]

b

18 / 36 67 / 80

Relative abundance

59 / 87

48 57

19

12 / 41 / 79

Time [min] 6

Time 2 nd dimension [s]

c

4

57 41

67

18

19

59

12 2

48

79

80

36

87

0 21.4

23.4

25.4

27.4

29.4

Time 1 st dimension [min]

31.4

33.4

35.4

37.4

38.6

Figr-2

6

Time 2 nd dimension [s]

a 4

2

0

5

10

15

25

20

Time

30

35

40

1 st dimension [min]

Relative abundance

b

Time [min] 6

Time 2 nd dimension [s]

c 4

benzaldehyde 10 11

12

2

0 21.8

79

4-methylnonane 22.8

23.8

24.8

25.8

Time 1st dimension [min]

26.8

27.8

28.2

Figr-3 6

Landfill biogas (site A) – 1 Liter

Time 2 nd dimension [s]

a 4

3 4

2

6 5

2

82

0

84 10

5

18 13 11 17 14 65 15 9 10 80

8

64

43

41

69

7

78 77 79

83 25 15

27

85 26 36

20

53

52

38

86

30

25

28 89 40

35

Time 1 st dimension [min] 6

Sewage sludge biogas (site C) – 5 Liters

Time 2 nd dimension [s]

b 4

34

2

41

15

13

2

80

84

25

78

79

85

26

27

53 86

0 10

5

15

20

Time 6

40

35

[min]

Agricultual biogas (site D) – 5 Liters

c Time 2 nd dimension [s]

30

25

1 st dimension

4

69 2 2

56

11 17 13

6

3 4 5

9

7 78

43

85

18 15 80

53

26

27

86 89

0 5

10

15

20

25

Time 1 st dimension [min]

30

35

40

Figr-4 100 90

89

Number of compounds

80

Unknown compounds

Oxygenates

Alkanes and cycloalkanes

Monoaromatic hydrocarbons

Siloxanes

Alkenes and terpenes

Halogenates

Polyaromatic hydrocarbons

Sulfurs

70 60 50

46

40 31

30 20 10

32 28

30 26

27

22

19 13 8

5

3 2 1

6

11

7 3 2 2

4 5

2

4

7 1

0

Site A (Landfill)

Site B (Landfill)

Site C (Sludge) Biogas type

Site D (Agricultural)

4

Figr-5 6

Agricultural biogas (site D) – 5 Liters

Time 2 nd dimension [s]

a 4

69 2 2

56

11 13 17

6

3 4 5

9

7 78

43

85

18

15 80

26

53 27

86 89

0

10

5

15

20

25

30

35

40

Time 1 st dimension [min] 6

Agricultural biomethane (site D) – 20 Liters

Time 2 nd dimension [s]

b 4

2

0

5

10

15

20

Time

25

1 st dimension [min]

30

35

40

Table 1: 89 compounds of the standard mixture with CAS number and molecular weight (MW).

N°

Compounds

CAS

MW

Monoaromatic hydrocarbons 1

Benzene

71-43-2

78

2

Toluene

108-88-3

92

3

Ethylbenzene

100-41-4

106

4

p-xylene

106-42-3

106

5

m-xylene

108-38-3

106

6

o-xylene

95-47-6

106

7

Styrene

100-42-5

104

8

Isopropylbenzene / Cumene

98-82-8

120

9

n-propylbenzene

103-65-1

120

10

p-ethyltoluene

622-96-8

120

11

m-ethyltoluene

620-14-4

120

12

1,3,5-trimethylbenzene

108-67-8

120

13

o-ethyltoluene

611-14-3

120

14

1,2,4-trimethylbenzene

25551-13-7

120

15

p-isopropyltoluene / p-cymene

99-87-6

134

16

1,2,4,5-tetramethylbenzene

95-93-2

134

17

1,2,3-trimethylbenzene

526-73-8

120

Polyaromatic hydrocarbons 18

Indane

496-11-7

118

19

Naphthalene

91-20-3

128

20

2-methylnaphthalene

91-57-6

142

21

1-methylnaphthalene

90-12-0

142

22

Benzo[a]pyrene

50-32-8

252

Alkanes 23

n-heptane

142-82-5

100

24

n-octane

111- 65 -9

114

25

n-nonane

111 - 84 -2

128

26

n-decane

124 - 18 -5

142

27

n-undecane

1120 - 21 -4

156

28

n-dodecane

112- 40 -3

170

29

n-tridecane

629- 50 -5

184

30

n-tetradecane

629 - 59 - 4

198

31

n-pentadecane

629-62-9

212

32

2,4-dimethylpentane

108-08-7

100

Cycloalkanes 33

Cyclohexane, Methyl

108 - 87- 2

98

34

Cyclopentane, Ethyl

1640 - 89 -7

98

35

Cyclohexane, Ethyl

1678-91-7

112

36

Cyclohexane, Butyl

1678 - 93 -9

140

37

Trans-decalin

91-17-8

138

38

Cis-decalin

106-44-5

108

Oxygen-organic compounds Alcohols 39

1-Propanol, 2-Methyl

78- 83 -1

74

40

1-Butanol

71 -36 -3

74

41

Phenol

108-95-2

94

Ketones 42

2-Pentanone

107 -87- 9

86

43

2-Pentanone, 4-Methyl

108 - 10 -1

100

44

2-Hexanone

591- 78 -6

100

Esters 45

Acetic acid, Butyl ester

105- 54 -4

116

46

Butanoic acid, Ethyl ester

539 -82 -2

130

47

Butanoic acid, Butyl ester

123- 66- 0

144

48

Hexanoic acid, Butyl ester

626-82-4

172

Furans 49

2-Ethylfuran

3208-16-0

96

50

2,5-Dimethylfuran

625-86-5

96

51

Dibenzofuran

132-64-9

168

Aldehydes 52

Nonanal

124-19-6

142

53

Decanal

112-31-2

156

Sulfur-organic compounds 54

Thiophene

110 - 02 -1

84

55

Disulfure de dimethyle

624 - 92 - 0

94

56

2-methylthiophene

554-14-3

98

57

Benzo(b)thiophene

95-15-8

134

58

Dibenzothiophene

132-65-0

184

59

Benzothiazole

95-16-9

135

78 - 87 -5

112

Halogenorganic-compounds 60

1,2-Dichloropropane

61

1,2-Dichloroethane

107- 06 -2

98

62

Trichloroethylene

79- 01-6

129

63

Tetrachloroethylene

127- 18- 4

163

64

Chlorobenzene

108 -90 -7

112

65

1,4-Dichlorobenzene (p-)

106- 46 -7

147

66

1,3-Dichlorobenzene

541- 73 -1

147

67

1,2-Dichlorobenzene (o-)

95 -50- 1

147

68

1,1,2-Trichloroethane

79-00-5

133

69

1,1,2,2-Tetrachloroethane

79-34-5

167

70

Fluorobenzene

462-06-6

96

71

1-bromo-4-fluorobenzene

460-00-4

174

72

Pentafluorobenzene

363-72-4

168

73

Bromodichloromethane

75-27-4

164

74

Dibromochloromethane

124-48-1

208

Alkenes 75

1-octene

111-66-0

112

76

1-decene

872-05-9

140

Terpenes 77

Camphene

79-92-5

136

78

Alpha-pinene

7785-70-8

136

79

Beta-pinene

127- 91 -3

136

80

Limonene

5989- 27 -5

176

81

Terpinolene

586 - 62 -9

136

Siloxanes 82

Hexamethyldisiloxane (MM)

107- 46-0

162

83

Octamethyltrisiloxanes (MDM)

107-51-7

236

84

Hexamethylcyclotrisiloxane (D3)

541- 05- 9

222

85

Octamethylcyclotetrasiloxane (D4)

556- 67 -2

296

86

Decamethylcyclopentasiloxane (D5)

541 - 02 -6

370

87

Dodecamethylpentasiloxane (MD2M)

141-63-9

384

88

Decamethyltetrasiloxane (MD3M)

141-62-8

310

89

Dodecamethylcyclohexasiloxane

540-97-6

444

Table 2: Samples investigated in this work

Site

Biomass type

A

MSW landfill

B

MSW landfill

C

Anaerobic Digester of sewage Sludge

D

Anaerobic Digester of agricultural residues

Sample gas Raw biogas Treated biogas Raw biogas Pre-treated biogas Treated biogas Raw biogas Treated biogas Biomethane Raw biogas Biomethane

Collected volume of gas (L) 1 20 1 5 20 5 20 20 5 20

Table 3: Experimental total VOC concentration by chemical families in three kinds of biogases analyzed in this study (range in ng.mL-1). For the families of compounds investigated in this work, references where quantitative data can be found in the literature are also indicated.

Biogas Landfill Sewage sludge Agricultural

Oxygenates

Halogenates

41,42

64,65,69

Alkanes

Terpenes

Siloxanes

25,26,27

78,79,80

85,86,89

0-300

Monoaromatic hydrocarbons2,3 100-16000

100-300

0-1400

0-4200

0-1300

0-100

n.d

0-400

0-60

0-40

0-210

n.d

n.d

0-300

n.d

0-450

0-20

Table 4: Estimation of the elimination yield of trace compounds in biogas in function of treatment processes

Site A

B

C

D

Biomass type Non-hazardous waste of landfill

Non-hazardous waste of landfill Sludge of wastewater treatment plant Agricultural residues

Treatment process

Elimination rate (E)

Input gas

Output gas

Raw biogas

Treated biogas

Adsorption on oxides + Filtration on activated charcoal

95 %

Raw biogas

Pre-treated biogas

Adsorption on oxides

86 %

Raw biogas

Treated biogas

Raw biogas

Treated biogas

Raw biogas

Biomethane

Raw biogas

Biomethane

Adsorption on oxides + Filtration on activated charcoal Filtration on activated charcoal Filtration on activated charcoal + Membranes Filtration on activated charcoal

94% 96% 99% 99%