Air-Blown Entrained-Flow Gasification of Biomass: Influence of

Article pubs.acs.org/EF

Air-Blown Entrained-Flow Gasification of Biomass: Influence of Operating Conditions on Tar Generation Ludwig Briesemeister,*,† Michael Kremling,† Sebastian Fendt,† and Hartmut Spliethoff†,‡ †

Department of Mechanical Engineering, Institute for Energy Systems, Technical University of Munich, Boltzmannstraße 15, 85748 Garching, Germany ‡ ZAE Bayern, Walther-Meißner-Straße 6, 85748 Garching, Germany ABSTRACT: The formation of tars in gasifiers based on fluidized- or fixed-bed technology is a major problem in biomass gasification. By pretreating biomass using hydrothermal carbonization (HTC), entrained-flow gasification becomes applicable. Oxygen-blown entrained-flow gasifiers (EFGs) operate at very high process temperatures, leading to an almost tar-free syngas. However, in decentralized small-scale units, preferably air is used as the gasification agent, which, in turn, causes lower gasifier temperatures. The specific impacts of air-blown gasification conditions and fuel properties of biocoal from HTC on tar formation require particular attention. Therefore, in this work, tar formation under air-blown gasification conditions is investigated using solid-phase adsorption at an electrically heated EFG with temperatures of 900−1300 °C and different air/fuel equivalence ratios λ. Furthermore, tars are measured in the hot syngas of an industrial-like autothermal EFG. HTC biocoals of various feedstocks (beech, biogenic residuals, municipal waste, and green waste), raw biomass (corn cobs), and fossil fuel (Rhenish lignite) are used as fuels. The results show that the main influencing parameter on tar loading in the syngas is the temperature, whereas the residence time and λ have less impact. However, in autothermal operation, the choice of λ controls the gasifier temperature and, thus, effectively affects the resulting tar loading. Identified tar compounds are mainly light polycyclic aromatic hydrocarbons, of which naphthalene is the most frequently occurring. At 1300 °C, tar loading is reduced to less than 0.2 g/Nm3, which allows for direct syngas use in internal combustion engines.

1. INTRODUCTION On the basis of the findings of the Intergovernmental Panel on Climate Change (IPCC), the massive exploitation of fossil fuels since the start of industrialization and the associated release of greenhouse gases have caused an increase in global temperatures.1 To limit negative consequences, such as rising sea levels or glacial melting, the European Union aims at a reduction of greenhouse gas emissions by 80−95% by 2050 compared to 1990.2 As a result of the inherent CO2 balance of biomass, its utilization for heat and power production has the potential of contributing to this goal. For the conversion of lignocellulosic biomass, state-of-the art solutions in decentralized applications are direct combustion and gasification in combination with heat and power generation. Most biomass gasifiers are fixed- or fluidized-bed-type because of the comparable low fuel quality requirements of these technologies. However, dependent upon specific fuel properties and gasifier design, the syngas of both types is polluted with tar compounds that result in highmaintenance and gas-cleaning costs. Typical tar loading in the syngas is between 1 g/Nm3 (downdraft) and 100 g/Nm3 (updraft) for fixed-bed gasifiers and around 10 g/Nm3 for fluidized-bed gasifiers.3 Entrained-flow gasifiers (EFGs) operate at higher process temperatures, which provoke thermal cracking of tars and produce syngas that is often considered tar-free.4,5 Thus far, EFGs are mainly used to convert fossil coals for large-scale synthesis processes, which require the utilization of pure oxygen as a gasifying agent as well as an operation at elevated pressures.5 Recently, the entrained-flow gasification of biomass is investigated for both pressurized operation using oxygen6,7 and air-blown operation at atmospheric pressure,8,9 © XXXX American Chemical Society

whereby the latter is preferable in decentralized applications for heat and power generation.10 To reach the high fuel quality required for the application of EFGs, biomass is thermally pretreated by torrefaction7,11,12 or hydrothermal carbonization (HTC).8,13,14 A short overview of recent activities concerning entrained-flow gasification of biomass is given by Schneider et al.15 The generation of tars from biomass gasification is wellinvestigated. A common definition considers hydrocarbons larger than benzene (>C6) as tars.16 During pyrolysis, tars are formed from the macromolecules of the fuel. Thermal degradation of cellulose, hemicellulose, and lignin produces primary tar compounds, such as levoglucosan and furfurals, that are further decomposed to phenolics and olefins.3 A further temperature increase favors the formation of aromatic and polycyclic aromatic hydrocarbons (PAHs) by recombination of the primary and secondary tars. PAHs can have high dew points of up to 500 °C.17 In the literature, different methods for measuring tars are used. The two most frequently applied methods are the tar protocol,16 which is based on gravimetric measurement of tars, and the solid-phase-adsorption (SPA) method.18 The SPA method is capable of detecting heterocyclic and aromatic tar compounds as well as PAHs, whereby the measurement of high-volatile compounds and very large PAHs is restricted. To improve the SPA method, recent research aims at the utilization Received: June 23, 2017 Revised: August 30, 2017 Published: August 30, 2017 A

DOI: 10.1021/acs.energyfuels.7b01801 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Figure 1. Simplified flow sheets of BabiTER (left) and BOOSTER (right).

cally heated EFG. Furthermore, gasification in an industrial-like autothermal EFG is applied to evaluate the transferability of the aforementioned results.

of a combination of different adsorption materials, which additionally allows for the measurement of volatile compounds, such as benzene.19,20 As a result of the large number of different occurring species, tars are classified in the literature to describe their nature and evolution in a comprehensive way. Evans and Milne use a classification based on the temperature-dependent occurrence of the individual tar compounds, which differentiates between primary, secondary, and tertiary tars.21,22 Another system developed at the Energy Research Centre of the Netherlands (ECN) assigns tars as a result of their chemical composition to five classes depending upon the size and number of their aromatic ring structures.23,24 In general, the most important influencing factors on tar evolution are temperature, pressure, gasifying agent, stoichiometry, residence time, and catalytic effects.25 Increasing temperatures cause a reduction of the total tar loading as a result of the conversion of oxygen-containing tars, such as phenols, cresols, and furans, that only exist up to 800 °C.26 At higher temperatures, tars consist mainly of aromatic compounds, such as benzene, naphthalene, and phenanthrene.26 Many authors are investigating the influences of the process conditions on tar composition and loading in the syngas of fixed- and fluidized-bed gasifiers, whereas only few works address the tar formation under EFG conditions. Yu et al. investigate the gasification of cellulose, hemicellulose, and lignin between 800 and 1100 °C and identify PAHs as the predominant species in EFG.27 Here, increases in both temperature and stoichiometric oxygen ratio reduce the tar loading. Hernández et al. investigate the gasification of marc of grape using air, steam, and mixtures of both in an EFG at temperatures up to 1200 °C.28 However, a detailed analysis of the tar evolution in autothermal EFG with pretreated biomasses and temperatures relevant for industrial EFG is only known to the authors from the work of Kremling et al.7 Herein, applied reaction conditions are representative of oxygen-blown gasifiers. In this work, tar loading and composition are investigated under air-blown gasification conditions using different biocoals from HTC, raw biomass, and lignite. Tars are measured using the SPA method to obtain quantitative information on tar loading and composition.18 Separate influences of gasification temperature and stoichiometry are evaluated using an electri-

2. EXPERIMENTAL SECTION Figure 1 shows the two different gasification test rigs used for the experiments. On the left side, the baby high-temperature entrainedflow reactor (BabiTER) consists of an electrically heated ceramic tube (1) 1.48 m long with an inner diameter of 40 mm. It is designed for atmospheric pressure and allows for gasification experiments in varying gas atmospheres at well-defined temperatures. Mass flow controllers (MFCs) adjust nitrogen and oxygen flows, so that the overall composition of the gasification agent is like air. Three individually controlled heating zones ensure a constant temperature along the reaction zone. Tar samples are taken after a constant plug-flow residence time of 1 s by varying the mass flows of the fuel and the gasification agent. The biomass pilot-scale entrained-flow gasifier (BOOSTER) on the right is designed for autothermal operation with a thermal fuel input of around 100 kW. The system uses a pneumatic dense-phase conveying system for supplying the fuel that is inserted with a swirl burner (2) together with preheated air (3). A refractory contained in the reactor vessel (5) builds the reaction chamber (4), which has an inner diameter of 250 mm and a total length of 2.3 m. At the end of the reactor, the syngas is water-quenched (6), filtered (8), and burnt in a flare (10). A regulation valve (9) maintains a constant system pressure of 0.2 barg. Tar samples are withdrawn 1.9 m downstream from the burner mouth through a heated sampling port (7). A constant thermal fuel input of 70 kW is used for the experiments. More information on both test rigs can be found in other papers.7,8,29 The applied SPA method for tar analysis was developed and successfully used in the context of fluidized-bed gasification by Mayerhofer.4,30,31 A similar sampling probe for SPA measurements is used for BabiTER and BOOSTER trials, as shown in Figure 2. Syngas is withdrawn from the hot reaction zone, and a glass fiber filter removes particles. After a constant flow is ensured, a syngas volume of

Figure 2. Schematic of SPA sampling. B


Article

Energy & Fuels Table 1. Calibrated Tar Compounds with ECN Classification and Minimum Detection Limit class

species

abbreviation

ctar,min (mg/Nm3)

class

species

abbreviation

ctar,min (mg/Nm3)

2 2 2 3 3 3 4

phenol cresol (o) cresol (m) toluene xylene (o) styrene indene

pho cre (o) cre (m) tol xyl (o) styr ind

18.01 16.40 17.64 13.21 12.41 13.43 12.72

4 4 4 4 4 5 5

naphthalene biphenyl fluorene anthracene phenanthrene fluoranthene pyrene

naph bip fluo ant phen fluor pyr

12.65 13.16 13.77 14.42 14.68 15.68 15.92

Table 2. Chemical and Physical Properties of the Fuels HTC compost (HCo)

HTC beech (HBe)

C H N S O (calculated)

59.96 3.67 0.68 0.57 20.15

55.62 4.25 0.35 0.31 25.56

moisture volatile yield ash fixed carbon (fc)

11.32 55.11 3.66 29.91

8.51 56.03 5.40 30.07

LHVfuel

22.00

20.34

d50

53.1

87.3

Na2O MgO Al2O3 SiO2 P2O5 SO3 K2O CaO TiO2 Fe2O3 traces

9.4 12.7 2.6 5.4 0.6 17.4 1.9 36.8 0.3 2.3 10.6

1.0 0.7 4.8 56.7 1.2 1.2 1.6 14.3 0.3 5.0 13.3

HTC green waste (HGW)

HTC municipal waste (HMW)

Ultimate Analysis (wt %, ar) 60.74 5.28 0.70 0.38 26.61 Proximate Analysis (wt %, ar) 2.29 59.71 3.99 34.01 Lower Heating Value (MJ/kg, ar) 22.75 Particle Size Analysis (μm) 56.2 Ash Composition (wt %) 1.2 0.8 4.5 29.2 3.5 3.1 1.8 40.1 0.3 4.0 11.6

100 mL is drawn over an amino-phase column (Supelco-Supelclean LC-NH2, 3 mL/500 mg) within 60 s using a syringe. Before and after sampling, the SPA columns are cooled, thus ensuring complete adsorption of the tars.32 To avoid condensation of tars, everything is electrically heated to above 200 °C in the sampling line. Tar samples are eluted from the amino phase using dichloromethane and analyzed in a gas chromatograph equipped with a flame ionization detector (GC−FID, Agilent model 7890A).30 An Agilent HP-5 column (30 m, 320 μm, 0.25 μm) is used with H2 as the mobile phase. Each GC analysis is repeated 3 times, and averaged values are used. The GC−FID is calibrated for 14 tar species frequently occurring in biomass gasification. Table 1 shows calibrated species, their classification according to the ECN system, and the minimum detection limit ctar,min for each compound. The calibration method is based on an external standard. Therefore, a representative mixture of the 14 tar species is dissolved in dichloromethane, thus representing a stock solution. For the quantitative calibration, various dilutions from 1:10 to 1:833 of the stock solution in dichloromethane are used, representing tar loadings of 0.0024−0.4816 g/Nm3 of each species in the syngas. Measurement errors of the SPA analysis are around 7% for all species.4 The given values for the tar loading of all calibrated species are quantitative and indicated as identified (id.). In addition, unknown

Rhenish lignite (RL)

corn cobs (CC)

35.28 2.74 1.85 1.48 11.06

60.30 3.79 0.67 0.52 17.48

44.73 5.52 0.27 0.14 39.57

11.13 45.10 36.46 7.31

13.32 46.19 3.92 36.56

8.51 73.53 1.25 16.71

13.21

22.06

16.13

52.2

56.5

220.8

0.6 5.3 8.0 27.1 5.5 2.9 1.4 34.1 0.6 3.9 10.6

3.9 15.7 2.6 2.2 0.8 17.1 3.0 32.3 0.2 7.6 14.6

2.4 3.0 1.7 11.4 4.6 0.8 59.7 0.0 0.1 0.2 16.2

species detected by the GC−FID are also considered and indicated as not identified (n.id.). Their concentrations in the syngas are approximated by a mean calibration factor and are therefore perceived as semi-quantitative values. As a result of the utilization of only one SPA stage (amino phase), high-volatile compounds cannot be detected completely.19 Thus, the given total tar loadings represent minimum values. However, particularly, tar compounds with high tar dew points are measured almost completely, which provides the technically most decisive information. The benzene signal is ignored in the data evaluation and is therefore not considered in the given tar-loading values. The fuels used for the investigations cover a wide range of different input materials. Four fuels represent HTC biocoals based on a mixture of biogenic residues for composting (HCo), beech wood (HBe), green waste (HGW), and municipal waste (HMW). HCo, HBe, and HGW were produced in a HTC demonstration plant by SunCoal Industries. Applied HTC conditions were around 210 °C at 20−21 bar with a residence time of 3 h. The Spanish company Ingelia delivered HMW. After HTC treatment, the fuels were mechanically dewatered, thermally dried, and ground to reach the final water content and particle size. Rhenish lignite (RL) was purchased as filter dust directly from the manufacturer and required no further processing. In this work, it represents a well-known fossil fuel. Corn cobs (CC) represent C


Article

Energy & Fuels

gasified under EFG conditions at 1200 °C, the overall tar loading level is quite low.28 In this work, tars are measured by means of a liquid chromatography method. Tar loading is reduced from 15.3 to 1 g/Nm3 by increasing λ from 0.19 to 0.67. Besides the incomplete representation of benzene, toluene, and xylene (BTX) by the SPA method, especially the HTC pretreatment of the biomass seems to provoke a significant reduction of the tar amount. To provide a more detailed view, Figure 4 shows the influences of the operating parameters on tar composition.

an untreated biomass with a high share of cellulose and hemicellulose delivered by JRS Rettenmaier. Table 2 summarizes the chemical and physical properties of all fuels.

3. RESULTS A detailed investigation of the individual influences of stoichiometry and temperature on tar evolution is carried out for HBe, HGW, and RL in BabiTER. Stoichiometry is varied in terms of the air/fuel equivalence ratio λ between 0.3 and 0.5. Therefore, λ is calculated by the ratio of the actual oxygen input ṁ O2 to the oxygen required for stoichiometric combustion ṁ O2,stoic (eq 1). ṁ O2 λ= ṁ O2,stoic (1) The temperature is varied between 900 and 1300 °C to cover a wide range of conditions relevant for air-blown EFG. Furthermore, 900 °C represents the upper limit of applicable temperatures in fluidized-bed gasifiers, thus allowing for good comparability of the results. Figure 3 shows the influences of λ

Figure 3. Influence of the temperature and λ on tar loading in BabiTER.

and temperature on total tar loading subdivided into identified and unknown species. Both increasing temperatures and λ cause a decrease of the tar loading, whereby the temperature effect is much stronger. By increasing λ, a part of the tar reduction results from the syngas dilution by additional nitrogen of air. This effect represents the actual impact of airblown gasification on syngas and is therefore not removed from the data by recalculation. When the fuels are compared, RL has the lowest tar loading, especially at low λ. Lignite is known to release few tar during pyrolysis.33 HBe causes the highest values with more than 1 g/ Nm3 at 900 °C. The most probable reason for the comparably high tar loadings of the HTC fuels is remaining lignin structures that do not convert completely during HTC.34,35 Lignin has low reactivity under gasification conditions and still decomposes at temperatures around 900 °C.36 At 1100 and 1300 °C, lignin is completely converted and the resulting tars are thermally cracked. In comparison to the results of a similar investigation by Hernández et al., in which raw biomass is

Figure 4. Influence of the temperature and λ on tar composition in BabiTER for HGW (top), HBe (middle), and RL (bottom).

Here, the ECN classification system is used considering the calibrated tar compounds listed in Table 1. Tars of all classes occur, but those of class 4 are the dominant species for all fuels. Looking at the results for RL, the temperature increase significantly reduces the amount of aromatic hydrocarbons, whereas the amount of class 2 tars is almost constant. An increase in λ reduces tars of classes 3 and 4, particularly at lower temperatures. The effect on class 4 tars is mainly caused by a D


Article

Energy & Fuels Table 3. Detailed Results of Tar Measurements in BabiTER tar loading (mg/Nm3)

operating conditions fuel RL

T (°C)

λ

tol

xyl (o)

ind

cre (o)

nap

flu

fluor

pyr

phen

id.

n.id.

total

900

0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5 0.3 0.4 0.5

60 40 45 20 0 0 0 0 0 52 47 43 33 0 27 20 19 19 75

31 0 0 0 0 0 0 0 0 33 38 33 0 0 0 0 0 22 40

0 0 0 0 0 0 0 0 0 33 47 39 0 0 0 0 0 0 51

37 34 36 36 34 35 39 34 34 0 0 0 0 0 0 0 0 0 34

235 170 180 80 64 32 0 0 0 406 450 324 162 0 60 0 0 0 482

24 25 25 26 26 27 28 26 25 26 26 24 0 0 0 0 0 0 26

0 0 0 0 0 0 0 0 0 91 0 60 29 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 101 0 61 65 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 105 44 72 22 0 0 0 0 0 50

388 268 286 163 124 94 67 60 59 847 652 656 310 0 87 20 19 40 758

0 0 88 102 51 71 42 37 72 231 302 243 276 67 106 48 23 24 285

388 268 374 265 175 165 109 97 131 1078 953 898 586 67 194 69 42 64 1043

42 31 0 0

29 0 0 0

0 0 0 0

35 39 34 35

256 201 66 44

24 25 26 24

0 0 0 0

0 0 0 0

0 0 0 0

386 295 125 103

132 175 63 72

518 470 189 174

0

0

0

35

0

27

0

0

0

61

71

133

1100

1300

HBe

900

1100

1300

HGW

900

1100

1300

methodology as described for BabiTER. Before sampling, a stable operating point is awaited to ensure thermal equilibrium of the refractory. As the main operating parameter, λ is varied between 0.35 and 0.6. Besides the measurement of tars, for the BOOSTER trials, also the dry syngas composition is measured using an online gas analyzer (Sick S700). The syngas composition represents averaged values over a period of at least 10 min in stable operation. A more detailed description of the experimental procedure is given in a previous paper.8 That paper presents measurement values for HGW in BOOSTER, which are considered here for reasons of comparability. Figure 5 shows the dry syngas compositions in BOOSTER for all fuels. By increasing λ, the evolution of the syngas is mainly influenced by the increasing temperature as well as additional oxygen and nitrogen. An increasing temperature causes an increase of char conversion and affects the syngas composition by the water-gas shift reaction.5 Additional oxygen is consumed by the combustion of the volatiles, which, in turn, increases the temperature. The general behavior is similar to results considering the equilibrium of the water-gas shift reaction.38 In contrast to oxygen-blown processes, the share of nitrogen increases by increasing λ. By only considering chemical equilibrium, methane would not be present as a result of the high temperatures reached. However, as a result of kinetic limitations of the methane conversion,39 it still reaches significant values, especially at low λ. Syngas from raw biomass (CC) has the highest methane content, whereas lignite produces only few methane. The HTC fuels show an intermediate methane level. This can be attributed to the reduction of the volatile yield in comparison to the raw fuel.40 As a result of the very high ash content in the

reduction in naphthalene, whose reactivity under EFG conditions develops strongly between 1000 and 1300 °C.37 The behavior of HGW is similar to RL, but the amount of class 4 tars is more than twice as high. For HGW and RL, the only identified tar species at 1300 °C are fluorene and cresol (o). Unlike HGW and RL, HBe produces no tars of class 2, but at low temperatures, class 5 tars, namely, fluoranthene and pyrene, are identified. Although their concentrations are comparably low, these large PAHs have high dew points and can cause severe problems in further processing steps of the syngas. The existence of class 5 tars is in accordance with observations from other authors gasifying raw beech under comparable conditions.7 Therefore, the aforementioned presumption of incomplete conversion of the biomass structure during HTC is confirmed. A further temperature increase to 1300 °C causes complete conversion of class 4 and 5 tars, so that toluene and xylene (o) are the only remaining species. Table 3 summarizes the detailed results of the tar measurements for the BabiTER trials. The calibrated species cresol (m), phenol, and styrene are not detected in the test series because their secondary cracking reactions are already completed at temperatures around 850 °C.21 Furthermore, neither anthracene nor biphenyl occurs, which is why these species are not listed in the table. In comparison to the results of Hernández et al.,28 measured xylene values are rather low, which indicates a possible underestimation as a result of the use of the SPA method. A significant share of the n.id. compounds probably is acenaphthylene, which is reported by Hernández et al. for similar reaction conditions.28 The autothermal test series in BOOSTER involves all fuels listed in Table 2. Tar sampling and analysis follow the same E


Article

Energy & Fuels

cause lower gasifier temperatures as a result of their low LHV and, in the case of HMW, a high ash content. The measured inner wall temperatures of HMW are up to 100 °C lower than for, e.g., HBe or HGW. Furthermore, on the basis of an ash-free consideration, both fuels have very high volatile yields. Because during pyrolysis tars form from the complex volatile matter, this is an additional reason for the comparable high tar loadings. By increasing λ, tars are efficiently reduced, so that, for all fuels, except for HMW, the tar loading is below 0.2 g/Nm3 at λ = 0.4. Similar values are reported by Kremling et al. for oxygen-blown operation in the case of lignite or pretreated biomass as fuels, whereas raw biomass causes tar loadings of up to 3.5 g/Nm3.7 In BOOSTER, the tar loadings of HBe, HGW, and RL are comparable to the results at λ ≥ 0.4 and T ≥ 1100 °C in BabiTER. Therefore, the considerably longer residence times achieved in BOOSTER do not seem to have a crucial effect on the tar loading if compared to other parameters, such as the temperature.26 Calculated tar dew points range between ≤ −20 and 55 °C, whereby the tar dew point mostly correlates to the tar loading. However, for CC at λ = 0.35, tar loading is much smaller than for HMW, but at the same time, the dew point is higher. For HMW, total tar loading decreases with increasing λ, but initially, the dew point increases. The reason for these observations is the formation of phenanthrene, which is exclusively present for these trials and has a high dew point of 340 °C. For all other trials, the only identified species are naphthalene and toluene. Because only class 3 and 4 tars occur in BOOSTER, a detailed consideration of the influences on tar classes is not shown. Instead, Figure 7 shows the tar loading of all occurring species for HMW, which is the most problematic fuel in terms of tar-related issues. Most of the tar consists of naphthalene. Phenanthrene is increasingly formed until λ = 0.45 and decreases after that. Besides the variation of λ, previously published results show that the addition of water steam has the potential to improve the performance of an autothermal EFG in terms of the fuel conversion and the cold gas efficiency.8,42 To investigate its influence on tar loading, the addition of water steam is applied for RL, HBe, and HGW in BOOSTER. Therefore, we use the steam addition ratio Π, which is calculated as the steam mass flow ṁ st related to the fuel mass flow ṁ f (eq 2).

Π=

Figure 5. Influence of λ on the syngas composition in BOOSTER.

ṁ st ṁ f(ar)

(2)

Steam addition is applied for a constant λ = 0.4, meaning that the supplied air flow is constant for a given fuel and the additional steam mass flow increases with Π. Similar to the results from these fuels without steam addition, the only identified tar species are toluene and naphthalene. Figure 8 shows tar loadings and tar dew points. In comparison to airblown gasification, the addition of steam tends to increase both the tar loadings and the tar dew points slightly. The reason for this is probably the reduced temperature as a result of the moderating effect of the steam,8 which results in decreased naphthalene conversion. All tar dew points are below 15 °C, therefore, from a technical point of view, the effect of steam addition on tar evolution is negligible. An increase in tar loading by steam addition is also reported by Hernández et al. for the gasification of marc of grape.28

fuel, the process temperatures for HMW are low, which leads to low methane conversion. Moreover, low temperatures promote the reduction of CO in favor of CO2 by the water-gas shift reaction. Figure 6 shows the results of the tar measurements for all fuels. Besides tar loading, the figure shows corresponding tar dew points calculated with the ECN tar dew point calculator considering the identified species.41 Because the utilization in a gas engine is the most probable application of the syngas from air-blown EFG, condensation of tars is a critical issue,17 whereas gaseous tar species are less problematic. Tar dew points shown as −20 °C involve calculated values of −20 °C and below. As expected, the highest tar loading occurs for low λ and therefore at the lowest temperatures. Tar loading is strongly dependent upon the fuel type, and CC and HMW have the highest values. In comparison to the other fuels, CC and HMW F


Article

Energy & Fuels

Figure 6. Influence of λ on tar loading in BOOSTER.

Figure 7. Influence of λ on detailed tar composition for HMW in BOOSTER. Figure 9. Relationship between the methane content in the syngas and total tar loading in BOOSTER.

function of the dry methane content for all fuels. Overall, an increasing methane content seems to indicate an increase in tar loading, but the effect is particularly dependent upon the different fuel properties. CC, for instance, produces much methane but only few tars, whereas HMW reaches comparably high tar loadings already at around 2 vol % CH4. For RL, there is no clear trend. Therefore, by considering only the methane content, a reliable statement on tar evolution is not possible. Nevertheless, for a given fuel, it provides valuable information on the stability of the process and an estimation of operating limits, where tar evolution becomes critical. In autothermal gasification, the fuel quality and operating parameters greatly influence the gasifier temperature. Because the temperature is considered the most prevailing influence parameter on tar generation, a correlation between both values is also examined for the BOOSTER trials. Therefore, the temperature T3 serves as a reference value (see Figure 1). It is measured for all trials and represents the inner wall temperature halfway down the reaction zone. Figure 10 shows the relationship of total tar loading as a function of the wall temperature for all BOOSTER trials. Here again, increasing temperatures clearly reduce tar loading, but the absolute amount is fuel-dependent. As already observed from the BabiTER trials, the temperature range for strong tar development is below 1000 °C. However, for fuels such as RL that only produce few tars, tar loading does not increase much, even at

Figure 8. Influence of steam addition on tar loading in BOOSTER at constant λ = 0.4.

Finally, simplified correlations are investigated on the basis of the experimental results from the BOOSTER trials. In industrial gasifiers, a sophisticated tar measurement is often not available and online monitoring of the process is desirable. Therefore, in the literature, easily measurable variables are considered indicators of tar evolution. Examples of these are acetylene and benzene43 as well as ethane or methane.44 In this work, the methane content in the syngas is measured for all trials, and its suitability as a tar indicator is therefore examined. Figure 9 shows an overview of the total tar loadings as a G


Energy & Fuels

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Ludwig Briesemeister: 0000-0002-2444-615X Michael Kremling: 0000-0002-8616-430X Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the German Federal Ministry for Economic Affairs and Energy (BMWi) in the framework of the project “FLUHKE” (FKZ 03KB074B). The authors thank the students involved in this work (Roland Balint and Martijn van Stiphout). The support of the Technical University of Munich (TUM) Graduate School is gratefully acknowledged.

Figure 10. Relationship between the gasifier wall temperature (T3) and total tar loading in BOOSTER.

■

comparably low temperatures, and a temperature correlation is only of limited significance. It is worth noting that the wall temperature T3 does not represent the gas temperature, which, in fact, can be significantly higher. In addition, pyrolysis and tar formation already occur in the flame region in the upper part of the reactor. Therefore, the crucial temperatures probably exceed those measured at T3 by far.

4. CONCLUSION AND OUTLOOK This study focuses on tar generation during entrained-flow gasification of HTC biocoal, raw biomass, and lignite specifically for the use of air as a gasification agent. Tar measurements using the SPA method are performed in an electrically heated EFG (BabiTER) as well as in an industriallike autothermal EFG (BOOSTER). Identified tar compounds are mainly light PAHs, of which naphthalene is the predominant species. Therefore, the majority of the tars are formed by recombination of primary and secondary pyrolysis products. Of all studied operating parameters, the temperature shows the greatest effect on tar evolution, whereas the residence time and λ have less influence. In the BabiTER trials, the measured tar loading of more than 1 g/Nm3 at 900 °C decreases to less than 0.2 g/Nm3 with a temperature increase to 1300 °C. In the BOOSTER trials, tar loading for most fuels is in the range of 0−0.2 g/Nm3, but for HMW, which has an ash content of almost 40 wt %, tar loading reaches a maximum value of 1 g/Nm3. The reason for this is the reduced gasifier temperatures as a result of the high proportion of inert material. A similar but less pronounced observation is made when adding steam to the gasifier. Thereby, tar loading and tar dew points increase slightly as well. In overall terms, syngas from autothermal air-blown EFG is not problematic in terms of possible tar condensation issues, e.g., when applying a gas engine. Measured tar dew points are below 15 °C for most of the experiments. In future work, improved measurements of BTX compounds should be carried out. Furthermore, measuring C2 compounds instead of methane in the syngas might provide a better indication of the presence and amount of tars. Tar measurements at varying pressures and residence times are desirable to obtain a more detailed understanding of tar formation in autothermal EFG.

■

NOMENCLATURE BabiTER = baby high-temperature entrained-flow reactor BOOSTER = biomass pilot-scale entrained-flow gasifier CC = corn cobs ECN = Energy Research Centre of the Netherlands EFG = entrained-flow gasifier fc = fixed carbon FID = flame ionization detector GC = gas chromatograph HBe = hydrothermal carbonized beech wood HCo = hydrothermal carbonized residues for composting HGW = hydrothermal carbonized green waste HMW = hydrothermal carbonized municipal waste HTC = hydrothermal carbonization id. = identified IPCC = Intergovernmental Panel on Climate Change LHV = lower heating value MFC = mass flow controller n.id. = not identified PAHs = polycyclic aromatic hydrocarbon RL = Rhenish lignite SPA = solid-phase adsorption REFERENCES

(1) Intergovernmental Panel on Climate Change (IPCC). Climate Change 2013: The Physical Science Basis; IPCC: New York, 2013. (2) Europäische Kommission. Energiefahrplan 2050; Publications Office of the European Union: Brussel, Belgium, 2011. (3) Milne, T.; Evans, R.; Abatzoglou, N. Biomass Gasifier “Tars”: Their Nature, Formation, and Conversion; National Renewable Energy Laboratory (NREL): Golden, CO, Nov 1998; NREL/TP-570-25357. (4) Mayerhofer, M. Teerentstehung und Teerminderung bei allothermer Wirbelschichtvergasung; Verlag Dr. Hut: München, Germany, 2014. (5) Higman, C.; van der Burgt, M. Gasification; Elsevier/Gulf Professional Publishing (GPP): Amsterdam, Netherlands, 2008. (6) Weiland, F.; Hedman, H.; Marklund, M.; Wiinikka, H.; Ö hrman, O.; Gebart, R. Energy Fuels 2013, 27, 932−941. (7) Kremling, M.; Briesemeister, L.; Gaderer, M.; Fendt, S.; Spliethoff, H. Energy Fuels 2017, 31, 3949−3959. (8) Briesemeister, L.; Kremling, M.; Fendt, S.; Spliethoff, H. Chem. Eng. Technol. 2017, 40, 270−277. (9) Hernández, J. J.; Aranda-Almansa, G.; Bula, A. Fuel Process. Technol. 2010, 91, 681−692. (10) Kobayashi, N.; Tanaka, M.; Piao, G.; Kobayashi, J.; Hatano, S.; Itaya, Y.; Mori, S. Waste Manage. 2009, 29, 245−251. (11) Couhert, C.; Salvador, S.; Commandré, J.-M. Fuel 2009, 88, 2286−2290. H


Article

Energy & Fuels

(44) Kaltschmitt, M.; Hartmann, H.; Hofbauer, H. Energie aus Biomasse; Springer: Berlin, Germany, 2016.

(12) Weiland, F.; Nordwaeger, M.; Olofsson, I.; Wiinikka, H.; Nordin, A. Fuel Process. Technol. 2014, 125, 51−58. (13) Tremel, A.; Stemann, J.; Herrmann, M.; Erlach, B.; Spliethoff, H. Fuel 2012, 102, 396−403. (14) Erlach, B.; Harder, B.; Tsatsaronis, G. Energy 2012, 45, 329− 338. (15) Schneider, J.; Grube, C.; Herrmann, A.; Rönsch, S. Fuel Process. Technol. 2016, 152, 72−82. (16) Deutsches Institut für Normung e.V. (DIN). BiomassevergasungTeer und Staub in ProduktgasenProbenahme und analytische Bestimmung; DIN: Berlin, Germany, 2006; DIN CEN/TS 15439. (17) Zwart, R. Gas Cleaning: Downstream Biomass Gasification: Status Report 2009; Energy Research Centre of the Netherlands (ECN): Petten, Netherlands, 2009. (18) Brage, C.; Yu, Q.; Chen, G.; Sjöström, K. Fuel 1997, 76, 137− 142. (19) Osipovs, S. Anal. Bioanal. Chem. 2008, 391, 1409−1417. (20) Osipovs, S. Fuel 2013, 103, 387−392. (21) Evans, R. J.; Milne, T. A. Energy Fuels 1987, 1, 123−137. (22) Evans, R. J.; Milne, T. A. Energy Fuels 1987, 1, 311−319. (23) Kiel, J.; van Paasen, J.; Devi, L.; Ptasinski, K.; Janssen, F.; Meijer, R.; Berends, R.; Temmink, H.; Brem, G.; Padban, N., Bramer, E. Primary Measures To Reduce Tar Formation in Fluidised-Bed Biomass Gasifiers: Final Report SDE Project P1999-012; Energy Research Centre of the Netherlands (ECN): Petten, Netherlands, 2004; ECN-C--04014. (24) Devi, L. Catalytic Removal of Biomass Tars: Olivine as Prospective in-Bed Catalyst for Fluidized-Bed Biomass Gasifiers; Technische Universiteit Eindhoven: Eindhoven, Netherlands, 2005; DOI: 10.6100/IR583960. (25) Devi, L.; Ptasinski, K. J.; Janssen, F. J. Biomass Bioenergy 2003, 24, 125−140. (26) Kinoshita, C. M.; Wang, Y.; Zhou, J. J. Anal. Appl. Pyrolysis 1994, 29, 169−181. (27) Yu, H.; Zhang, Z.; Li, Z.; Chen, D. Fuel 2014, 118, 250−256. (28) Hernández, J. J.; Ballesteros, R.; Aranda, G. Energy 2013, 50, 333−342. (29) Tremel, A.; Haselsteiner, T.; Kunze, C.; Spliethoff, H. Appl. Energy 2012, 92, 279−285. (30) Mayerhofer, M.; Fendt, S.; Spliethoff, H.; Gaderer, M. Fuel 2014, 117, 1248−1255. (31) Mayerhofer, M.; Mitsakis, P.; Meng, X.; de Jong, W.; Spliethoff, H.; Gaderer, M. Fuel 2012, 99, 204−209. (32) Ortiz Gonzalez, I.; Perez Pastor, R. M.; Sanchez Hervas, J. M. Anal. Bioanal. Chem. 2012, 403, 2039−2046. (33) Suuberg, E. M.; Peters, W. A.; Howard, J. B. Ind. Eng. Chem. Process Des. Dev. 1978, 17, 37−46. (34) Gunarathne, D. S.; Mueller, A.; Fleck, S.; Kolb, T.; Chmielewski, J. K.; Yang, W.; Blasiak, W. Energy Fuels 2014, 28, 1992−2002. (35) Lin, Y.; Ma, X.; Peng, X.; Yu, Z.; Fang, S.; Lin, Y.; Fan, Y. Fuel 2016, 181, 905−915. (36) Yang, H.; Yan, R.; Chen, H.; Lee, D. H.; Zheng, C. Fuel 2007, 86, 1781−1788. (37) Jess, A. Fuel 1996, 75, 1441−1448. (38) Weiland, F.; Wiinikka, H.; Hedman, H.; Wennebro, J.; Pettersson, E.; Gebart, R. Fuel 2015, 153, 510−519. (39) Dufour, A.; Valin, S.; Castelli, P.; Thiery, S.; Boissonnet, G.; Zoulalian, A.; Glaude, P.-A. Ind. Eng. Chem. Res. 2009, 48, 6564−6572. (40) Funke, A.; Ziegler, F. Biofuels, Bioprod. Biorefin. 2010, 4, 160− 177. (41) Energy Research Centre of the Netherlands (ECN). Tar Dew Point Calculation: Complete Model; ECN: Petten, Netherlands, 2009; http://www.thersites.nl/completemodel.aspx (accessed May 17, 2017). (42) Tremel, A. Reaction kinetics of solid fuels during entrained flow gasification. Ph.D. Dissertation, Technische Universität München, München, Germany, 2013. (43) Ö hrman, O. G. W.; Molinder, R.; Weiland, F.; Johansson, A.-C. Environ. Prog. Sustainable Energy 2014, 33, 699−705. I


Air-Blown Entrained-Flow Gasification of Biomass: Influence of

Recommend Documents