Study of the equilibrium of air-blown gasification of biomass to coal evolution fuels

Energy Conversion and Management 128 (2016) 120–133

Contents lists available at ScienceDirect

Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

Study of the equilibrium of air-blown gasification of biomass to coal evolution fuels Enrico Biagini Dipartimento di Ingegneria Civile e Industriale (DICI), Università di Pisa, L. Lazzarino, 56126 Pisa, Italy

a r t i c l e

i n f o

Article history: Received 28 July 2016 Received in revised form 31 August 2016 Accepted 22 September 2016

Keywords: Pyrolysis Energy Combustion Wood Residues Wastes

a b s t r a c t A non-stoichiometric equilibrium model based on the minimization of the Gibbs free energy was used to study the isothermal and adiabatic air-blown gasification of solid fuels on a carbonization curve from fossil (hard/brown coals, peat) to renewable (green biomasses and cellulose) fuels, including torrefied biofuels. The maps of syngas composition, heating value and process efficiency were provided as functions of equivalent ratio (oxygen-to-fuel ratio) in the range 0–0.6, temperature in 500–2000 K, and a fuel parameter, which allowed different cases to be quantitatively compared. The effect of fuel moisture, unconverted carbon and conditions to limit the tar formation was also studied. Cold gas efficiency >0.75 can be achieved for coals at high temperature, using entrained beds (which give low unconverted carbon), and improved by moisture/added steam. The bigger efficiency of green biomasses is only potential, as the practical limits (high temperature required to limit tar formation, moisture content and unconverted carbon in small gasifiers) strongly reduce the gasification performance. Torrefied biomasses (and plastics having an intermediate fuel parameter between coals and green biomasses) can attain high efficiency also in real conditions. The results shown in this work can be useful to evaluate the most promising feedstock (depending on its composition and possible pre-treatment/upgrading), define the operating conditions for maximizing the syngas heating value or the global efficiency, assess the techno-economicenvironmental feasibility of a gasification-based system. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction The use of unconventional fuels (such as battle coals, peats, tar sands) and renewable biofuels (forest-agro-food residues, energy crops, algae) has found current applications and studies on most promising options and energy efficient solutions increased significantly in the last two decades. Furthermore, the emergence of ‘pretreated’ fuels (torrefied, steam exploded, and hydrotreated biomasses, biochar, see for instance [14,35,48,36,17,4]) and mixtures of them with fossil derived (co-combustion, waste-toenergy) have extended the range of compositions available for energy fuels. The use of these solid fuels can be seen as a valuable option to face the near depletion of traditional fossil resources, extend/enhance the availability of local energy sources, reduce the global warming/climate change. Due to the different properties of these fuels, the energy conversion options are numerous. Among them, the gasification showed high efficiency and versatility in feedstock selection, technology choice and product opportunity. The gasification was applied to coals, biomasses and industrial/municipal/agricultural wastes to E-mail address: [email protected] http://dx.doi.org/10.1016/j.enconman.2016.09.068 0196-8904/Ó 2016 Elsevier Ltd. All rights reserved.

give a combustible/synthetic gas with a greater efficiency than the direct combustion (see reviews of [34,22,2,51,6]). The plant size ranged from small gasifiers for dislocated applications (50– 500 kW) to large IGCC plants (100–1000 MW). The gasifier configuration can be based on fixed (downdraft, countercurrent), fluidized (bubbling, circulating), or entrained beds. The use of different feedstocks and operating conditions (temperature, pressure, residence time, gasifying agent, catalyst) can give a wide range of gasification products/byproducts: heat, electricity, biochar, syngas for further conversion, e.g. production of methane, hydrogen, ammonia, Fischer-Tropsch fuels [29,15]. The most common gasifying agents are air, pure oxygen, steam and CO2, also in mixtures. The gasification is an ensemble of homogeneous and heterogeneous reactions, starting from the pyrolysis of the solid fuels and involving the gasifying species and pyrolysis products. The entire process can be autothermal, in which the reactor temperature is achieved by the balance of exothermic and endothermic reactions, as in a partial oxidation, or allothermal, in which an external energy source provides the heat necessary for the reactor to achieving the desired temperature, for instance the hot sand in cir-

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

121

Nomenclature a, b CB CGE EBP ER exp F G gas G0f H/C i IFRF IGCC j LHV M

correlation parameters subscript for Carbon Boundary Cold Gas Efficiency Evolution Biomass Parameter (EBP = H/C ⁄ O/C) Equivalent Ratio abbreviation for experimental abbreviation for fuel when followed by index j Gibbs function superscript for gaseous state free energy of formation on a molar basis hydrogen to carbon mass ratio generic index for gaseous product International Flame Research Foundation Integrated Gasification Combined Cycle generic index for fuel F Lower Heating Value (dry) [MJ/kg for solids, MJ/m3 for gas] moisture content of the fuel (mass basis)

culating fluidized beds. In the choice of the most profitable option for the gasification configuration, the following are general points: d

d

d

the presence of oxygen gives an exothermic contribution, so that the higher the oxygen content in the feeding stream, the higher the achieved temperature, but the lower the heating value of the produced syngas; the presence of H2O gives an endothermic contribution, so that the higher the water in the feed (as fuel moisture as well as added steam), the lower the temperature in the reactor, but the higher the potential heating value of the produced syngas; the presence of nitrogen from air reduces the achievable temperature inside the reactor, dilutes the syngas and reduces its heating value.

The composition and properties of the syngas can be related to the feedstock characteristics, gasifying agent and gasifier conditions, for estimating the process parameters and optimizing the efficiency. Innumerable models exist to simulate the gasification and predict the composition of the syngas. Their classification can be based on the accuracy, complexity, and specificity (see for instance the reviews by [20,40,37]). The starting point of all models is the equilibrium approach, which assures a general applicability. It consists in calculating the composition of gasification products at the thermodynamic equilibrium, so it is based on the feedstock composition, in terms of ultimate analysis, and process conditions (temperature, pressure, gasifying agent-to-fuel ratio), but it is independent of the feedstock structure and gasifier design/operation. The calculation may follow two approaches: stoichiometric, which defines the equilibrium constants of constituent gasification reactions, and non-stoichiometric, which minimizes the Gibbs free energy of the gasification products. The results of the two approaches were proved to be equivalent (see [40] and references therein). Therefore, the non-stoichiometric approach is more advantageous as it only requires the definition of the list of chemical species expected in the product mixture [5]. The equilibrium model is simple, is based on non-specific process parameters, requires a small calculation effort and allows a wide range of conditions (process parameters and fuel compositions) to be studied. These are the reasons for its widespread use as a general tool or starting point for modified versions and comparison with more detailed models. Its predictability has been

mod n N O/C PKS Qext R SFDB SGP sol T T0 TG tot UC w

abbreviation for model molar amount number of gaseous species oxygen to carbon mass ratio Palm Kernel Shells external heat flow rate needs [MW] ideal gas constant Solid Fuel DataBase Specific Gas Productivity [kg syngas/kg fuel] superscript for solid state absolute temperature [K] initial temperature of the feed [K] Thermo-Gravimetry subscript for total amount of gas unconverted carbon (mass basis) flow rate [kg/s or kg/h for solids, m3/s or m3/h for gas]

validated only for specific fuels and single reactor configurations. In those cases, the limited reliability and accuracy of the equilibrium model results have been imputed to the fact that real plants operate under conditions, which may be far from equilibrium. The equilibrium approach is indeed useful for predicting what is thermodynamically attainable, indicates the maximum efficiency of gasification [39] and can be a guide for process design, evaluation and optimization [28]. The aim of this work is to validate an equilibrium nonstoichiometric model for fuels ranging from coals to biomasses and quantify its accuracy. Hence the equilibrium model is used to study the gasification process and evaluate the effect of fuel composition, temperature and oxygen-to-fuel ratio on syngas composition, heating value and process efficiency. A fuel parameter is defined to verify a comprehensive connection of different fuels, involving also brown coals, torrefied biomasses and plastics. The maps of the results obtained in isothermal and adiabatic conditions may represent useful tools for producers and operators of small and medium gasifiers to evaluate the most promising feedstock (depending on its composition and possible pretreatments), predict syngas composition and heating value, and determine the most efficient conditions (temperature, oxygen-to-fuel ratio) of gasification. They are also useful for process and system analysis, as preliminary study and application to techno-economic and environmental assessments (see for instance [47]).

2. Model description Every fuel was treated as an exclusively CAHAO system, since carbon, hydrogen and oxygen are by far the most abundant elements in coal and biomass gasification (see similar assumptions in [16,28,39,3]). Each fuel was characterized by the H/C and O/C mass ratios from the ultimate analysis normalized on a CAHAO basis. The fuel parameter EBP (Evolution Biomass Parameter) was defined as the product of H/C and O/C. The values of H, C and EBP are listed in Table 1. The smallest values of EBP are specific of hard coals, the greatest ones are specific of green biomasses. Intermediate values of EBP may denote brown coals, peats and torrefied biomasses. 18 fuels were selected to run the simulations described in this work. The fuel F1 is a low volatile coal (Tower coal, from [11]), for which EBP = 0.0040, F2–F6 are different rank

122

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

Table 1 Elemental composition, Evolution Biomass Parameter and Lower Heating Value of the fuels. Fuel

C (wt%)

H (wt%)

EBP (w/w)

LHV (MJ/kg)

f1-hard coal f2-hard coal f3-hard coal f4-hard coal f5-hard coal f6-brown coal f7-brown coal f8-torrefied PKS f9-torrefied PKS f10-torrefied PKS f11-peat f12-PKS f13-pine wood f14-vine pruning f15-rice husks f16-corn cobs f17-sorghum f18-cellulose PMMA PET Eucalyptus Olive pits Yorkshire coal Koflach coal

87.3 84.0 81.3 78.2 75.5 72.7 69.0 67.0 60.1 58.9 56.5 54.4 53.0 50.8 49.4 47.6 46.0 44.4 60.0 62.5 49.0 51.3 86.5 68.7

3.17 3.59 3.89 4.20 4.45 4.69 4.98 5.13 5.57 5.65 5.78 5.89 5.97 6.07 6.14 6.22 6.29 6.36 8.00 4.17 5.90 6.70 5.33 5.79

0.0040 0.0063 0.0087 0.0121 0.0157 0.0200 0.0272 0.0318 0.0530 0.0577 0.0683 0.0791 0.0871 0.1015 0.1119 0.1268 0.1419 0.1584 0.0711 0.0356 0.1108 0.1069 0.0059 0.0312

32.1 31.1 30.2 29.2 28.3 27.3 25.9 25.2 22.5 22.1 21.1 20.3 19.7 18.9 18.3 17.6 16.9 16.3 25.4 21.9 17.8 19.8 34.4 26.8

coals (from [45]), F7 is a brown coal (from [38]), F8–F10 are palm kernel shells under different levels of torrefaction [27], F11 is a peat (from [38]), F12 is the untreated palm kernel shells, F13– F17 are different lignin-cellulosic biomasses, namely, pine wood, vine pruning, rice husks, corn cobs, and sorghum from TG biomassdevo database [12], F18 is cellulose, for which EBP = 0.1584. They are represented by the cross symbols in Fig. 1. The specific selection of the fuels was operated with the presumption that they may represent an ideal evolution curve from biomass to coal, starting from pure cellulose, through different grades of lignincellulosic biomasses and then different levels of torrefied/carbonized biofuels, to different rank coals. This may represent the natural path of coalification, which consists in the intensification of dehydration and decarboxylation and a reduction in the H/C and O/C ratios. The selected fuels were connected in the Van Krevelen graph (Fig. 1) with a translated hyperbole having the following interpolating equation:

ðO=C  aÞ2 ðH=CÞ2  ¼ 1 with a ¼ 0:8913 and b ¼ 0:0713 2 a2 b 0.16

0.12

Koflach coal

0.1

H/C

Olive pits

PMMA

0.14

BIOMASS PKS

coal

0.06

PET

F1

0.02 0

F18 Cellulose Eucalyptus

Peat

0.08 Yorkshire

0.04

ð1Þ

Torrefied PKS

Lignites

HARD COALS 0

0.5

1

1.5

O/C Fig. 1. Van Krevelen graph of the biomass evolution curve (black curve), fuels from F1 to F18 in the evolution curves (cross symbols, labels in Italics), coal and biomass regions (grey dashed contours), and alternative fuels (circle symbols, labels in bold) studied for comparison.

Although many biomass evolution curves can be speculatively defined, the one proposed in Eq. (1) allows real fuels in a wide range of composition to be studied and related. As shown in the results and discussion sections, the main conclusions of this work are actually based on the fuel parameter EBP, independently of the specific evolution curve. The simulations were carried out also with some fuels outside the evolution curve of Fig. 1 (circle symbols in the figure, for two coals, two plastics and two biofuels) to corroborate these conclusions. The Lower Heating Value LHV (in MJ/kg) of the fuels was obtained by the following equation [33]:

LHVfuel ¼ 0:341C þ 1:104H  0:12O

ð2Þ

where C, H and O are expressed as percentage mass fractions of the material. The calculated values are shown in Table 1. The gasification process was assumed to be operated with air as gasifying agent, under 1 atm pressure, at constant temperature T and then heated or cooled as needed from external sources, or in adiabatic conditions with no heat exchange with external bodies. The equivalence ratio ER was used to define the level of gasification and represents the oxygen-to-fuel ratio. ER is defined as the ratio between the actual oxygen in the gasifier feed and the oxygen required for a stoichiometric oxidation of the fuel. Air and fuel were assumed to be at 300 K and 1 atm for the input to the equilibrium reactor. The reference case was studied for a feeding rate of 100 kg/h of dry fuel, that is moisture M = 0. Some cases were studied with M = 0.1, while maintaining the same feeding rate of dry fuel. The gasification reactor was based on a single stage of equilibrium (for more details see also [10]). The solver scheme is shown in Fig. 2. The input consisted of the fuel characteristics and operating conditions. In particular, for the isothermal case the value of T was constrained, while for the adiabatic case Qext = 0. The model solved the mass and energy balances and minimized the Gibbs energy of the system:

G¼

  N X ngas 0 i þ nsol G0f ;sol ngas G þ RTln gas f ;i i n tot i¼1

ð3Þ

The unknowns (products yields and T or Qext) were determined, starting from initial guess values, according to the Nelder-Mead

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

123

Fig. 2. Solver scheme of the gasification model.

method . All calculations were executed in Matlab. The calculations were reiterated for all fuels in the evolution biomass curve, and the results of each fuel were taken as guess values for the fuel in the subsequent reiteration. This was beneficial to reduce the computational effort. The reaction products were divided into gaseous and solid phases, namely syngas (formed of CO, CO2, H2, H2O, CH4, O2 and N2) and char (formed of carbon, assumed in the graphitic form). Nitrogen was assumed to be not involved in any chemical reaction and participate only to the energy balance. Similar assumptions were previously adopted by Cairns and Tevebaugh [16], Li et al. [28], Prins et al. [39], Baratieri et al. [5], Azzone et al. [3], Tapasvi et al. [46]. Products in the liquid phase and further ones in the gaseous phase (e.g. light C2-C4 hydrocarbons) were not considered as previous [10] and literature works [5,44] revealed they were in negligible concentrations in the results of equilibrium models under the usual conditions of gasification. This simplification was assumed to have insignificant influence on the results of the equilibrium model, while positively reduced the computational effort and thus allowed to perform a vast number of simulations for all the fuels listed above (18 fuels in the evolution curve from hard coals to cellulose + 6 alternative fuels outside this curve), for a temperature range between 500 and 2000 K, for ER values between 0 and 0.6. The results were expressed in terms of molar fractions of the products (excluding N2, although it is present in the syngas). O2 was consumed in all cases, and no O2 was detected in the syngas. The reference case was studied with no constrain on the potential C conversion, while some simulations were also studied to compare the case of imposed Unconverted Carbon UC = 0.1. The syngas was characterized in terms of LHV (MJ/m3), which was calculated as the weighted sum of the contributions from the chemical species:

LHV gas ¼

N X ni LHV i n i¼1 tot

ð4Þ

With the assumptions made above, the positive contribution to LHVgas is given by the presence of CO, H2 and CH4. It is worth noting that ntot is the sum of all gaseous species in the syngas, N2

included. The efficiency of the gasification process was evaluated by introducing the Cold Gas Efficiency CGE defined as:

CGE ¼

wgas LHV syngas wfuel LHV fuel þ Q ext

ð5Þ

where w is the flow rate and Qext the external heat flow rate needed to maintain the reactor temperature in case of isothermal simulations: this term was accounted for in the previous expression only when Qext > 0, while it was omitted in case of Qext < 0 (this means no heat recovery is performed in the reactor, although extra heat is available, and thus Qext is considered as thermal dispersion). Qext = 0 for the case of adiabatic equilibrium. For each fuel, wgas, wfuel and Qext are proportional, so that the Cold Gas Efficiency is a normalized efficiency of the process with a more general significance, and can be considered independent of the assumed fuel rate of 100 kg/h mentioned at the beginning of this section. 3. Results 3.1. Validation of the equilibrium model As stated in the introduction, the equilibrium model has been used in many literature works to simulate the gasification data for specific fuels and with a certain approximation for different reactor configurations, but the proof of a general adequacy can be hardly found. The validation of the same model with tests carried out on fuels from hard coals to green biomasses allows to prove the general quality of this model and quantify its accuracy as a prediction tool. The validation requires a complete dataset of fuel characteristics, operating conditions and experimental results. The model developed in the previous section is validated here with the experimental data from home made and literature data on different fuels. The home made data refer to the experimental campaigns carried out at CRIBE (Center of biomass-to-energy Research in Pise – Italy) and summarized in Biagini et al. [8], Biagini et al. [9], Biagini et al. [10]. Different biomass fuels, namely vine pruning, rice husks, wood pellets and corn cobs, were gasified in an air-blown downdraft reactor (feed rate 50–80 kg/h) with ER in the range 0.25–0.4. The work of Li et al. [28] reported the gasi-

124

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

fication tests of a sub-bituminous coal in an air-blown circulating fluidized bed reactor (20–30 kg/h), with ER in the range 0.3–0.5. The work of Kwapinska et al. [26] reported the gasification tests of torrefied and green Miscanthus in an air-blown bubbling fluidized bed gasifier (0.4–0.6 kg/h) with ER in the range 0.18–0.3. The results of the simulations are shown in Fig. 3, where the comparison of the experimental and model data is plotted for coals, torrefied and green biomasses. The Specific Gas Productivity SGP is predicted with a good accuracy for all cases (mean deviation 2.7%, max deviation 12%), while the deviation between the experimental and model gas composition, as shown for the CO to H2 ratio, is more evident for the bubbling fluidized bed, which operated at relatively lower temperatures than the other two reactors. In all cases however, the mole fraction of CH4 is under predicted as the experimental data are in the range 0.005–0.05, while the equilibrium model predicts values of CH4 less than 0.005 at the relatively high gasification temperatures. Nevertheless, as a sort of counterbalanced equilibrium among the CO/H2/CH4 species, the Lower Heating Value of the gas is generally predicted with a good accuracy (mean deviation 3.6%, max deviation 12%), as can be seen in Fig. 3. As a general consideration, the equilibrium model can be used to predict the SGP and LHVgas of different fuels and gasifiers, while a certain discrepancy, which can be large when the reactor conditions are far from those of equilibrium (see also the considerations in the discussion section), should be expected in the syngas composition.

–

–

–

–

3.2. Results of the isothermal case for fuels on the biomass evolution curve The simulations of the equilibrium model in isothermal conditions were carried out for all selected fuels (F1–F18), temperatures between 500 and 2000 K, and ER between 0.2 and 0.4. These ranges include most of the conditions used in gasification plants with different reactor configurations and different fuels. The resulting mole fractions of all gasification products (N2 excluded) are shown as cumulative values in Fig. 4 for some fuels. The results are shown as contours in the T-EBP plane for the case of ER = 0.3 in Fig. 5. The results of the isothermal equilibrium simulations can be summarized in the following general observations: – char (C solid), methane, CO2 and H2O are favored at relatively low temperatures, while hydrogen and CO are favored at relatively high temperatures. Based on the stoichiometric reactions, it means that carbon gasification reactions are favored at high temperature. For instance, when studying the Boudouard equilibrium (C + CO2 = 2CO), a temperature can be defined (in the range 900–940 K, according to Cairns and Tevebaugh [16] and

CO/H2, SGP, LHVgas

7

CO/H2 exp CO/H2 mod

–

Li et al. [28] to discriminate between reactants (C + CO2 favored at lower temperatures) and products (CO favored at higher temperatures); the relatively high availability of oxygen (deriving from the operation conditions at high ER as well as from the natural content of green biomasses) may favor a significant presence of CO2 and H2O (oxidation products of CO and H2) even at high temperature, while it makes the production of char to decrease; the production of char can be significant even at high temperatures for fuels at low values of EBP. For instance, the product char from the gasification of F1 at 2000 K and ER = 0.3 is 0.227 (mole fraction of C among all gasification products, excluding N2), 0.140 for F4, 0.0295 for F8. The temperature at which the conversion of char is complete (carbon boundary temperature, see [16,28,39]) is 1090 K for F12, 1005 K for F15, 995 K for F18; the production of CH4 has a maximum for temperatures between 610 and 625 K (at ER = 0.3) and its value increases with EBP: the maximum mole fraction of CH4 (among all gasification products, excluding N2) is 0.012 at 625 K for F1, 0.021 at 620 K for F4, 0.057 at 612 K for F15, 0.065 at 612 K for F18. The maximum value decreases with ER: for F4 it passes from 0.029 to 0.017 (from 0.0675 to 0.049 for F15) when ER passes from 0.2 to 0.4, with negligible variations in the corresponding temperature; the production of H2 increases with the gasification temperature to reach a stable value for fuels with EBP < 0.035 at ER = 0.3 (for EBP < 0.08 at ER = 0.2, for EBP < 0.012 at ER = 0.4), while it shows a maximum for fuels with EBP > 0.035. The maximum mole fraction of H2 (among all gasification products, excluding N2) is 0.178 for F1, 0.242 for F4, 0.362 at 1025 K for F15, 0.364 at 1020 K for F18, in all cases corresponding to the carbon boundary temperature. This value slightly decreases with ER: for F4 it passes from 0.243 to 0.235 (from 0.400 to 0.319 for F15) when ER passes from 0.2 to 0.4; the production of CO increases with the gasification temperature and achieve a stable value at a temperature less than 2000 K (the maximum temperature studied in this work) only for specific ranges of ER and EBP: for ER = 0.2 and EBP < 0.101 the mole fraction of CO (among all gasification products, excluding N2) reaches an asymptotic value with the maximum value registered of 0.598 for F13; for ER = 0.3 and EBP < 0.053 CO reaches an asymptotic value with the maximum value registered of 0.657 for F8; for ER = 0.4 and EBP < 0.012 CO reaches an asymptotic value with the maximum value registered of 0.776 for F2.

The results of the isothermal simulations were elaborated to study LHVgas according to the assumptions made in the previous section. The results are shown in the contours of Fig. 6. In general,

SGP exp SGP mod

LHV exp LHV mod

6 5 4 3 2 1 0

Fig. 3. Comparison of the experimental and model results for different gasification tests (in the abscissa is reference/EBP/ER, where reference L: [28], K: [26], B: [10].

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

125

Fig. 4. Gasification products for the isothermal equilibrium of fuel (a) F1, (b) F4, (c) F8, (d) F12, (e) F15, and (f) F18, as functions of the temperature at ER = 0.20 (continuous black curves), 0.30 (continuous grey curves), 0.40 (dashed black curves). M = 0, UC = 0, mole fractions of products normalized without N2.

for a fixed value of ER and for each fuel, LHVgas increases with the temperature of gasification up to a certain temperature, above which LHVgas remains practically constant. It is worth noting that in correspondence with this temperature, the production of char remains constant too or becomes negligible, for the fuels with high EBP. For a fixed temperature and for each fuel, LHVgas decreases with ER. For instance, for F4 the maximum LHVgas is 6.33 and 5.30 MJ/m3 (7.74 and 4.48 MJ/m3 for F15) when ER is 0.2 and 0.4, respectively. For a fixed value of ER and a fixed temperature (below 900–1100 K, depending on ER), LHVgas increases with EBP. For a fixed value of ER and a fixed temperature (above 900–1100 K, depending on ER), LHVgas shows a maximum for a value of EBP depending on the conditions. The maximum value of LHVgas = 8 MJ/m3 registered in this work among all the isothermal simulations is for ER = 0.2, T > 1200 K and EBP between 0.080 and 0.085, corresponding to palm kernel shells and pine wood, respectively. This result is only indicative and should not been generalized, as the condition of ER = 0.2 is at the boundary of the region studied with the model in the isothermal conditions. The maximum of LHVgas decreases with ER, along with the range of EBP corresponding to that value. For ER = 0.3 the maximum LHVgas is 6.5 MJ/m3, achieved for EBP around 0.04 (corresponding to torrefied biomasses), for ER = 0.4 the maximum LHVgas is 5.35 MJ/m3, achieved for EBP around 0.009 (bituminous coal). The LHVgas is a basic index of the quality of the syngas produced. The process efficiency can be quantified by the CGE (see

Eq. (5)), which accounts for, besides the LVHgas, the specific gas productivity SGP (that is the amount of syngas produced by gasification of a fixed amount of fuel), the heating value of the solid fuel and the heat needed to reach the gasification temperature. The results of CGE for the isothermal simulations are shown in the contours of Fig. 6. To draw some general considerations, besides the observations on LHVgas described above, it is important to note that SGP increases with ER and T, to a maximum value which is then maintained constant, while the external heat needed for achieving the gasification temperature increases with T and decreases with ER. For fixed values of ER and T, SGP decreases with EBP increasing, as well as the heat needed for achieving the gasification temperature, as in both cases the air required for gasification decreases with EBP increasing. Considering different ranges of EBP, T and ER, the optimal conditions can be found for maximizing the CGE. For ER = 0.2, very high values of CGE can be found, well above 0.80. These values can be achieved in a quite wide region of conditions, i.e. for temperatures between 1000 and 1400 K and fuels with EBP > 0.08 (biomasses). Lower values can be found outside this region: at lower temperatures LHVgas is too low, at higher temperatures the heat needed is too high, for decreasing values of EBP LHVgas decreases and LHVfuel (in the denominator in Eq. (5)) increases. The maximum CGE (>0.85) can be found for EBP > 0.10 and temperatures between 1000 and 1100 K. For ER = 0.3 the mean values of CGE show a general decrease with respect to the previous case,

126

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

Fig. 5. Gasification products for the isothermal equilibrium: mole fractions of (a) char (solid), (b) CH4, (c) CO, (d) CO2, (e) H2, and (f) H2O, as contours of the fuel parameter EBP and equilibrium temperature. ER = 0.30, M = 0, UC = 0, mole fractions of products normalized without N2.

although values above 0.80 can be found in a relatively narrower region of conditions. The temperature is comprised between 1100 and 1200 K for fuels with EBP = 0.04–0.08, and between 1000 and 1100 K for fuels with EBP = 0.08–0.16. Even lower values of CGE can be observed for ER = 0.4. In this case the maximum values of CGE are between 0.75 and 0.78, and are achieved at temperatures between 1100 and 1400 K for fuels with EBP = 0.005–0.03, and between 1000 and 1300 K for fuels with EBP = 0.03–0.07. At lower temperatures LHVgas is too low, at higher temperatures the heat needed is too high, at higher values of EBP both the LHVgas and SGP decrease. Therefore, as a general observation, the region for the maximum CGE is for intermediate temperatures (1000– 1400 K) and moves to lower values of EBP when ER increases. The values of maximum CGE, which can be found in the T-EBP plane, decreases with ER.

3.3. Results of the adiabatic case for fuels on the biomass evolution curve The adiabatic simulations were performed by varying the value of the gasification air and thus increasing ER from 0 to 0.6. The results are shown in Fig. 7 for some fuels as cumulative mole fractions of all gasification products (N2 excluded). The results of the adiabatic simulations are qualitatively similar to those obtained in isothermal conditions, if one considers that in general the greater the value of ER, the higher the temperature: – the production of C (solid char) decreases with ER for each fuel, and decreases with EBP for a fixed ER. The conversion of char is complete for ER > 0.42 (F1), 0.34 (F8), 0.30 (F18). Accordingly,

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

127

Fig. 6. Gasification results for the isothermal equilibrium (M = 0, UC = 0): (a, c, e) LHVgas (in MJ/m3 in the contour curves) and (b, d, f) CGE as contours of the fuel parameter EBP and temperature at (a, b) ER = 0.20, (c, d) 0.30, (e, f) 0.40.

the carbon boundary conditions are defined when exactly enough oxygen is added to achieve a complete char conversion (see [16,28,39]); – the gasification temperature increases with ER for each fuel (the contour of the temperature in the ER-EBP plane is shown in Fig. 8a), according to two regimes: when char is present among the products (and thus the conditions of carbon boundary are not achieved), the temperature raises slowly; when the char is completely converted in the gasification products, the temperature raises more rapidly. For a fixed ER the temperature decreases with EBP. The temperature denoting the change of regime (TCB) is 1450 K for F1, 1265 K for F4, 1090 K for F8, 945 K for F18; – the production of H2 and CO shows a maximum with ER in correspondence with the carbon boundary conditions: for lower

values of ER there is not enough O2 to gasify the char, for higher values the excess O2 consumes H2 and CO for giving the oxidation products (H2O and CO2); – the production of H2O and CO2 is favored at relatively low values of ER, which correspond to low temperatures as noted in the isothermal case, and relatively high values of ER, which correspond to temperatures greater than TCB, as products of oxidation of H2 and CO. The production of H2O and CO2 increases with EBP for a fixed ER; – the production of CH4 is favored at relatively low values of ER, which correspond to low temperatures. For a fixed ER it increases with EBP. As in the case of char production, the production of CH4 becomes negligible for temperatures greater than TCB;

128

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

Fig. 7. Gasification products (black curves) and temperature (grey curves) for the adiabatic equilibrium of fuel (a) F1, (b) F8, (c) F18, (d) eucalyptus wood, (e) Koflach coal, and (f) PET, as functions of ER (M = 0, UC = 0, mole fractions of products normalized without N2).

– the trend of LHVgas with ER shows three different zones (Fig. 8b). For low values of ER (that is for ER between 0 and 0.1, which correspond to pyrolysis conditions) LHVgas decreases rapidly with ER. At ER = 0, LHVgas passes from 11.5 MJ/m3 for F1, to 7.7 MJ/m3 for F18. For intermediate values of ER (that is between 0.1 to the ER corresponding to TCB), LHVgas decreases slowly with ER, showing a sort of plateau. Mean values of LHVgas in this zone are comprised between 5 and 7 MJ/m3. For high values of ER, LHVgas decreases linearly with ER. In general, for a fixed value of ER in the first and third zones, LHVgas decreases with EBP. In the second zone, LHVgas shows a maximum with EBP: at ER = 0.3 LHVgas is 5.32 MJ/m3 for F1, reaches the maximum of 5.90 MJ/m3 for F8 (corresponding to torrefied PKS, with EBP around 0.03), and then decreases to 5.46 MJ/m3 for F18; – the cold gas efficiency shows a maximum with ER in all cases (Fig. 8c). The maximum CGE corresponds to the value of ER at carbon boundary, that is when the char reaches the complete conversion. The corresponding values of ER (ERCB) and temperature (TCB) are shown in Fig. 9. ERCB decreases with EBP, rapidly for EBP between 0 and 0.03 and then slowly for EBP > 0.03. TCB shows a similar trend. The trend of the maximum value of CGE shows a feeble maximum with EBP. It is 0.757 for F1, 0.800 for F4, 0.834 for F8, 0.835 for F12, 0.831 for F15, 0.825 for F18. The trend of LHVgas, in correspondence of ERCB, shows a more evident maximum with EBP.

3.4. Results for some fuels outside the biomass evolution curve The results shown above were obtained for fuels in the biomass evolution curve of Fig. 1. Qualitative similar results can be obtained for other fuels outside this curve. As a matter of fact, the results can be considered even quantitative if based on the EBP. Six more fuels were studied, outside the curve of fuels studied above, as shown in Fig. 1: two biomasses (Eucalyptus wood and olive pits), two oxygenated plastics (PET and PMMA), and two coals of different ranks (Koflach and Yorkshire coals). Simulations of the adiabatic equilibrium were carried out with these fuels. The composition of syngas obtained for some of them is shown as function of ER in Fig. 7d–f. The trends of the gaseous products for the two biomasses are similar to the biomasses previously studied. The results for the two coals are similar to those of coals of similar rank. The results for the two plastics are similar to those of fuels with similar EBP. As mentioned above, a particular attention was paid to the conditions for the maximum efficiency of the process. As seen with the previous fuels, the maximum CGE is obtained at the carbon boundary, that is once the conversion of char is completed. The results of the corresponding values of ER, adiabatic temperature, LHVgas, and CGECB are superimposed to those obtained for the previous fuels in Fig. 9. As can be observed, the points lie onto the same curves obtained for the previous fuels. So, the results expressed in terms of EBP can be generalized. The following equations,

129

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

Fig. 8. Adiabatic equilibrium results: (a) temperature (in K in the contour curves), (b) LHVgas (in MJ/m3 in the contour curves), (c) CGE in case of UC = 0, and (d) CGE in case of UC = 0.10, as functions of EBP and ER.

ER

ERCB

Eucalyptus

Yorkshire coal

Koflach coal

0

0

1400 1200

0.2 0.1

(a)

PMMA

PET Olive pits PMMA

0.05

Eucalyptus

0.1

1000

LHVgas (MJ/m3)

Olive pits PET

T (K)

Koflach coal

0.3

6.6

1600

6.2

PET

5.8

Koflach coal

5.4

800

(b)

5

Eucalyptus Olive pits

0

CGECB

0.8 0.6

PMMA Eucalyptus

0.4

Olive pits

LHVCB 0.2

Yorkshire coal

TCB 0.15

PMMA

Koflach coal Yorkshire coal

CGE

PET

Yorkshire coal

0.4

0.05

0.1

0.15

0

EBP

EBP

Fig. 9. (a) ER and T, (b) LHVgas and CGE, for the adiabatic equilibrium in the carbon boundary conditions, corresponding to the maximum value of CGE. Comparison of the results for fuels in the evolution curve (continuous lines) and alternative fuels (symbols) as defined in Fig. 1.

obtained by regression of the data of this work, represent the results for a CAHAO system, that is a fuel expressed with its EBP (=H/C ⁄ O/C):

CGECB ðmaximum CGE achievableÞ ¼ ð1 þ 418:1 EBP þ 310:1 EBP2 Þ =ð1:609 þ 487:57 EBP þ 477:64 EBP2 Þ ERCB ðER at max CGEÞ ¼ ð0:2397 þ 0:06773 EBPÞ  EBP0:1045 TCB ðadiabatic T at max CGEÞ ¼ ð0:2734 þ 764:09 EBPÞ  EBP1:1035 ½K LHVgas ðat max CGEÞ ¼ ð1 þ 98:37 EBP þ 625:83 EBP3 Þ =ð0:219 þ 14:04 EBP þ 41:276 EBP2 Þ ½MJ=m3  with EBP in the range ½0:0040  0:1584

ð6Þ

4. Discussion The results and graphs shown in the previous section can be used to predict the equilibrium of isothermal and adiabatic gasification of every fuel with a composition ranging from hard coals to cellulose based on the evolution biomass parameter EBP. As a matter of fact, some practical issues should be accounted for when transferring data from the thermodynamic model to a real system. Some comments to the results obtained above are given in this section, concerning the causes of non-equilibrium (including incomplete carbon conversion), syngas utilization related issues (presence of tar), effect of fuel moisture. In general, the higher the temperature of the gasifier, the more accurate the predictions of the equilibrium model. Patra and Sheth [37] suggested a temperature >1500 K in the reactor to find a good accordance between

130

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

thermodynamic equilibrium and real systems. Also the configuration of the reactor can be favorable to the achievement of the equilibrium conditions. Puig-Arnavat et al. [40] noticed that the equilibrium model applied to entrained flow gasifiers, and to a certain extent also to downdraft gasifiers, can give satisfactory predictions, while for updraft and fluidized bed gasifiers the agreement is less accurate. As seen in the validation section, a good agreement can be found for the prediction of SGP and LHVgas, while some deviation can be expected for the syngas composition. The recognized weak points of the equilibrium model are in the scarce prediction of: d

d

d

the concentration of CH4 and light hydrocarbons in the syngas, which is generally underestimated in equilibrium models; the conversion of char, which is rarely complete in real plants, although the equilibrium model may predict a complete conversion; the presence of undesired products (tar and dust), which is not predicted by the equilibrium approach.

Significant divergences on these points can be found between data from real plants and equilibrium model predictions, and are mainly due to the specific reactor characteristics (e.g. insufficient residence time, inefficient air/fuel mixing, inappropriate particle size of the fuel, thermal gradients, catalytic effects of ash/minerals/metals), which make the practical systems to deviate from the ideal behavior of the chemical equilibrium. In these cases, modifications to the equilibrium approach are recommended to improve the predictability of the model. Some examples can be found in literature works: d

d

d

in the restricted temperature equilibrium approach, an equilibrium temperature lower than the one measured in the gasifier is adopted. The difference of the temperatures can be significant: Li et al. [28] determined a difference of 250 K for coal gasification, Huang and Ramaswamy [23] a difference of 125 K for the gasification of sawdust. In all cases, the optimal temperature for the restricted approach depends strongly on the specific reactor; the adoption of constrains as the introduction of equations related to CH4 kinetics, char conversion and/or tar formation, can give a better prediction of the experimental data on solid fuel gasification (see for instance [23,3,42,24,31]). Also in these cases the model improvement depends strongly on the real system; more complex approaches, based on kinetic models and a multizonal definition of the reactor, can be used to achieve more accurate predictions of the experimental data on specific gasification systems (see for instance, [41,18,7,1,43,25]).

As stated above, although it is sometimes not accurate, only the equilibrium model has a general significance, independent of the reactor configuration. The results obtained in equilibrium conditions should be considered as the attainable limits for the fuel and operating parameters adopted in real systems. The systems can be indeed improved by limiting or removing the causes of non-equilibrium. 4.1. Effect of the unconverted carbon The incomplete conversion of the char is a relevant cause of divergence between equilibrium and real results. Extremely variable values of unconverted carbon can be found at the reactor exit, depending on reactor configuration, operating conditions and fuel characteristics, generally on the order of some percentage points, up to less than 1% for entrained beds (see for instance [7]). As seen

in the results section, the conditions for a complete conversion of the char should be adopted to get the maximum process efficiency. Clearly, the presence of carbon in the solid product of the gasification plant represents an intrinsic inefficiency and a handling trouble for the disposal of the residue. Therefore, every measure, compatible with the system, should be pursued to increase the char conversion, e.g. high temperature, long residence time, good air/fuel mixing, optimal fuel particle size distribution. To quantify the effect of the incomplete char conversion on the gasification performance, some simulations were carried out by imposing a fixed Unconverted Carbon UC = 0.1. This means that 10% of the carbon in the feeding fuel is separated (consisting in 8.7 kg/h for F1 and 4.4 kg/h for F18, respectively, based on a feeding rate of 100 kg/h) and considered inert in the equilibrium model. It is noted that the energy associated to the carbon in the gasifier residue is assumed to be lost and thus not accounted for in the calculation of the efficiency. The case UC = 0.1 was studied in adiabatic simulations and compared to the reference case with UC = 0 (Figs. 8 and 10). Focusing on the conditions of maximum efficiency attainable CGECB, the assumption of UC = 0.1 has a feeble influence on the adiabatic temperature TCB and LHVgas. On the other hand, the values of ERCB decreases significantly, as only 90% of the fuel carbon is gasified: ERCB passes from 0.425 to 0.379 for F1, and from 0.301 to 0.260 for F18 when UC passes from 0 to 0.1. SGP decreases significantly when UC passes from 0 to 0.1 and this causes a decrease in the maximum value of the CGE. It passes from 0.757 to 0.694 for F1, and from 0.825 to 0.750 for F18 when UC passes from 0 to 0.1. On average (see Fig. 8c and d), a reduction of 5–7 percentage points in the maximum process efficiency can be observed for all fuels, corresponding to lower values of ER. 4.2. Effect of the primary measures to reduce the tar The presence of CH4 in the syngas of real plants can be on the order of 0.5–5% (mole basis) and is likely to be the result of an incomplete conversion of pyrolysis products, as well as that of light hydrocarbons and tar [13,19,10]. As the gasification is operated at a temperature above say 1000 K, the concentration of CH4 predicted by the equilibrium model in these conditions is less than 0.5% (see Fig. 5b). This difference might represent a leak in the approach if one were interested in the accurate composition of the syngas, e.g. for processes aiming at further syngas conversion, such as methanation or hydrogen production. In these cases, a system specific model should be applied to quantify the non-equilibrium causes and evaluate the conversion of pyrolysis products. However, if one were interested in the combustion of the syngas, only SGP and LHVgas should be considered. Similar conclusions can be deduced from the results of Li et al. [28], Melgar et al. [30], Azzone et al. [3], Mendiburu et al. [31], Biagini et al. [10]. In these cases, the results of the process efficiency obtained with the equilibrium model can be transferred to the real systems. The heating value is commonly used to evaluate the quality of the syngas. A minimum value around 4.5–5 MJ/m3 should be assured for an efficient conversion in internal combustion engines. These values can be obtained for every fuel, depending on the temperature (>850–1000 K) and ER (<0.4), in either isothermal or adiabatic condition. Although the LHVgas can be maximized by operating at high fuel-to-air ratios (that is low values of ER, as seen in the results of the previous section and similar conclusions found by Melgar et al. [30]), these conditions are difficult to be reached and maintained. As a matter of fact, also the concentration of tar and dust is a crucial index of syngas quality, and extremely low values are desired for the utilization of the syngas. Hasler and Nussbaumer [21] refer to extremely variable values for tar and dust at the exit of common gasifiers, ranging from 100 mg/m3 to 100 g/m3, depending on the reactor configuration. Values less than

131

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

0.2

1200

TCB

0.1

(a)

0

0

0.05

0.1

1000

0.15

800

EBP

LHVgas (MJ/m3)

1400

T (K)

ER

0.3

0.8

6.6

1600

ERCB

CGECB 0.6

6.2

LHVCB

5.8

0.4 0.2

5.4

(b)

5

CGE

0.4

0

0.05

0.1

0.15

0

EBP

Fig. 10. (a) ER and T, (b) LHVgas and CGE, for the adiabatic equilibrium in the carbon boundary conditions, corresponding to the maximum value of CGE. Comparison of the results for M = 0 and UC = 0 (continuous curves), M = 0.10 and UC = 0 (dashed curves), M = 0 and UC = 0.10 (dotted curves).

50 mg/m3 are required for combustion in an engine, less than 30 mg/m3 for combustion in a turbine, and less than 0.1 mg/m3 for conversion in catalytic reactors and fuel cells [21,50]. The dust content in the syngas is strongly dependent on reactor configuration and plant operations, but a model can hardly predict its value. More consistent effort is devoted to the tar content, as its removal is of foremost concern. The tar is a liquid product of the pyrolysis and consists of numerous organic compounds with high molecular weight [32]. They are in the vapor phase at the high temperature of gasification, and condense in the downstream units (400–750 K) giving extremely viscous deposits, which can be hardly removed, and compromising the durable utilization of the plant (tubes, valves, heat exchangers) and syngas conversion units (combustion engine, turbine, catalytic converter). It is desirable to promote the conditions for limiting the formation of tar in the gasification reactor with primary measures and reducing the more onerous secondary measures, e.g. catalytic crackers, filters, scrubbers, plasma [50]. Every representative tar compound (e.g., toluene, phenol, naphthalene) can be hardly predicted by the equilibrium model under the common conditions of gasification. As a matter of fact, due to the relatively low amount of tar produced in the gasification, this limitation does not invalidate the equilibrium predictions on syngas composition and process efficiency. The material and energy balances are not sensitive to such negligible amounts of products. By the way, the same reasoning can be applied also to the pollutant formation, such as HCN, NH3, H2S, COS, HCl, which undoubtedly should be considered when evaluating the entire process, but generally have a negligible importance on the global balances. Nevertheless, the tar related issues impose operating choices, which have a strong influence on the process. The primary measure to limit the tar formation is the high temperature of reaction. The tar can be thermally cracked to light gases when exposed for a sufficient time (less than 1 s) to a temperature above 1200– 1300 K (see for instance [52]). For auto-thermal systems, this means that the gasification is operated at a higher ER than the one required for an optimal performance. This does not represent a loss of efficiency for fuels with high carbon boundary temperatures, such as hard coals, while it does for fuels with low TCB, such as raw and torrefied biomasses (see Fig. 9a). In these cases, the fuel is over-oxidized to achieve a high temperature and operate above its carbon boundary temperature, where the efficiency is significantly reduced (similar conclusions in [39,31]). The results of the adiabatic simulations were rearranged at fixed temperatures in the range 1000–1300 K to compare LHVgas and CGE of different fuels (Fig. 11). As can be seen in Fig. 11a for the case M = 0 and UC = 0, an average reduction of 0.35 MJ/m3 is predicted for LHVgas when the gasification temperature is increased by 100 K (this is operated by increasing ER). The gasification efficiency increases for fuels with small EBP, and decreases for fuels with large EBP.

At 1200 K, the maximum CGE = 0.81 is for EBP = 0.02 (corresponding to the brown coal F6): for lower values of EBP (hard coals) some char still remains unconverted and thus the efficiency strongly decreases; for greater values of EBP (torrefied and raw biomasses) the fuel is over-oxidized and thus the efficiency decreases linearly. The splitting value of EBP depends on the gasification temperature: 0.08 (1000 K), 0.03 (1100 K), 0.02 (1200 K), 0.01 (1300 K). When comparing the adiabatic cases of UC = 0 and UC = 0.1 at fixed temperatures, no significant variations in LHVgas was noted (graph not shown), while a strong decrease in the efficiency can be observed in Fig. 11b, especially for fuels at high EBP. As a matter of fact, for fuels and conditions under which the unconverted carbon is greater than 0.1, no difference in the CGE can be observed when the constrain on the unconverted carbon is applied. In the other cases, an average reduction of 7 percentage points are registered when the unconverted carbon passes from 0 to 0.1. 4.3. Effect of the fuel moisture The initial moisture of the fuel has a strong influence on the process parameters. The case of M = 0.1 was studied in adiabatic simulations and compared to the reference case with M = 0 (Fig. 10), by maintaining the same feeding rate of the dry fuel. The presence of water in the feed significantly decreases the temperature achieved in the equilibrium reactor. The decrease on the carbon boundary temperature is more significant for fuels with low values of EBP, TCB passes from 1452 K to 1240 K for F1, and from 944 K to 913 K for F18 when the moisture content passes from 0 to 0.1. The presence of water in the feed acts also as an additional gasifying agent and thus gives a greater conversion of char, especially for fuels at low values of EBP. The two contributions are opposite but quantitatively similar, so only a slight difference can be observed in the maximum values of CGE: it increases for F1 (from 0.757 to 0.801) and decreases for F18 (from 0.825 to 0.806) when the moisture content passes from 0 to 0.1. The value of ERCB, corresponding to the maximum value of CGE, decreases only for fuels with EBP < 0.03 when moisture is present. The LHVgas decreases only for fuels with EBP > 0.017: it is 5.06 (5.49) MJ/m3 for F1, has a maximum of 5.87 (5.74) MJ/m3 for EBP = 0.03 (0.02), is 5.46 (5.09) for F18, when M = 0 (M = 0.1). Fig. 11a and c compare the cases of M = 0 and M = 0.1, at fixed values of the temperature in adiabatic simulations. Significant variations in LHVgas can be noted. On average, at a fixed temperature, LHVgas is reduced by 0.5 MJ/m3 when the moisture content passes from 0 to 0.1. Although this can be a negative outcome on the process efficiency, one should consider that the presence of water in the system, under the same conditions of temperature and ER, can promote the gasification of char, if present. This eventuality increases the syngas productivity and consequently the efficiency increases for

E. Biagini / Energy Conversion and Management 128 (2016) 120–133

(a)

7

0.9

6

0.8

CGE

LHVgas (MJ/m3)

132

5

4

3

0.7

0.6

0

0.05

EBP

0.1

0.15

(b)

0.5

0

0.05

0.1

0.15

EBP 0.9

CGE

0.8

0.7

0.6

(c) 0.5

0

0.05

EBP

0.1

0.15

Fig. 11. LHVgas (a) and CGE (b, c) for the adiabatic equilibrium at 1000 K (continuous curves), 1100 K (long dashed curves), 1200 K (short dashed curves), and 1300 K (dotted curves). Comparison of the results for M = 0 and UC = 0 (black curves in a, b, c), M = 0 and UC = 0.10 (light grey curves in b), M = 0.10 and UC = 0 (dark grey curves in a, c).

fuels having low values of EBP. Vice versa, for fuels at high EBP and thus having no char to be gasified, the presence of water gives a significant decrease in the CGE. In these cases, an average reduction of 3 percentage points are registered when the fuel moisture content passes from 0 to 0.1.

the ideal conditions of equilibrium. A pre-treatment (e.g. torrefaction) moves the EBP of a green biomass to lower values, so that a higher gasification temperature can be achieved to attain the maximum heating value of the syngas (around 5.8 MJ/m3) and a limited reduction in the process efficiency.

5. Conclusion

References

The air-blown gasification of solid fuels on a biomass-to-coal evolution curve was studied with a non-stoichiometric equilibrium model based on the minimization of the Gibbs free energy under both the isothermal and adiabatic conditions. The Equivalent Ratio was varied between 0 to 0.6, the temperature in the range 500– 2000 K, the fuel parameter EBP (=H/C ⁄ O/C) in the range 0.004 (hard coal) – 0.158 (cellulose). The introduction of this parameter allowed the gasification of fuels of different origin to be quantitatively compared, including even plastics: the gasification of PET and torrefied biomasses (EBP around 0.03) gave similar results, and so also PMMA and peat (EBP = 0.07). The maximum efficiency of the process (cold gas efficiency as high as 0.8) was achieved for every fuel at the carbon boundary conditions, that is at the temperature or ER for the complete conversion of the char. For fuels with small EBP (coals), the high values of ER (around 0.4) and T (1200–1400 K), which are known to be sustainable for the low tar formation, were proven to assure a high efficiency of gasification, which can be increased by the presence of moisture or added steam, while the unconverted carbon can be minimized by the adoption of an entrained bed reactor. The potential high efficiency of green biomasses was proven to be strongly reduced by the practical conditions of high gasification temperature (at least 1200 K to limit the formation of tar), initial moisture content and unconverted carbon, due to the discrepancy between the real conditions in small fixed or fluidized bed reactors and

[1] Abdelouahed L, Authier O, Mauviel G, Corriou JP, Verdier G, Dufour A. Detailed modeling of biomass gasification in dual fluidized bed reactors under Aspen Plus. Energy Fuels 2012;26:3840–55. [2] Arena U. Process and technological aspects of municipal solid waste gasification. Waste Manage 2012;32:625–39. [3] Azzone E, Morini M, Pinelli M. Development of an equilibrium model for the simulation of thermochemical gasification and application to agricultural residues. Renewable Energy 2012;46:248–54. [4] Bach QV, Skreiberg O. Upgrading biomass fuels via wet torrefaction: a review and comparison with dry torrefaction. Renew Sustain Energy Rev 2016;54:665–77. [5] Baratieri M, Baggio P, Fiori L, Grigiante M. Biomass as an energy source: thermodynamic constraints on the performance of the conversion process. Bioresour Technol 2008;99:7063–73. [6] Basu P. Biomass gasification, pyrolysis and torrefaction. 2nd ed. London: Academic Press, Elsevier; 2013. [7] Biagini E, Bardi A, Pannocchia G, Tognotti L. Development of an entrained flow gasifier model for process optimization study. Ind Eng Chem Res 2009;48:9028–33. [8] Biagini E, Barontini F, Tognotti L. Gasification of agricultural residues in a demonstrative plant: corn cobs. Bioresour Technol 2014;173:110–6. [9] Biagini E, Barontini F, Tognotti L. Gasification of agricultural residues in a demonstrative plant: vine pruning and rice husks. Bioresour Technol 2015;194:36–42. [10] Biagini E, Barontini F, Tognotti L. Development of a bi-equilibrium model for biomass gasification in a downdraft bed reactor. Bioresour Technol 2016;201:156–65. [11] Biagini E, Fantozzi C, Tognotti L. Characterization of devolatilization of secondary fuels in different conditions. Combust Sci Technol 2004;176:685–703. [12] Biagini E, Tognotti L. A generalized procedure for the devolatilization of biomass fuels based on the chemical components. Energy Fuels 2014;28:614–23.

E. Biagini / Energy Conversion and Management 128 (2016) 120–133 [13] Boroson ML, Howard JB, Longwell JP, Peters AW. Products yields and kinetics from the vapor phase cracking of wood pyrolysis tars. AIChE J 1989;35:120–8. [14] Bridgeman TG, Jones JM, Shield I, Williams PT. Torrefaction of reed canary grass, wheat straw and willow to enhance solid fuel qualities and combustion properties. Fuel 2008;87:844–56. [15] Bridgwater AV. Renewable fuels and chemicals by thermal processing of biomass. Chem Eng J 2003;91:87–102. [16] Cairns EJ, Tevebaugh AD. CHO gas phase composition in equilibrium with carbon, and carbon deposition boundary at one atmosphere. J Chem Eng Data 1964;9:453–62. [17] Chen WH, Peng J, Bi XT. A state-of-the-art review of biomass torrefaction, densification and applications. Renew Sustain Energy Rev 2015;44:847–66. [18] Corella J, Sanz A. Modeling circulating fluidized bed biomass gasifiers. A pseudo-rigorous model for stationary state. Fuel Process Technol 2005;86:1021–53. [19] Fagbemi L, Kezhami L, Capart R. Pyrolysis products from different biomasses: application to the thermal cracking of tar. Appl Energy 2001;69:293–306. [20] Gomez-Barea A, Leckner B. Modeling of biomass gasification in fluidized bed. Prog Energy Combust Sci 2010;36:444–509. [21] Hasler P, Nussbaumer T. Gas cleaning for IC engine applications from fixed bed biomass gasification. Biomass Bioenergy 1999;16:385–95. [22] Higman C, Van der Burgt M. Gasification. Elsevier; 2008. [23] Huang HJ, Ramaswamy S. Modeling biomass gasification using thermodynamic equilibrium approach. Appl Biochem Biotechnol 2009;154:193–204. [24] Kangas P, Hannula I, Koukkari P, Hupa M. Modelling super-equilibrium in biomass gasification with the constrained Gibbs energy method. Fuel 2014;129:86–94. [25] Kong X, Zhong W, Du W, Qian F. Compartment modeling of coal gasification in an entrained flow gasifier: a study on the influence of operating conditions. Energy Convers Manage 2014;82:202–11. [26] Kwapinska M, Xue G, Horvat A, Rabou LP, Dooley S, Kwapinski W, et al. Fluidized bed gasification of torrefied and raw grassy biomass (Miscanthus  giganteous). Energy Fuels 2015;29:7290–300. [27] Li J, Biagini E, Yang W, Tognotti L, Blasiak W. Flame characteristics of pulverized torrefied biomass combusted with high temperature air. Combust Flame 2013;160:2585–94. [28] Li X, Grace JR, Watkinson AP, Lim CJ, Ergudenler A. Equilibrium modeling of gasification: a free energy minimization approach and its application to a circulating fluidized bed coal gasifier. Fuel 2001;80:195–207. [29] McKendry P. Energy production from biomass (part 3): gasification technologies. Bioresour Technol 2002;83:55–63. [30] Melgar A, Perez JF, Laget H, Horillo A. Thermochemical equilibrium modelling of a gasifying process. Energy Convers Manage 2007;48:59–67. [31] Mendiburu AZ, Carvalho JAJ, Zanzi R, Coronado CR, Silveira AZ. Thermochemical equilibrium modeling of a biomass downdraft gasifier: constrained and unconstrained non stoichiometric models. Energy 2014;71:624–37.

133

[32] Milne TA, Abatzoglou N, Evans RJ. Biomass gasifier tars: their nature, formation and conversion. Golden, Co: National Renewable Energy Laboratory NREL/TP-570-25357; 1998. [33] Milne TA, Brennan AH, Glenn BH. Sourcebook of methods of analysis for biomass and biomass conversion processes. Elsevier Applied Science; 1990. ISBN 1-85166-527-7. [34] Minchener AJ. Coal gasification for advanced power generation. Fuel 2005;84:2222–35. [35] Oliveira I, Blohse D, Ramke HG. Hydrothermal carbonization of agricultural residues. Bioresour Technol 2013;142:138–46. [36] Pandey A, Negi S, Binod P, Larroche C. Pretreatment of biomass – processes and technologies. Amsterdam, Netherlands: Elsevier; 2015. [37] Patra TK, Sheth PN. Biomass gasification models for downdraft gasifier: a stateof-the-art review. Renew Sustain Energy Rev 2015;50:583–93. [38] Phyllis2 Database for biomass and waste, Energy research Center of the Netherlands ECN, https://ww.ecn.nl/phyllis2 last accessed 15 June 2016. [39] Prins MJ, Ptasinski KJ, Janssen FJJG. From coal to biomass gasification: comparison of thermodynamic efficiency. Energy 2007;32:1248–59. [40] Puig-Arnavat M, Bruno JC, Coronas A. Review and analysis of biomass gasification models. Renew Sustain Energy Rev 2010;14:2841–51. [41] Radulovic PT, Ghani MU, Smoot LD. An improved model for fixed bed coal combustion and gasification. Fuel 1995;74:582–94. [42] Rofouie P, Moshkelani M, Perrier M, Paris J. A modified thermodynamic equilibrium model for woody biomass gasification. Can J Chem Eng 2014;92:593–602. [43] Roy PC, Datta A, Chakraborty N. An assessment of different biomass feedstocks in a downdraft gasifier for engine application. Fuel 2013;106:864–8. [44] Shabbar S, Janajreh I. Thermodynamic equilibrium analysis of coal gasification using Gibbs energy minimization method. Energy Convers Manage 2013;65:755–63. [45] Solid Fuel Database. SFDB, International Flame Research Foundation, IFRF at http://sfdb.ifrf.net [accessed 15.06.16]. [46] Tapasvi D, Kempegowda RS, Tran KQ, Skreiberg O, Gronli M. A simulation study on the torrefied biomass gasification. Energy Convers Manage 2015;90:446–57. [47] Tay DHS, Kheireddine H, Ng DKS, El-Halwagi MM, Tan RR. Conceptual synthesis of gasification-based biorefineries using thermodynamic equilibrium optimization models. Ind Eng Chem Res 2011;50:10681–95. [48] Tekin K, Karagoz S, Bektas S. A review of hydrothermal biomass processing. Renew Sustain Energy Rev 2014;40:673–87. [50] Woolcock PJ, Brown RC. A review of cleaning technologies for biomass-derived syngas. Biomass Bioenergy 2013;52:54–84. [51] Worley M, Yale J. Biomass gasification technology assessment, Consolidated Report of the U.S. National Renewable Energy Laboratory NREL/SR-510057085, November; 2012. [52] Yuehong Z, Hao W, Zhihong X. Conceptual design and simulation study of a cogasification technology. Energy Convers Manage 2006;47:1416–28.