Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization Antonio C.D. Freitas, Reginaldo Guirardello* School of Chemical Engineering, University of Campinas (UNICAMP), Av. Albert Einstein 500, 13083-852 Campinas, SP, Brazil

article info

abstract

Article history:

This paper focuses on the comparison of different glycerol reforming technologies aimed

Received 12 December 2013

to hydrogen and syngas production. The reactions of steam reforming, partial oxidation,

Received in revised form

autothermal reforming, dry reforming and supercritical water gasification were analyzed.

26 February 2014

For this, the Gibbs energy minimization approach was used in combination with the virial

Accepted 18 March 2014

equation of state. The validation of the model was made between the simulations of the

Available online xxx

proposed model and both, simulated and experimental data obtained in the literature. The effects of modifications in the operational temperature, operational pressure and reactants

Keywords:

composition were analyzed with regard to composition of the products. The effect of coke

Reforming processes

formation was discussed too. Generally, higher temperatures and lower pressures resulted

Gibbs energy minimization

in higher hydrogen and syngas production. All reforming technologies demonstrated to be

Hydrogen

feasible for use in hydrogen or synthesis gas production in respect of the products

Syngas

composition. The proposed model showed good predictive ability and low computational time (close to 1 s) to perform the calculation of the combined chemical and phase equilibrium for all systems analyzed. Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction In the past decades, biomass has received a great deal of attention as a new source for clean energy because it is renewable, alternative and producing less carbon dioxide than fossil fuels [1]. In the last decade, biodiesel has been becoming a major substitute for fossil diesel, as its environmental benign characteristics can reduce carbon dioxide emission pollution and mitigate climate change problems [2,3].

Glycerol is obtained as a by-product in biodiesel production using vegetable oils by a base-catalyzed transesterification reaction. Glycerol of high purity is an important industrial feedstock for applications in food, cosmetics, pharmaceutical and other industries [4]; however, it is costly to refine crude glycerol, especially for medium and small-sized plants. In any case, since glycerol production and utilization have a notable impact on the economics and sustainability of biodiesel production, the development of novel processes for glycerol valorization is essential. Among all possible routes, glycerol conversion into valuable gases like as hydrogen and

* Corresponding author. Tel.: þ55 19 3521 3955; fax: þ55 19 3521 3910. E-mail addresses: [email protected], [email protected] (R. Guirardello). http://dx.doi.org/10.1016/j.ijhydene.2014.03.130 0360-3199/Copyright ª 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

syngas as an energy carrier using different reforming processes are one of the most attractive ways to aggregate value in the glycerol [4]. The reforming processes used to produce hydrogen and syngas from glycerol include some methods, and in the literature, it is possible to find some articles about then. The articles principally discussing the catalytic behavior of the systems, the main related methods for glycerol reforming are steam reforming (SR) [5e11], partial oxidation (or oxidative reforming) (PO) [12e14], autothermal reforming (ATR) [12,14] and supercritical water gasification (SCWG) [15e19]. Other methods like as dry reforming (or CO2 reforming) (DR) [20,21], dry autothermal (a combination between dry, oxidative and steam reforming) [22] and pyrolysis [23], also, were evaluated as possible routes for glycerol conversion. Normally catalysts were used to promote and accelerate these reforming processes. Ni, Co and noble metals like as Pt, Pd and Rh based catalysts were related to promote the reforming process. Among then, the Ni based catalysts are the most commonly used. The main reactions that can be occurred in the reforming processes of glycerol are presented in Table 1. In the following sections, the steam, oxidative, autothermal, dry and supercritical water reforming reactions will be discussed, in terms of reactions conditions and characteristics of the reactions for hydrogen and syngas production.

Steam reforming of glycerol (SR) SR is the most commonly used technology for producing hydrogen in the chemical industry especially from methane [24,25]. The SR of natural gas produces approximately 90% of

Table 1 e Main reactions in glycerol reforming processes. Reaction Glycerol oxidation reactions C3H8O3 þ 0.5O2 / 2CO þ CO2 þ 4H2 C3H8O3 þ 3.5O2 / 3CO2 þ 4H2O C3H8O3 þ 1.5O2 / 3CO2 þ 4H2 Steam reforming of glycerol/SCWG C3H8O3 þ 3H2O 4 3CO2 þ 7H2 Glycerol decomposition C3H8O3 4 4H2 þ 3CO Water gas shift reaction CO þ H2O 4 CO2 þ H2 CO disproportionation reaction 2CO 4 CO2 þ C(s) Methanation of CO CO þ 3H2 4 CH4 þ H2O Methanation of CO2 CO2 þ 4H2 4 CH4 þ 2H2O Methane steam reforming CH4 þ H2O 4 3H2 þ CO Methane dry reforming CO2 þ CH4 4 2H2 þ 2CO Methane decomposition CH4 4 2H2 þ C(s) Methane oxidative reforming CH4 þ 0.5O2 4 2H2 þ CO Carbon gasification C(s) þ 2H2O 4 2H2 þ CO2

DHR298 K ðkJ$mol1 Þ 32 1565 598

(1)

128

(2)

250

(3)

41

(4)

172

(5) (6)

206 165 207

(7)

247

(8)

76

(9)

38

(10)

90

(11)

the hydrogen available on the market. In this process the substrate is reacted with steam in the presence of a catalyst to produce hydrogen or syngas. The SR is an endothermic process, i.e. energy is needed for this process. This energy is used to break up the CeC and HeC bonds in the chain of the reactant. The high-energy consumption in the steam reforming process is the principal drawback of this process. The principal advantages of the SR are that it can convert the substrate in hydrogen at low pressures and can produce high concentrations of hydrogen with high substrate conversions when compared with other reforming process [26]. Many papers have been performed on hydrogen production from hydrocarbons and biomass materials using SR processes, like as Adhikari et al. [7,24,27,28], Slinn et al. [29], Oliveira et al. [30] and Rass-Hansen et al. [26]. Different authors evaluated the SR of glycerol in the last years [8,10,31e36], most of these works talk about the catalytic behavior and about the effects of different catalysts in SR systems. The conditions related to be the more favorable to reforming reactions are 500e900  C, 1 atm and 6-9/1 of water/ glycerol molar ratio.

Partial oxidation of glycerol (PO) In the PO the substrate is reacted with oxygen or air. The reaction results in great amounts of energy and high temperatures are observed, temperatures close to 1000  C are reported in the literature [37]. As a result, no external heat is required to drive the reactions in a PO process once the reaction is initiated. This reaction may be conducted with or without catalysts and the production of syngas is favored because the carbon monoxide (CO) selectivity is higher in the PO process when compared with the SR process [37]. Although the PO process presents low energy requirements than SR it commonly operates at very high temperature in non-catalytic process (above 1200  C). In the last decade, some studies analyzed the catalytic PO promoted by different metallic catalysts. Ni, Pt and Rh based catalysts are the most commonly used in the PO reaction. Normally was observed that in the catalytic PO high conversions are achieved and equilibrium compositions are observed in temperatures above 600  C [38,39].

Autothermal reforming of glycerol (ATR) The ATR of glycerol is a combination of SR and PO. ATR, as well as PO, does not require an additional heat supply once the reaction is initiated. Thus, ATR can be run up automatically by properly adjusting the reactive conditions [37e40]. Additionally the reactor temperature can be optimized and the catalyst deactivation can be avoided by manipulation of the reactants compositions in the ATR [41]. The main advantages of the ATR are that lower energy is supplied than in steam reforming and the composition of the products out-stream can be easily manipulated by modification in the inlet-stream. Different studies analyzed the ATR of glycerol experimentally Dauenhauer et al. [42] and Douette et al. [43]

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

3

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

evaluated the behavior of the ATR of glycerol in the presence of Pt and Ni based catalysts respectively. Conversions close to equilibrium ones are reported by the authors.

Dry reforming of glycerol (DR) The DR process for syngas production has attracted much attention from both industrial and environmental aspect. Glycerol reforming with CO2 could be an attractive process mainly because of the biodiesel-derived glycerol is considered renewable and the CO2 will be converted in syngas or high value added gases. However the coke (C(s)) formation in the DR is more severe than in other reforming process, due to the lower H/C ratio of this reaction [44]. One of the obstacles encountered in the application of this technology is the rapid deactivation of the catalyst, which is mainly due to coke accumulation and sintering of both the support and the active metal particles [20]. No papers evaluating the experimental behavior of the DR of glycerol were found in the literature. Wang et al. [20] performed the thermodynamic analysis of glycerol dry reforming using the Lagrange’s undetermined method in Matlab software. The authors observed that higher pressures have a negative effect on both hydrogen and syngas yield, the hydrogen yield increases with increasing temperature and large amounts of CO2 in the feed results in poor hydrogen generation.

Supercritical water gasification of glycerol (SCWG) The SCWG reaction occurs in an aqueous fluid phase under conditions that exceed the critical point of water (TC ¼ 647.15 K, PC ¼ 21.9 MPa). Supercritical water has unique features with respect to its density, dielectric constant, ion product, viscosity, diffusivity, electric conductance and solvent ability [45]. SCWG of biomass seems to be a promising technology for the production of hydrogen and other useful gases from wet biomass [46]. The main advantage of the SCWG are that the process does not require the expensive drying process, necessary for others thermochemical process for transformation of biomass [47e49]. However high-energy costs are associated with the compression steps. Other important factor in SCWG is that high temperature and pressures of the process ensure that intrinsic reaction rates are high, and near equilibrium conversions being reached [50]. It is generally believed that the ionic product and dielectric constant of water below its critical point are the major factors controlling hydrolysis reactions of organic materials [51]. May et al. [52] and Byrd et al. [15] evaluated the SCWG of glycerol in a catalytic system. The authors observed that glycerol was completely gasified to hydrogen, carbon dioxide and methane with only small amounts of carbon monoxide and the hydrogen yield were found to increase directly with temperature. In this paper, we focused on the comparison of several glycerol reforming technologies. The SR, PO, ATR, DR and SCWG of glycerol were analyzed. For this, we used the Gibbs

energy minimization approach in combination with the virial EoS. Comparisons between experimental and simulated data obtained from literature for these systems were performed. In addition, the effects of the process variables (temperature, pressure and inlet composition) were evaluated with respect to the product compositions. The effects of the presence of catalysts in the system and formation of coke were discussed too.

Methodology Model formulation Gibbs energy minimization model The thermodynamic equilibrium condition for reactive multicomponent closed system, at constant pressure and temperature, with given initial composition, can be obtained by the minimization of Gibbs energy (G) of the system, with respect of the number of moles of each component in each phase, given by: minG ¼

NC X

g g

ni mi þ

i¼1

NC X

nli mli þ

i¼1

NC X

nsi msi

(12)

i¼1

While satisfying the restrictions of non-negative number of moles of each component in each phase: g

ni ; nli ; nsi  0

(13)

In addition, the restriction of mole balances, given by atom balance for reactive systems (non-stoichiometric formulation): NC P i¼1

NC  g  P ami $ ni þ nli þ nsi ¼ ami $n0i

m ¼ 1; .; NE

(14)

i¼1

Smith and Missen [53] demonstrated that the stoichiometric formulation is equivalent to the non-stoichiometric one, if all independent reactions are considered. g The values of mi can be calculated from the formation values given at some reference conditions, using the following thermodynamic conditions: vHi  vT

P

¼ Cpi

i ¼ 1; .; NC

v  mi  Hi ¼ vT RT P R$T2

(15)

i ¼ 1; .; NC

(16)

In this work, which considers only a gas phase and a possible solid phase as carbon for coke formation, the Gibbs energy can be expressed as: G¼

NC X i¼1

NC   X g ni m0i þ RT ln P þ ln yi þ ln 4i þ nsCðsÞ m0CðsÞ

(17)

i¼1

Since, some of the systems analyzed by the present work was at high pressure, the virial equations of state (EoS) truncated at second virial coefficient, were utilized to determine the fugacity coefficient. The second virial coefficient was calculated by the correlation of Pitzer and Curl [54], which was modified by Tsonopoulos [55]. The following relation determined the fugacity coefficient:

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

2 b i ¼ 42 ln f

m X j

3 yj Bij  B5

P RT

(18)

The Gibbs energy minimization method is widely studied in the literature, being applied in a wide range of systems [12,13,20,24,28,56e64]. Some of then related to glycerol reforming processes [13,20,24,28,56], but none of them presents a complete thermodynamic study about these reactions.

Model implementation During the process of optimization, utilizing the Gibbs energy g minimization method the number of moles of the gaseous (ni ) and solid (nsi ) phase are considered decision variables, while T, P and the chemical potential of the pure component in the reference state (m0i ) are considered parameters. The software GAMS 23.2.1, (General Algebraic Modeling System) with the CONOPT solver was used in the resolution of the combined chemical and phase equilibrium problem. A description of GAMS software can be found in Brooke et al. [65]. The solid phase formed was considered as solid carbon (pure component). These methodologies and considerations were applied in previous works in literature with good predictive results for similar systems with methane and biomass [58e62]. A total of 15 output compounds were selected as representative of the main compounds, which can be found in the output stream of these reactive systems. The thermodynamic data used in the calculations were obtained from literature [66e68]. In almost all the thermodynamic analyses of glycerol reforming processes, the systems were represented considering primary species, such as hydrogen, methane, carbon monoxide, carbon dioxide, water and glycerol, and in some papers the solid carbon (C(s)) was considered too. The consideration of the formation of solid carbon is very important because the C(s) can poison the catalyst in reforming reactions [44]. However, some experimental works in the literature related the presences of secondary components in these reactions, some of the secondary compounds are methanol, ethanol, acetaldehyde, acetone and ethane [23,69]. In this work, a complete thermodynamic analysis of these reactions was performed and all species mentioned above and some others have been take into account. The complete list of output compounds, along with their critical parameters, is reported in Table 2.

Results and discussion Production of hydrogen and syngas has been analyzed in different pressure, temperature and inlet composition conditions in the SR, PO, ATR, DR and SCWG from glycerol. In the following section, the proposed model was validated with simulated and experimental data obtained in the literature for the different glycerol reforming technologies evaluated by the present work.

Table 2 e Chemical compounds considered in the simulations and your thermodynamic properties. Compound

Chemical TC PC VC u ()a a a 3 a formula (K) (bar) (m /kmol)

Water Glycerol Hydrogen Methane Carbon dioxide Carbon monoxide Oxygen Ethane Propane Ethylene Propylene Methanol Ethanol Acetone Acetaldehyde

H2O C3H8O3 H2 CH4 CO2 CO O2 C2H6 C3H8 C2H4 C3H6 CH3OH C2H5OH C3H6O C2H4O

a

647.3 850.0 33.0 191.1 304.2 133.0 154.6 305.4 369.9 283.1 369.9 512.6 514.0 508.1 461.0

220.0 75.0 13.0 45.8 73.9 35.0 50.4 48.2 42.0 50.5 45.4 80.1 61.4 47.0 55.7

0.056 0.260 0.064 0.099 0.094 0.093 0.073 0.148 0.200 0.124 0.182 0.116 0.168 0.209 0.154

0.348 0.513 0.000 0.013 0.420 0.041 0.022 0.105 0.152 0.073 0.143 0.559 0.644 0.304 0.303

Source: Refs. [66e68].

Model validation The validation was performed using simulated and experimental data obtained in literature in comparison with the simulations performed using the proposed virial model. The objective of the validating the model is to verify the predictive ability of the model against experimental data obtained in the literature and verify the ability of the virial equation in represent the non-ideality’s of the studied systems. In order to compare the results calculated using the proposed approach with the values found in literature, the mean relative error (MRE) was used, according to equation (19). MRE ¼

  NPE X NCE xlit  xsim  X 1 j;i   j;i $    xlit  NPE$NCE j j;i i

(19)

where NPE is the number of experimental points, NCE is the is number of components for each experimental point, xsim j;i the value calculated by the present work using the proposed virial model and xlit j;i is the experimental or simulated value obtained in the literature. Fig. 1 shows the validation of the virial proposed model in comparison with the experimental results of Byrd et al. [15] and the simulated results from the work of Castello and Fiori [56]. The experiments of Byrd et al. [15] were performed at the following conditions: temperature of 800  C (1073.15 K), pressure of 241 bar and with feed concentration in the range of 5e40 wt%. The simulations of the present work and that obtained in the work of Castello and Fiori [56] were performed under the same above-mentioned conditions. However, Castello and Fiori [56] used the PengeRobinson EoS in the Gibbs energy minimization model. Analyzing Fig. 1 it is possible to verify that the virial proposed model predictions are in good agreement with the experimental behavior reported by Byrd et al. [15]. The highest deviations were observed in the higher initial feed concentration region (above 25 wt%). It can also be seen that the model predictions for H2 production is better than that

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

5

Fig. 1 e Virial model validation with the experimental data from Byrd et al. [15] and the simulated data from Castello and Fiori [56] for the SCWG of glycerol. Symbols: solid line: virial proposed model predictions; black symbols: experimental data from Byrd et al. [15]; dashed line D white symbols: simulated data from Castello and Fiori [56].

observed for the other compounds (CO2, CH4 and CO). The MRE between the experimental and simulated data was 0.692. Still in Fig. 1, it is also interesting to note that the use of the virial equation instead of PengeRobinson EoS does not result in changes in the predictive ability of the Gibbs energy model, although it leads to a significant mathematical simplification of the Gibbs energy minimization problem. The experimental results for the H2 yield were obtained in the work of Byrd et al. [15] too. The simulations were performed at the same above-mentioned conditions and the comparison is presented in Fig. 2(a). It is interesting to emphasize that the model predictions for the H2 yield were good. The MRE between the experimental and simulated data was 0.104. Analyzing Fig. 2(a) it is possible to verify that increasing the concentration of glycerol in the feed is associated with decrease in the hydrogen concentration and an accompanying increase in the methane concentration in the product. This trend can be explained by considering that less water is present at the higher concentrations of glycerol in the feed. In Fig. 2(b) the effect of modifications in the reaction temperature was analyzed, the simulations were performed at the following conditions: pressure of 241 bar; residence time of 1 s and with 5 wt% of glycerol in the feed stream. The observed MRE was low: 0.44 indicating good ability of the model to predict the effect of reaction temperature. Analyzing Fig. 2(b) it’s possible to verify that the elevation of the reaction temperature resulted in improved prediction capability of the model in comparison with the behavior of the experimental data. It can also be seen that the methane composition is slightly higher at lower temperatures, and this trend was well represented by the model predictions. Analyzing Figs. 1 and 2(a) and (b) it’s possible to see that the experimental gas compositions and H2 yields closely mirror the equilibrium compositions calculated by the proposed virial model, this behavior indicate that the reaction is near its

Fig. 2 e Comparison between experimental data from Byrd et al. [15] and the results obtained using the proposed virial model for (a) H2 yield for the SCWG of glycerol; (b) product gas composition as function of reaction temperature.

thermodynamic equilibrium. For higher glycerol concentration in the feed, was observed that the experimental carbon monoxide concentrations are smaller than predicted by the thermodynamic model, the behavior for H2 and CO was the opposite, and the experimental data was higher than predicted by the thermodynamic model. This behavior can be explained by the water gas shift reaction (Eq. 4, in Table 1), it is possible that this reaction has not reached the equilibrium condition in the experimental conditions. In Fig. 3, the validation of the virial model is performed with the simulated and experimental data obtained in the work of Wang et al. [70] for the steam reforming of glycerol. The experiments of the Wang et al. [70] were performed at 923 K, in the presence of Ni/ZrO2 catalyst and with two different water to glycerol molar ratios in the feed, 6 and 9. Fig. 3(a) presents the results obtained to a WGR of 6 and Fig. 3(b) the results to a WGR of 9. The simulations performed using the proposed virial model and the simulations obtained in the work of Wang et al. [70] were performed in the same aforementioned experimental conditions. The simulations of Wang and coworkers were performed using the ideal gas formulation and the Lagrange’s undetermined multiplier method. Analyzing Fig. 3 is possible to verify that the description of the experimental data is better for the simulations performed

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

Fig. 3 e Comparison of product composition (dry basis) between experimental and simulated data from Wang et al. [70] and the predictions obtained using the virial proposed model for the steam reforming of glycerol with different WGR in the feed (a) WGR-6; (b) WGR-9.

by the present work, using the virial model, than the simulations of Wang et al. [70] that used the consideration of ideality. The product compositions are close to the thermodynamic equilibrium values over the investigated reaction conditions. The MRE between the experimental and simulated data was 0.232 for the WGR of 6 and was 0.310 for the WGR of 9. Douette et al. [43] studied the autothermal reforming of glycerol using a 23 factorial experimental design, in a fixed be reactor of nickel based catalyst. The effects of the oxygen to carbon ratio, steam to carbon ratio and reformer temperature were evaluated over H2 yield, CO yield and CO2 yield. Table 3 presents the experimental conditions of the 12 experiments performed by Douette and coworkers, all experiments were conducted at atmospheric pressure. The experimental results obtained by Douette et al. [43] for the conditions presented in Table 3 and the predictions obtained by the present work using the virial model can be visualized in Fig. 4(a)e(c) for the H2 yield, CO yield and CO2 yield, respectively, as a function of the conditions presented in Table 3.

Table 3 e Experimental conditions in the work of Douette et al. [43] for the autothermal reforming of glycerol. Condition number

1 2 3 4 5 6 7 8 9 10 11 12

Variable values Oxygen to carbon ratio

Steam to carbon ratio

Temperature (K)

0.40 0.70 0.40 0.70 0.40 0.70 0.40 0.70 0.55 0.55 0.30 0.00

2.00 2.00 2.70 2.70 2.00 2.00 2.70 2.70 2.40 2.40 2.30 2.20

1043.15 1043.15 1043.15 1043.15 1123.15 1123.15 1123.15 1123.15 1083.15 1083.15 1080.15 1077.15

Fig. 4 e Comparison between experimental results of Douette et al. [43] and the predictions obtained using the virial model for the autothermal reforming of glycerol for the (a) gas yield of H2; (b) gas yield of CO; and (c) gas yield of CO2.

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

7

literature for this system, thus the comparison with experimental data was not possible for DR of glycerol. With the results was possible to verify that the proposed model has good predictive ability, and can therefore be used to predict the thermodynamic behavior of the different reforming technologies of glycerol evaluated by the present work. In the following sections, the model was used to study different reforming technologies of glycerol for hydrogen and syngas production, evaluating the effect of modification in the operational temperature, pressure and in the reactants composition.

Hydrogen and syngas production Fig. 5 e Comparison between simulated data of Wang et al. [20] and the simulations performed by the present work using the virial proposed model for the dry reforming of glycerol. Symbols: solid line: syngas production; dashed line: hydrogen production; black symbols: simulated data from Wang et al. [20] for syngas production; white symbols: simulated data from Wang et al. [20] for hydrogen production.

The MRE between the experimental and simulated data was 0.292 for H2 yield (Fig. 4(a)), 0.054 for CO yield (Fig. 4(b)) and 0.039 for CO2 yield (Fig. 4(c)). The highest deviations were observed for H2 (Fig. 4(a)), especially in conditions 2, 6, 8, 9, 10 and 12. No direct relationship was observed to relate the points with larger deviations, however, we point out that for many of these cases large deviations were observed as the experimental data is for the composition of CO2, CO and H2. These deviations may be associated with larger deviations in conditions 2, 6, 8, 9, 10 and 12. Wang et al. [20] performed the thermodynamic analysis of the dry reforming of glycerol using the Gibbs energy minimization in the Matlab language. The Lagrange undetermined multiplier method and the ideal gas consideration are used. Fig. 5 presents the comparison between the simulated data obtained in the work of Wang et al. [20] and the simulations performed using the virial proposed model for the DR of glycerol. The simulations performed by the present work are in excellent agreement with the simulations obtained in the work of Wang et al. [20]. The MRE between the data was 0.063 for the number of moles of H2 and 0.058 for the number of moles of syngas. No experimental data was obtained in the

In the following sections, the results obtained using the virial model for different glycerol reforming technologies are presented. For each reaction a detailed thermodynamic evaluation is presented, based on thermodynamic effects of temperature, pressure and reactant composition. The conditions under each one of the glycerol reforming technologies were evaluated are presented in Table 4, and the effect of each one of these variables is evaluated under the production of H2 and syngas.

Steam reforming Fig. 6 shows the results obtained using the Gibbs energy minimization combined with virial model for the effects of temperature (Fig. 6(a)), pressure (Fig. 6(b)) and glycerol/water molar ratio (Fig. 6(c)) in the SR of glycerol. The effect of changes on operating temperature, at atmospheric pressure, on the equilibrium compositions under H2, CO and syngas production are show in Fig. 6(a). The glycerol/ water molar ratio was set at 1/3. The elevation of the operating temperature results in the increased productions of H2 and CO until the temperature of 1173.15 K. Above this temperature the elevation of operating temperature did not result in increases in H2, CO, and consequently syngas, production. Similar behavior was observed in the simulations performed by Adhikari et al. [24]. This behavior can be explained by the occurrence of reverse water gas shift reaction (Eq. (4) in Table 1) in conditions of high temperature. Fig. 6(b) shows the effect of pressure under H2, CO and syngas production. The temperature and the glycerol/water molar ratio were fixed in 973.15 K and 1/3, respectively. The elevation of the pressure results in decreased production of the interest products, this effect was observed in all range of pressure analyzed (1e20 bar). For example, the increase in the pressure of 1e20 bar results in a decrease of 58%

Table 4 e Conditions evaluated in the different glycerol reforming processes evaluated. Reforming technology Steam reforming (SR) Partial oxidation (PO) Autothermal reforming (ATR) Dry reforming (DR) Supercritical water gasification (SCWG)

Reaction temperature range (K)

Reaction pressure range (bar)

Feed composition range

773.15e1273.15 773.15e1273.15 773.15e1273.15 773.15e1273.15 773.15e1273.15

1e20 bar 1e20 bar 1e20 bar 1e20 bar 250e350 bar

1/0.5e3.0 glycerol/H2O molar ratio 1.0/0.25e1.5 glycerol/O2 molar ratio 1/0.5e1.5/0.0e1.5 glycerol/H2O/O2 molar ratio 1/0.5e3.0 glycerol/CO2 molar ratio 10e35 wt% of glycerol

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

8

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

Fig. 7 e H2/CO molar ratio for the steam reforming of glycerol as function of glycerol/water molar ratio in the feed. Simulation conditions: pressure: 1 bar; temperature: 973.15 K.

The increase of the H2O concentration in the feed results in increased production of H2 and consequently in the elevation of the H2/CO molar ratio, for example, the H2/CO molar ratio was 1.99 for the 1.0/1.5 glycerol/water molar ratio and 2.68 in the 1.0/3.0 glycerol/water molar ratio. Under the conditions evaluated was not observed significant formation of solid carbon in the steam reforming of glycerol. In SR of glycerol was observed the formation of H2, CO, CO2, CH4, unreacted H2O and small amounts of C2H6. At lower temperatures, below 973.15 K, the main compound in the output stream are H2 and CO2. With increasing temperature (above 973.15 K), the main compounds become H2 and CO. Solid carbon formation was observed at conditions of low H2O molar ratios in the feed and at low temperatures (below 873.15 K). Fig. 6 e Hydrogen, carbon monoxide and syngas molar composition in the dry gas for steam reforming of glycerol as function of (a) temperature; (b) pressure and (c) glycerol/ water molar ratio.

of the number of moles of H2 and of 60% of the number of moles of syngas produced in the SR system. The effect of modification in the glycerol/water molar ratio with respect to H2, CO and syngas production is presented in Fig. 6(c). The pressure and the temperature were set at 1 bar and 973.15 K, respectively. It is interesting to verify that higher amounts of H2O in the feed results in higher concentration of H2 in the product. Already CO formation decreases with higher amounts of H2O in the feed. The increase in the H2O molar ratio in the feed of 0.5e3.0 resulted in a 43% of increase in the number of moles of H2 in the product stream. The H2/CO molar ratio obtained as function of the glycerol/water molar ratio is presented in Fig. 7. The conditions used in the simulations were the same used in Fig. 6(c) (973.15 K and 1 bar). H2/CO molar ratio of 2.0 (ideal for FischereTropsch applications) was obtained with a glycerol/water molar ratio of 1.0/1.5 in the feed.

Partial oxidation The PO of glycerol were thermodynamic evaluated using the proposed virial model, and Fig. 8 shows the results obtained. Fig. 8(a) presents the effect of modification in operating temperature, Fig. 8(b) presents the effect of pressure and Fig. 8(c) presents the effect of glycerol/O2 molar ratio. For evaluate the effect of the operating temperature, the pressure and the glycerol/O2 molar ratio in the feed were set at 1 bar and 1.0/0.5, respectively. The elevation of operating temperature results in both, elevation of H2 and CO production, and consequently syngas production (see Fig. 8(a)). This effect was observed until the temperature of 1173.15 K, above this temperature the production of CO and H2 decreased. The effect of the pressure under H2 and CO formation is show in Fig. 8(b). The simulations were performed at constant temperature and glycerol/O2 molar ratio of 973.15 K and 1.0/ 0.5. The elevation of the operation pressure results in reduced production of H2 and CO. The same effect was observed in the steam reforming of glycerol. Fig. 8(c) depicts the effect of modifications in the glycerol/ O2 molar ratio under the production of H2, CO and syngas. The simulations were performed at constant temperature and pressure of 973.15 K and 1 bar. It is interesting to verify that

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

9

Fig. 9 e Solid carbon formation in the partial oxidation of glycerol as function of temperature (a) and pressure (b).

Fig. 8 e Hydrogen, carbon monoxide and syngas molar composition in the dry gas for partial oxidation of glycerol as function of (a) temperature; (b) pressure and (c) glycerol/ water molar ratio.

the elevation of O2 ratio in the feed results in reduced productions of H2 and CO. Higher amounts of CO2 and H2O were observed in that conditions. For example, raising the glycerol/ O2 molar ratio of 1.0/0.25 to 1.0/1.50 resulted in a 26% decrease in the number of moles of H2 and 38% in the number of moles of CO in the system; however, the variation of the molar fraction is small. The PO of glycerol presented significant C(s) formation under the conditions in which the reaction was analyzed by the present work. Thus, Fig. 9 presents the formation of solid carbon as function of the operating temperature (Fig. 9(a)) and pressure (Fig. 9(b)). The simulations conducted to evaluate the temperature effect were performed at constant pressure (1 bar) and the simulations performed to evaluate the pressure effect were performed at constant temperature (973.15 K) in both cases the glycerol/O2 molar ratio was kept constant (1.0/0.5).

Analyzing Fig. 9(a) it is possible to verify that the elevation of the process temperature results in significant reductions in C(s) formation and at 973.15 K the C(s) formation has been eliminated by thermodynamic effects. Fig. 9(b) presents the effect of the pressure over the C(s) production. The elevation of the pressure results in an increase in the C(s) formation. At 1 bar no C(s) formation was observed. The coke formation showed a linear increase with increasing of the operating pressure of the system throughout the pressure range examined. In the PO of glycerol was observed the formation of H2, CO, CO2, CH4, and small amounts of C2H6 and C2H4 at low temperature conditions (below 93.15 K). The O2 and the glycerol were completely converted, the formation of solid carbon was observed in the PO mainly because the reaction was conducted at low O2 concentrations in the feed, to avoid the occurrence of the complete combustion reaction. However, at some interesting conditions for H2 and syngas production the formation of solid carbon was not observed, even using low concentrations of O2.

Autothermal reforming Fig. 10(a) shows the effect of the temperature in respect of H2 and CO formation in the ATR of glycerol. The simulations were performed at constant pressure and glycerol/H2O/O2 molar ratio of 1 bar and 1.0/1.0/1.0 respectively. As observed in the reactions previously showed (SR and PO), the formation of the interest products increased with increasing in temperature for CO in the whole range of temperature examined. Increasing in the H2 production was

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

10

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

to verify that the number of moles of H2 significantly increases with H2O addition and the production of CO decreases at the same conditions. Fig. 11 shows the behavior of the H2/CO molar ratio as a function of the glycerol/H2O/O2 molar ratio. For these simulations the pressure and the temperature was kept constant in 1 bar and 973.15 K. It is interesting to emphasize that the addition of water, with a constant molar ratio of O2 (of 0.5), results in an increase in the H2/CO molar ratio. Molar ratios close to two are achieved with a glycerol/H2O/O2 molar ratio of 1.0/1.5/0.5. Another interesting effect is that the composition of the products stream can be easily modified by changes in the feed composition the amount of O2 is a key issue in the ATR process because it can strongly affect the H2 production in this process. No C(s) formation was observed in the conditions analyzed for this reforming technology. Significant amounts of H2 and CO were produced at high temperature conditions. The formation of C2H6 was also observed in the ATR of glycerol. Glycerol and O2 presented complete conversion in all conditions analyzed, and in all cases were observed H2O production.

Dry reforming

Fig. 10 e Hydrogen, carbon monoxide and syngas molar composition in the dry gas for autothermal reforming of glycerol as function of (a) temperature; (b) pressure and (c) glycerol/H2O/O2 molar ratio.

observed until 973.15 K, above this temperature the number of moles of H2 produced decrease with increasing the temperature. Fig. 10(b) presents the behavior of formation of H2 and CO as function of the pressure of the system. The simulations were performed at constant pressure and glycerol/H2O/O2 molar ratio, of 973.15 K and 1.0/1.0/1.0 respectively. It is interesting to emphasize that the elevation of the pressure results in decreased formation of interest products. Fig. 10(c) depicts the results obtained by the simulations for different glycerol/H2O/O2 molar ratios at constant pressure (1 bar) and temperature (973.15 K). First, the molar ratio of H2O was set at 0.5 and the molar ratio of O2 varied between 0.0 and 1.5. Under these conditions, it is interesting to emphasize that the molar fractions of H2 and CO decreased with the addition of O2. Secondly, the molar ratio of O2 was set at 0.5 and the molar ratio of H2O varied between 0.0 and 2.0. It is interesting

Fig. 12 presents the effects of the temperature (Fig. 12(a)), pressure (Fig. 12(b)) and glycerol/CO2 molar composition in the feed stream (Fig. 12(c)) over the DR of glycerol. The effect of temperature was evaluated at constant pressure (1 bar) and at constant glycerol/CO2 molar ratio (1.0/ 1.0), and the results obtained are presented in Fig. 12(a). This reforming technology presents a peculiar feature in relation to other technologies evaluated; the CO molar ratio was high and H2/CO relation is close to one for temperatures higher than 973.15 K. This behavior was expected due to high amounts of carbon (as CO2) in the feed stream. As observed in the other reforming technologies, the H2 and syngas production significantly increased with increases in the operational temperature. Hydrogen production

Fig. 11 e H2/CO molar ratio for the autothermal reforming of glycerol as function of glycerol/H2O/O2 molar ratio in the feed. Simulation conditions: pressure: 1 bar; temperature: 973.15 K.

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

11

reforming processes analyzed here. The effects of temperature and pressure on the solid carbon formation were presented at Fig. 13(a) and (b) respectively. The elevation of the temperature results in the reduction of the solid carbon formation; the simulations were performed at constant pressure of 1 bar and constant glycerol/CO2 molar ratio, of 1.0/1.0. The elevation of the operating pressure presents a negative effect, favoring the formation of solid carbon in the system; the simulations were performed at constant temperature (973.15 K) and constant glycerol/CO2 molar ratio (1.0/1.0). These effects were also observed in the partial oxidation of glycerol too. The DR of glycerol presented the highest production of CO among all reforming processes analyzed, this behavior can be explained by the increase of the C and O atoms in the feed. The formation of small amounts of C2H4 and C2H6 were observed in conditions of elevated pressure (above 3 bar) and low temperatures (below 873.15 K), in that conditions significant amounts of solid carbon was also observed.

Supercritical water gasification The proposed model was used to evaluate the SCWG reaction. The effects of temperature, pressure and inlet glycerol composition were analyzed and the results are presented in Fig. 14(a)e(c), respectively. The effect of modifications in the operating temperature was analyzed at constant pressure of 250 bar and at constant glycerol/water molar ratio (1.0/0.9, that is the same of 36.2 wt %) and the results are presented at Fig. 14(a). The elevation of the operating temperature resulted in increased production of H2 in the system, as has been

Fig. 12 e Hydrogen, carbon monoxide and syngas molar composition in the dry gas for dry reforming of glycerol as function of (a) temperature; (b) pressure and (c) glycerol/CO2 molar ratio.

increases with increases in the temperature and reaches the maximum at 1073.15 K. The effect of operating pressure was studied at constant temperature (973.15 K) and feed composition (1.0/1.0 e glycerol/CO2). Higher pressures have a negative effect on hydrogen and syngas production, the maximum amounts of hydrogen and syngas were achieved at atmospheric pressure. Fig. 12(c) presents the H2, CO and syngas composition in dry gas as a function of glycerol/CO2 molar ratio in the inletstream at constant temperature and pressure of 973.15 K and 1 bar, respectively. The addition of CO2 results in significant increases of 61% in number of moles of CO produced in the reaction. The addition of CO2 presents another interesting effect, the reduction of solid carbon formation at constant temperature. The dry reforming of glycerol was the process that showed greater susceptibility to solid carbon formation from all

Fig. 13 e Solid carbon formation in the dry reforming of glycerol as function of temperature (a) and pressure (b).

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

12

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

temperature (973.15 K) and constant glycerol/water molar ratio (1.0/9.0). The elevation of the operational pressure has no significant effect on the production of H2 or syngas. Similar results were observed in previous published papers of our group for the SCWG of glucose, cellulose and microalgal biomass [62,71]. The effect of the feed concentration over the production of H2 and syngas is presented in Fig. 14(c) as function of the mass concentration of glycerol in the feed. The simulations were performed at constant pressure and temperature of 250 bar and 973.15 K respectively. The addition of glycerol results on reduced productions of H2 and slightly increased the production of CO. This characteristic of the SCWG reaction is interesting when the H2 production is the main objective, because the low concentration of CO in the system facilitate the subsequent separation processes [46]. The SCWG reaction had the lowest CO formation from all processes evaluated. The excess of water used can explain this behavior. The water reacts with the CO formed through the water gas shift reaction (Eq. (4) in Table 1) to produce CO2 and more H2. In previous works of our group, the use of CO2 as a co-reactant proved to be an effective way to improve the CO production in the SCWG of different biomass materials [62,71]. Significant amounts of CH4 are observed at low temperature conditions (873.15 and 973.15 K). At 873.15 K, the main products are CH4 and CO2, with the elevation of reaction temperature H2 become the main product in the SCWG of glycerol. The high pressures used in this process could explain the formation of small amounts of C2H4, C2H6 and C3H8 especially at low temperatures. In any of the conditions analyzed was observed the formation of solid carbon. In the following section, all technologies were compared with respect to hydrogen and syngas production.

Fig. 14 e Hydrogen, carbon monoxide and syngas molar composition in the dry gas for supercritical water gasification of glycerol as function of (a) temperature; (b) pressure and (c) glycerol/water molar ratio.

observed in all other processes analyzed. The CO formation also increased with increasing temperature, shuffled remained low throughout the range of temperature analyzed. Fig. 14(b) shows the effect of modifications in the operating pressure on the composition of H2, CO and syngas in the product stream, the simulations were performed at constant

Comparison between the technologies Table 5 shows the better conditions for H2 production observed in all reaction analyzed with the operational temperature fixed at 1073.15 K. The SR, PO, ATR and DR were analyzed at 1 bar and the SCWG was analyzed at 250 bar. Analyzing Table 5 it’s interesting to emphasize that higher molar fraction of H2 was observed in the SR reaction (with a molar ratio glycerol/water ¼ 1.0/3.0 in the feed stream). The lower molar fraction of H2 was observed for the SCWG reaction. However this process presented the lowest production of CO between all processes analyzed, with that, the subsequent purification processes of H2 become simpler for the SCWG.

Table 5 e Better conditions for H2 production for the different reforming technologies analyzed. Process

SR PO ATR DR SCWG

Pressure (bar)

1 1 1 1 250

Temperature (K)

1073.15 1073.15 1073.15 1073.15 1073.15

Molar ratio

1/3 (Gly/H2O) 1/0.5 (Gly/O2) 1/1/1 (Gly/H2O/O2) 1/1 (Gly/CO2) 1/9 (Gly/H2O)

Molar fraction H2

Syngas

61.9 52.9 51.0 46.2 40.5

88.4 92.8 81.5 92.7 48.2

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

13

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

Table 6 e Better conditions for syngas production for the different reforming technologies analyzed. Process

SR PO ATR DR SCWG

Pressure (bar)

1 1 1 1 250

Temperature (K)

1273.15 873.15 973.15 873.15 1273.15

Molar ratio

Molar fraction

1/3 (Gly/H2O) 1/0.5 (Gly/O2) 1/1/1 (Gly/H2O/O2) 1/1 (Gly/CO2) 1/9 (Gly./H2O)

The better conditions observed for syngas production (with a molar ratio more close to 2.0 as possible) was presented in Table 6. The SR, PO, ATR and DR were analyzed at 1 bar and the SCWG was analyzed at 250 bar again. Each case presented a different temperature of reaction to the better condition for syngas production. Analyzing Table 6 it is interesting to emphasize that the highest molar fraction of H2 was observed to the SR reaction (91.4% of the dry gas), with 1273.15 K of operational temperature and with a glycerol/water molar ratio of 1.0/3.0, in that conditions the H2/CO molar ratio was 1.99. The SCWG of glycerol presented the highest H2/CO molar ratio between all analyzed processes, 3.21 at 1273.15 K and with a glycerol/H2O molar ratio of 1.0/9.0. The lower molar fraction of syngas was observed for the dry reforming of glycerol (62.5%) at 873.15 K and with a glycerol/CO2 molar ratio of 1.0/1.0, in that conditions the H2/CO molar ratio was 2.00, however was observed solid carbon formation (1.16 mol of C(s)). Solid carbon formation was observed in the PO too (0.65 mol of C(s) at 873.15 K and with a glycerol/O2 molar ratio of 1.0/0.5). Table 7 shows the comparison between all reforming processes with respect to H2 and CO production and the results obtained for the H2/CO molar ratio. The same glycerol/X (where X are H2O, O2, CO2 or combinations between then) molar ratio (1.0/3.0) was used in all simulations performed. The SR, PO, ATR and DR were simulated at 1 bar and the SCWG was simulated at 250 bar and three operational temperatures (873.15, 973.15 and 1073.15 K) were evaluated.

Ratio

Moles

H2

Syngas

H2/CO

C(s)

60.8 46.5 51.6 41.7 57.7

91.4 65.4 78.0 62.5 75.7

1.99 2.44 1.95 2.00 3.21

0.00 0.65 0.00 1.16 0.00

Analyzing Table 7 it’s possible to verify that the highest concentrations of H2 are observed in the steam reforming reaction. Still in Table 7 it’s possible to verify that the SR and ATR presented H2/CO molar ratio close to 2.0 in the conditions in which the reactions were analyzed. The SCWG presented the highest H2/CO molar ratio between all reforming processes analyzed. Another interesting effect is that the elevation of operational temperature results in reductions of the H2/CO molar ratio in the product, by inducing the production of CO in the system. This behavior was observed in all reforming processes analyzed.

Conclusion A thermodynamic analysis for hydrogen and syngas production from glycerol using different reforming technologies has been performed. The SR, PO, ATR, DR and SCWG were evaluated. The composition of the products was calculated using the Gibbs energy minimization approach in combination with the virial equation of state. The model was validated with simulated and experimental data obtained in literature with good agreement between then. Hydrogen yields were found to increase directly with temperature in all reforming processes analyzed. The increase of the operating pressure results in large amounts of solid carbon in the product stream in the partial oxidation and dry

Table 7 e Comparison between all reforming technologies of glycerol. Reforming technology

SCWG

SR

ATR

OR

DR

Temperature (K)

873.15 973.15 1073.15 873.15 973.15 1073.15 873.15 973.15 1073.15 873.15 973.15 1073.15 873.15 973.15 1073.15

Pressure (bar)

250 250 250 1 1 1 1 1 1 1 1 1 1 1 1

Molar fraction (mol%)

Molar ratio

H2

CO

Syngas

H2/CO

8.5 15.8 25.3 55.0 61.7 61.9 46.1 47.8 46.3 19.8 17.9 16.0 29.0 32.1 31.8

1.6 4.7 10.5 14.8 23.0 26.5 14.5 21.0 25.3 6.8 9.5 12.0 24.7 43.7 47.7

10.1 20.6 35.7 69.8 84.8 88.4 60.6 68.8 71.5 26.6 27.4 28 53.7 75.8 79.4

5.34 3.34 2.42 3.72 2.68 2.34 3.17 2.28 1.84 2.92 1.88 1.34 1.17 0.73 0.67

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

14

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

Subscripts i refers to component i m refers to element m in a component

reforming processes. In the other processes the formation of solid carbon was not observed. The syngas production with H2/CO molar ratio close to 2.0 was found without solid carbon formation only in the steam reforming of glycerol. The proposed model implemented in the GAMS software and solved with the CONOPT solver, show to be fast and reliable (since computational time close to 1 s were observed in the simulations and the validation with experimental and simulated data obtained in the literature showed that the model presents a good predictive ability) to perform the thermodynamic analysis of the different glycerol reforming technologies analyzed.

Greek symbols m chemical potential m0i chemical potential of component i at reference condition Chemical potential of solid carbon at reference m0CðsÞ condition fugacity coefficient of component i fi bi f fugacity coefficient of component i in the mixture u acentric factor

Acknowledgments

references

The authors gratefully acknowledge the financial support from CAPES e Coordenac¸a˜o de Aperfeic¸oamento de Pessoal de Nı´vel Superior, FAPESP e Sa˜o Paulo Research Foundation (Process: 2011/20666-8) and CNPq e Conselho Nacional de Desenvolvimento Cientı´fico e Tecnolo´gico, Brazil.

Nomenclature ami ATR B Bij Cpi DR G Hi nki n0i nsCðsÞ MRE NC NCE NE NPE P PC PO R SCWG SR T TC VC xliterature j;i xsimulated j;i yi

number of atoms of element m for component i autothermal reforming second virial coefficient cross second virial coefficient heat capacity of pure component i Dry reforming Gibbs Energy partial molar enthalpy of component i number of moles of component i in the phase k initial number of moles of component i number of moles of solid carbon in solid phase mean relative error number of components number of components for each experimental point number of elements number of experimental points pressure critical pressure partial oxidation universal gas constant supercritical water gasification steam reforming temperature critical temperature critical volume experimental or simulated value obtained in the literature value obtained by the simulations of the present work molar fraction of component i

Superscripts g gas phase l liquid phase s solid phase

[1] Saxena RC, Seal D, Kumar S, Goyal HB. Thermo-chemical routes for hydrogen rich gas from biomass: a review. Renew Sustain Energy Rev 2008;12:1909e27. [2] Ma F, Hanna MA. Biodiesel production: a review. Bioresour Technol 1999;70:1e15. [3] Pagliaro M, Ciriminna R, Kimura H, Rossi M, Della Pina C. From glycerol to value added products. Angew Chem Int Ed 2007;46:4434e40. [4] Vaidya PD, Rodrigues AE. Glycerol reforming for hydrogen production: a review. Chem Eng Technol 2009;32:1463e9. [5] Zhang B, Tang X, Li Y, Xu Y, Shen W. Hydrogen production from steam reforming of ethanol and glycerol over ceria supported metal catalysts. Int J Hydrogen Energy 2007;32:2367e73. [6] Buffoni IN, Pompeo F, Santori GF, Nichio NN. Nickel catalysts applied in steam reforming of glycerol for hydrogen production. Catal Commun 2009;10:1656e60. [7] Adhikari S, Fernando SD, Haryanto A. Hydrogen production from glicerin by steam reforming over nickel catalysts. Renew Energy 2008;33:1097e100. [8] Thyssen VV, Maia TA, Assaf EM. Ni supported on La2O3eSiO2 used to catalyze glycerol steam reforming. Fuel 2013;105:358e63. [9] Iriondo A, Barrio VL, Cambra JF, Arias PL, Guemez MB, Sanchez MC, et al. Glycerol steam reforming over Ni catalysts supported on ceria and ceria promoted alumina. Int J Hydrogen Energy 2010;35:11622e33. [10] Pompeo F, Santori GF, Nichio NN. Hydrogen production by glycerol steam reforming with Pt/SiO2 and Ni/SiO2 catalysts. Catal Today 2011;172:183e8. [11] Pompeo F, Santori GF, Nichio NN. Hydrogen and/or syngas from steam reforming of glycerol. Study of platinum catalysts. Int J Hydrogen Energy 2010;35:8912e20. [12] Wang H, Wang X, Li M, Li S, Wang S, Ma X. Thermodynamic analysis of hydrogen production from glycerol autothermal reforming. Int J Hydrogen Energy 2009;34:5683e90. [13] Wang W. Thermodynamic analysis of glycerol partial oxidation for hydrogen production. Fuel Process Technol 2010;91:1401e8. [14] Yang G, Yu H, Peng F, Wang H, Yang J, Xie D. Thermodynamic analysis of hydrogen generation via oxidative steam reforming of glycerol. Renew Energy 2011;36:2120e7. [15] Byrd AJ, Pant KK, Gupta RB. Hydrogen production from glycerol by reforming in supercritical water over Ru/Al2O3 catalyst. Fuel 2008;87:2956e60. [16] Chakinala AG, Brilman DWF, van Swaaij WPM, Kersten SRA. Catalytic and non catalytic supercritical water gasification of

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34] [35]

microalgae and glycerol. Indust Eng Chem Res 2009;49:1113e22. Guo S, Guo L, Cao C, Yin J, Lu Y, Zhang X. Hydrogen production from glycerol by supercritical water gasification in a continuos flow tubular reactor. Int J Hydrogen Energy 2012;37:5559e68. van Bennekon JG, Venderbosch RH, Assink D, Heeres HJ. Reforming of methanol and glycerol in supercritical water. J Supercrit Fluids 2011;58:99e113. Voll FAP, Rossi CCRS, Silva C, Guirardello R, Souza ROMA, Cabral VF, et al. Thermodynamic analysis of supercritical water gasification of methanol, ethanol, glycerol, glucose and cellulose. Int J Hydrogen Energy 2009;34:9737e44. Wang X, Li M, Wang M, Wang H, Li S, Wang S, et al. Thermodynamic analysis of glycerol dry reforming for hydrogen and synthesis gas production. Fuel 2009;88:2148e53. Ferna´ndez Y, Arennilas A, Bermu´dez JM, Mene´ndez JA. Comparative study of conventional and microwave assisted pyrolysis, steam and dry reforming of glycerol for syngas production, using a carbonaceous catalyst. J Anal Appl Pyrolysis 2010;88:155e9. Kale GR, Kulkarni BD. Thermodynamic analysis of dry autothermal reforming of glycerol. Fuel Process Technol 2010;91:520e30. Valliyappan T, Bakhshi NN, Dalai AK. Pyrolysis of glycerol for the production of hydrogen or syngas. Bioresour Technol 2008;99:4476e83. Adhikari S, Fernando S, Gwaltney SR, Filip To SD, Bricka RM, Steele PH, et al. A thermodynamic analysis of hydrogen production by steam reforming of glycerol. Int J Hydrogen Energy 2007;32:2875e80. Holladay JD, Hu J, King DL, Wang Y. An overview of hydrogen productions technologies. Catal Today 2009;139:244e60. Rass-Hansen J, Johansson R, MØller M, Christensen CH. Steam reforming of technical bioethanol for hydrogen production. Int J Hydrogen Energy 2008;33:4547e54. Adhikari S, Fernando S, Haryanto A. Production of hydrogen by steam reforming of glicerin over alumina supported metal catalysts. Catal Today 2007;129:355e64. Adhikari S, Fernando S, Haryanto A. A comparative thermodynamic and experimental analysis on hydrogen production by steam reforming of glicerin. Energy Fuels 2007;21:2306e10. Slinn M, Kendall K, Mallon C, Andrews J. Steam reforming of biodiesel by-product to make renewable hydrogen. Bioresour Technol 2008;99:5851e8. Oliveira ELG, Grande CA, Rodrigues ARE. Methane steam reforming in large pore catalyst. Chem Eng Sci 2010;65:1539e50. Sanchez EA, Comelli RLA. Hydrogen by glycerol steam reforming on a nickel alumina catalyst: deactivation processes and regeneration. Int J Hydrogen Energy 2012;37:14740e6. Dieuzeide ML, Iannibeli V, Jobbagi M, Amadeo N. Steam reforming of glycerol over Ni/Mg/g-Al2O3 catalysts. Effect of calcination temperatures. Int J Hydrogen Energy 2012;37:14926e30. ´ lvarez A, Domı´nguez MI, Romero-Sarria F, Bobadilla LF, A Centeno MA, Montes M, et al. Influence of the shape of Ni catalyst in the glycerol steam reforming. Appl Catal B e Environ 2012;123:379e90. Cheng GK, Foo SY, Adesina AA. Steam reforming of glycerol over Ni/Al2O3 catalysts. Catal Today 2011;178:25e33. Wang C, Dou B, Chen H, Song Y, Xu Y, Du X, et al. Hydrogen production from steam reforming of glycerol by NieMgeAl based catalysts in fixed bed reactor. Chem Eng J 2013;220:133e42.

15

[36] Wang C, Dou B, Chen H, Song Y, Xu Y, Du X, et al. Renewable hydrogen production from steam reforming of glycerol by NieCueAl, NieCueMg, NieMg catalysts. Int J Hydrogen Energy 2013;38:3562e71. [37] Ahmed S, Krumpelt. Hydrogen from hydrocarbon fuels for fuel cells. Int J Hydrogen Energy 2001;26:291e301. [38] Carrettin S, McMorn P, Johnston P, Griffin K, Kiely CJ, Hutchings GJ. Oxidation of glycerol using supported Pt, Pd and Au catalysts. Phys Chem Chem Phys 2003;5:1329e36. [39] Rennard DC, Kruger JS, Schmidt LD. Autothermal catalytic partial oxidation of glycerol to syngas and non-equilibrium products. Chem Sus Chem 2009;2:89e98. [40] Rabenstein G, Hacker V. Hydrogen for fuel cells from ethanol by steam reforming, partial oxidation and combined autothermal reforming: a thermodynamic analysis. J Power Sources 2008;185:1293e304. [41] Qi A, Wang S, Fu G, Wu D. Autothermal reforming of noctane on Ru-based catalysts. Appl Catal A e Gen 2005;293:71e82. [42] Dauenhauer PJ, Salge JR, Schmidt LD. Renewable hydrogen by autothermal steam reforming of volatile carbohydrates. J Catal 2006;244:238e47. [43] Douette AMD, Turn SQ, Wang W, Keffer VI. Experimental investigation of hydrogen production from glycerin reforming. Energy Fuels 2007;21:3499e504. [44] Edwards JH, Maitra AM. The chemistry of methane reforming with carbon dioxide and its current and potential applications. Fuel Process Technol 1995;42: 269e89. [45] Mizuno T, Goto M, Kodama A, Hirose T. Supercritical water oxidation of a model municipal solid waste. Indust Eng Chem Res 2000;39:2807e10. [46] Kruse A. Supercritical water gasification. Biofuels Bioprod Biorefining 2008;2:415e37. [47] Savage PE. Heterogenous catalysis in supercritical water. Catal Today 2000;62:167e73. [48] Calzavara Y, Joussot-Dubien C, Boissonnet G, Sarrade S. Evaluation of biomass gasification in supercritical water process for hydrogen production. Energy Convers Manag 2005;46:615e31. [49] Guo Y, Wang SZ, Xu DH, Gong YM, Ma HH, Tang XY. Review of catalytic supercritical water gasification for hydrogen production from biomass. Renew Sustain Energy Rev 2010;14:334e43. [50] Hao XH, Guo LJ, Mao X, Zhang XM, Chen XJ. Hydrogen production from glucose used as a model compound of biomass gasified in supercritical water. Int J Hydrogen Energy 2003;28:55e64. [51] Clifford T. Fundamentals of supercritical fluids. New York: Oxford University Press; 1998. [52] May A, Salvado´ J, Torras C, Montane´ D. Catalytic gasification of glycerol in supercritical water. Chem Eng J 2010;160:751e9. [53] Smith WRM, Missen RW. Chemical reaction equilibrium analysis: theory and algorithms. JohnWiley & Sons; 1982. [54] Pitzer KS, Curl RF. The volumetric and thermodynamic properties of fluids III. Empirical equation of the second virial coefficient. J Am Chem Soc 1957;20:263e72. [55] Tsonopoulos C. An empirical correlation of second virial coefficients. AIChE J 1974;20:263e72. [56] Castello D, Fiori L. Supercritical water gasification of biomass: thermodynamic constraints. Bioresour Technol 2011;102:7574e82. [57] Castillo J, Grossmann IE. Computation of phase and chemical equilibria. Comput Chem Eng 1981;5:99e108. [58] Freitas ACD, Guirardello R. Oxidative reforming of methane for hydrogen and synthesis gas production: thermodynamic equilibrium analysis. J Nat Gas Chem 2012;21:571e80.

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130

16

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 4 ) 1 e1 6

[59] Lu Y, Guo X, Zhang X, Yan Q. Thermodynamic modeling and analysis of biomass gasification for hydrogen production in supercritical water. Chem Eng J 2007;131:233e44. [60] Nichita DV, Gomez S, Luna E. Multiphase equilibria calculation by direct minimization of Gibbs free energy with a global optimization method. Comput Chem Eng 2002;26:1703e24. [61] Rossi CCRS, Berezuk ME, Cardozo-Filho L, Guirardello R. Simultaneous calculation of chemical and phase equilibria using convexity analysis. Comput Chem Eng 2011;35:1226e37. [62] Freitas ACD, Guirardello R. Supercritical water gasification of glucose and cellulose for hydrogen and syngas production. Chem Eng Trans 2012;27:361e6. [63] Castier M. Solution of isochoric-isoenergetic flash problem by direct entropy maximization. Fluid Phase Equilibria 2009;276:7e17. [64] White WB, Johnson SM, Danzig GB. Chemical equilibrium in complex mixtures. J Chem Phys 1958;28:751e5. [65] Brooke K, Meeraus D, Raman R. GAMS e a user’s manual; 1998.

[66] Polling BP, Prausnitz JM, O’Connel PJ. The properties of gases and liquids. 5th ed. New York: McGraw Hill; 2000. [67] Reid RP, Prausnitz JM, Sherwood TK. The properties of gases and liquids. 4th ed. New York: McGraw Hill; 1987. [68] DIPPR. DIADWM Public v. 1.2. Design Institute for physical property data. Information and data evaluation manager; 2000. [69] Adhikari S, Fernando SD, To SDF, Bricka RM, Steele PH, Haryanto A. Conversion of glycerol to hydrogen via a steam reforming process over nickel catalysts. Energy Fuels 2008;22:1220e6. [70] Wang X, Li M, Li S, Wang H, Wang S, Ma X. Hydrogen production by glycerol steam reforming with/without calcium oxide sorbent: a comparative study of thermodynamic and experimental work. Fuel Process Technol 2010;91:1812e8. [71] Freitas ACD, Guirardello R. Thermodynamic analysis of supercritical water gasification of microalgae biomass for hydrogen and syngas production. Chem Eng trans 2013;32:553e8.

Please cite this article in press as: Freitas ACD, Guirardello R, Comparison of several glycerol reforming methods for hydrogen and syngas production using Gibbs energy minimization, International Journal of Hydrogen Energy (2014), http://dx.doi.org/ 10.1016/j.ijhydene.2014.03.130