- Email: [email protected]
Comparison between a micro reactor with multiple air inlets and a monolith reactor for oxidative steam reforming of diesel
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Comparison between a micro reactor with multiple air inlets and a monolith reactor for oxidative steam reforming of diesel Moa Z. Granlund a,*, Oliver Go€rke b, Peter Pfeifer b, Lars J. Pettersson a a
KTH Royal Institute of Technology, Department of Chemical Engineering and Technology, SE-100 44 Stockholm, Sweden b Karlsruhe Institute of Technology, Institute for Micro Process Engineering (IMVT), D-76344 Eggenstein-Leopoldshafen, Germany
article info
abstract
Article history:
In order to lower the emission from idling heavy-duty trucks auxiliary power units can be
Available online 11 July 2014
implemented. Due to limited space available on-board the truck the units needs to be both efficient and compact. One alternative for these units is a fuel cell supplied with hydrogen
Keywords:
from a fuel reformer. Today, mostly monolithic reactors are used in the field of oxidative
Oxidative steam reforming
steam reforming of fuels, which has some challenges that need to be addressed before a
Diesel
possible breakthrough occurs on the market. One is the temperature gradient developed
Micro reactor
over the length of the monolith as a consequence of the sequential reactions. This could be
Monolith reactor
improved by using a metallic micro reactor with better heat integration between the re-
Multiple air inlets
action zones and further improving the integration with multiple air inlets along the catalytic bed. The aim with this study was to compare a conventional monolith reactor for oxidative steam reforming of fuel with a novel micro reactor design where air was dosed at four different positions along the reactor channels. The experiments were not necessarily conducted autothermal, i.e. a heating jacket was applied for operation. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction It is estimated that heavy-duty vehicles are responsible for about 30% of the total emissions of nitrogen oxides (NOx) and particulate matters (PM) [1]. In addition the engine is idling 30e40% of the time just to supply the vehicle with electricity [1,2]. The emission during idling is a considerable part of the total emission for the vehicle and there is an urgent need to
decrease those in order to meet future emission legislations [3]. One alternative to solve this could be to implement a fuel cell system where the hydrogen is produced via on-board reforming of diesel, a diesel fuel cell auxiliary power unit (FC APU) [4]. There are three different processes for reforming diesel to hydrogen-rich gas: partial oxidation, steam reforming and oxidative steam reforming (OSR); the latter could be operated self-sustaining as autothermal reforming (ATR). The three
* Corresponding author. Tel.: þ46 8 790 91 50. E-mail address: [email protected] (M.Z. Granlund). http://dx.doi.org/10.1016/j.ijhydene.2014.06.096 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
18038
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
processes have different advantages and disadvantages and for on-board fuel reforming OSR is the most suitable option [5]. The main reason is that the process is autothermal and therefore no external heating is needed. The autothermal characteristic is a consequence of the process being a combination of the exothermic partial oxidation and endothermic steam reforming. Other advantages are the short start-up time due to the fast thermal response and the relatively high hydrogen yield [6]. Due to the limitations of space onboard heavy-duty vehicles the design of the FC APU must be compact. An OSR reactor is normally based on a tubular reactor where the catalytic material is coated on a honeycomb cordierite monolith. This minimizes the pressure drop and also makes the reformer resistant to vibrations [7]. One of the most prominent disadvantages with this reactor design is the characteristic temperature gradient, developing along the monolith as a consequence of the combination of the two sequential reactions. The fast and exothermic partial oxidation reaction takes place at the entrance, causing a sharp increase in temperature and the slower, endothermic steam reforming reaction consumes the heat towards the outlet of reactor [8]. The heat transport from the initial partial oxidation reaction is transferred along the monolith only by convection and the interaction between the two reaction zones is poor. The initial raise in temperature can cause thermal deactivation of the catalytic material by the potential formation of hot spots. In order to limit the temperature increase in the beginning, better heat integration between the two reaction zones is necessary. One option to decrease the temperature gradient over the reactor and to increase the heat and mass transfer could be to use a micro reactor with better heat integration [9,10]. With a micro reactor it is possible to make a design closer connecting the exothermic and the endothermic part of the reactor. Another advantage with micro reactors is that metal is used which increases the heat conductivity of the reactor structure compared to ceramic monoliths [11]. One completely different possibility, already successfully tested at the Institute for Micro Process Engineering at Karlsruhe Institute of Technology, is to indirectly couple endothermic steam reforming with the exothermic combustion of anode-off gas in two different foil arrangements inside a metallic micro reactor [12]. In the current publication, however, the attempt was to further improve the heat integration between the two reaction zones in a direct coupling. Therefore, a micro reactor with multiple air dosing points axially distributed along the reaction channels was designed. The idea is to further improve the heat management by dosing the air axially along the reactor channels by adapting the kinetics between the partial oxidation and steam reforming reactions. With a better heat integration between the partial oxidation and steam reforming parts of the catalyst the temperature stress that usually occurs in the beginning of a monolith would be decreased. The integration could potentially also lead to higher temperatures in the steam reforming part, increasing the rate of the steam reforming reaction with a higher H2 yield as result. The scope of our work is to evaluate if micro structured reactors are a viable alternative to monolith reactors for on-
board production of hydrogen. The deactivation of the catalytic material caused by thermal aging, i.e. sintering and collapsing of the pore structure, might therefore be decreased as previously shown in Ref. [11].
Experimental Catalyst preparation The catalytic material used in the comparative study between the two different reactor designs was 3 wt.% Rh supported on high surface area CeO2eZrO2 (MEL Chemicals, 16.5% CeO2/ ZrO2) prepared by the incipient wetness method. As metal precursor, Rh nitrate solution was used (Rh(NO3)3 solution, 9.53 w/w, SigmaeAldrich). After the impregnation the catalytic material was calcined in air at 800 C for 3 h. Subsequently, the catalytic material was suspended in ethanol and ball milled for 24 h. For the monolith reactor the catalytic material was applied to a cordierite monolith (400 cpsi) by dip coating until the washcoat loading was approximately 20 wt.%. After the coating the monolith was calcined in air at 800 C for 3 h. The inserted catalyst mass was 1.15 g. The diameter and length of the monolith was 20 mm and 30 mm, respectively, resulting in an inner reaction volume of approximately 5.35 cm3 [13]. Before coating the micro structured reactor, coating tests with single metallic micro structured foils were conducted. Prior to coating the Nicrofer® foils were calcined at 800 C for 3 h in air to generate a thin oxide film on the surface. Subsequently the foils were coated dropwise with the catalyst slurry. The coated foils were then investigated with scanning electron microscope and microprobe analysis. The obtained coating showed an excellent adhesion on the micro channel foils and the catalytic material was homogeneously distributed on the foils, which can be seen in Fig. 1. For the micro reactor with multiple air inlets the micro structured foil stack was welded by diffusion bonding and air inlets were attached to the flow chambers by tungsten inertgas welding. The stack was calcined at 800 C for 3 h and afterwards flow-coated with the same slurry as it was used for dipcoating of the single micro structured foils. After drying and calcination of the catalyst for 3 h at 800 C in air, the flow chambers were welded onto the coated stack on the reaction side by tungsten inert gas welding. The inserted catalyst mass was also 1.15 g. The fuel used in the study was synthetic diesel of type biomass-to-liquid (NExBTL, Neste Oil). The main advantage with using synthetic diesel is its high content of alkanes, implying low content of aromatic and polyaromatic hydrocarbons that are prone to coke and lead to catalyst deactivation. Another benefit with synthetic diesel is its low sulfur content [14]. The characteristics of the synthetic diesel are summarized in Table 1.
Reactor set-up The OSR experiments were performed in a bench-scale setup where either the micro reactor or the monolith reactor was
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
18039
Fig. 1 e Scanning electron microscopy picture of a micro structured Nicrofer® foil coated with 3 wt.% Rh/CeO2eZrO2.
mounted. For the monolith reactor experiments a horizontally mounted high-temperature stainless steel (Sandvik 253 MA, Øinner ¼ 24.3 mm) reactor was used. This reactor configuration is described in detail elsewhere [15]. The integration of the micro reactor in the reactor setup can be seen in Fig. 2. The temperature measured on the outside of the micro reactor was assumed to be the reactor temperature, since the temperature gradient was negligible. Due to limitations of the external heating system of the micro reactor the maximum operating temperature was 800 C. The reformate was analyzed with an FTIR instrument, a mass spectrometer (V&F Inc., H-sense) and a gas chromatograph (Varian CP-3800). The FTIR instrument (MKS Multigas™ 230 HS) was equipped with a high speed spectrometer and was used to analyze the concentrations of H2O, CO, CO2, CH4, short-chain hydrocarbons and unconverted fuel online in the wet reformate. The mass spectrometer monitored the H2 concentrations in the dry reformate. The gas chromatograph was equipped with two sequentially packed columns, a Porapak followed by a molecular sieve (5 Å), and a thermal conductivity detector used to analyze the concentrations of H2, CO, CO2, O2 and N2 in the dry reformate. Prior to the experiments in the monolith reactor the catalyst was pre-treated in N2 at 950 C for 30 min to accelerate the thermal aging of the material. The reactor design evaluation experiments were performed in the temperature range of 700e900 C with steady state measurements at 50 C intervals. The experiments were performed at a gas hourly space velocity (GHSV) of 50,000 h1.
Table 1 e Characteristics of NExBTL [14]. Properties Sulfur content Min. boiling point Dist. 95% evap. Aromaticsa Carbon, C Hydrogen, H a
Value
Unit
<1 180 293 <0.3 84.9 15.1
ppm C C wt.% wt.% wt.%
Includes mono-, di-, and tri þ - aromatics.
To evaluate the heat integration capacity of the micro reactor, 100% of the air feed was dosed through the first air inlet (see Fig. 3). The influence of dosing air along the reactor channels, through the multiple air inlets (see Fig. 3) of the micro reactor was evaluated at 750 C and GHSV ~40,000 h1. During all experiments, both monolithic reactor and micro reactor, the operation parameters H2O/C and O2/C were kept constant at 2.5 and 0.45, respectively.
Micro reactor design The metallic micro reactor with multiple air inlets (fuel inlet/air inlet 1 þ 3 additional air inlets distributed axially along the reactor channels) was manufactured at the Institute for Micro Process Engineering (IMVT) at Karlsruhe Institute of Technology (KIT) [16]. The idea with the design is to combine three strategies into one micro reactor design, all to minimize the typical temperature gradient developing axially in the reactor channels of a OSR reactor. The three strategies are: Enhancing the heat transfer, both in axial and radial direction, by manufacturing the reactor in metal. Further improving the heat transfer between the exothermic partial oxidation and the endothermic steam reforming by dosing them in opposite directions. Dosing the air at multiple places axially along the reactor channels to distribute the partial oxidation of diesel. The micro reactor consists of multiple foils (each containing 27 reaction channels) welded together by diffusion bonding and mounted in the outer shell. A picture of the bonded foils is shown in Fig. 4. The fuel inlet/air inlet 1 and the outlet are part of the outer shell and are not included in the picture. The air fed in air inlet 1,2 and 3 is introduced in each reactor channel via holes positioned: 5 mm, 40 and 80 mm, downstream the fuel inlet/air inlet 1. After air inlet 3, the gas flow is redirected 180 and returned towards the outlet. A schematic picture of one reactor channel is shown in Fig. 4. The geometric data of the micro reactor is summarized in Table 2. When the reactor channels are coated with catalyst, both the height and width are reduced by approximately 150 mm (75 mm coating thickness).
18040
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 2 e Schematic picture of the experimental setup with the micro reactor installed. A detailed experimental setup with the monolith reactor installed can be found in Ref. [15].
Product analysis and calculations SHx Cy ¼ In the calculations the diesel is approximated to C14H26. Fuel conversion is defined as the amount of fuel converted to CO, CO2 or CH4 according to Eq. (1), where Fi denotes the respective molar flows. Xfuel ð%Þ ¼
FCO þ FCO2 þ FCH4 $100 14 Ffuel
(1)
The H2 yield is defined as the percentage of the theoretical, maximum amount H2 formed when assuming the only products are CO2 and H2, i.e. all CO is converted to CO2 through the water-gas shift reaction (WGS): YH2 ð%Þ ¼
FH2 þ FCO $100 FH2 max
(2)
The selectivity for CO2 compared to CO is calculated according to Eq. (3). SCO2 ð%Þ ¼
FCO2 $100 FCO2 þ FCO
The selectivity of hydrocarbon is shown in Eq. (4).
Fig. 3 e Schematic picture of one reactor channel in the micro reactor design.
(3)
y$FHx Cy FC1 ;in
(4)
Results and discussion The design of the micro reactor is based on results from previous experiments with ceramic monolith reactors. The designing of the micro reactor was performed in multiple steps. The first step was to ensure that the total flow is equally distributed even in the case when air dosing is performed downstream the reactor channels. The results from the CFD simulations can be seen in Fig. 5 where the concentration of O2 is correlated to the color. In Fig. 5a the flow is simulated when air is dosed through air inlet 2 while in Fig. 6b the air is dosed, equally divided, through air inlets 2e4. From the CFD simulations it was concluded that the flow distribution was unaffected by the introduction of additional air at a later stage in the reactor channels. The second step was to evaluate the design concept of the foils that were going to be used in the micro reactor. This was executed in a clamped reactor setup, where the foils were clamped together after they had been coated with catalyst (3 wt.% Rh supported on CeO2eZrO2). The step enabled evaluation of the substrate's effect on the catalyst and also to evaluate the coating technique. The clamped reactor was evaluated through OSR of propane. The results from the clamped reactor design showed that the catalyst was not affected by the change in substrate. The next step was to evaluate the distribution of catalytic washcoat inside the bonded micro reactor. An uneven coating would mean that some channels are not coated and no catalytic activity appears here. This also influences the pressure drop, which decreases in the uncoated channels, causing some diesel slip. The experiments were conducted with propane in the temperature interval 700e850 C. The propane conversion and slip are shown in Fig. 6. From these results it was concluded that the washcoat was not homogeneously distributed inside the micro reactor. There are possibly channels that are uncoated and others channels with higher washcoat loading, resulting in a relatively high fuel slip.
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
18041
Fig. 4 e The welded foils for the micro reactor prior to catalyst coating and mounting in the outer shell. The outer shell includes the fuel inlet/air inlet 1 and the outlet. The next step was to investigate the fuel reforming reaction in the bonded micro reactor and compare that to the monolith reactor. The fuel conversions for the two reactors are compared in Fig. 7. The conversions for the reactors look different, the monolith reactors leveling out around 750 C meanwhile the micro reactor seems to have a sharp increase in conversion around 800 C. Due to limitations in the external heating system for the micro reactor it was not possible to go higher in temperature to observe the behavior. The conversion is lower for the micro reactor, but this originates from the high fuel slip due to uneven coating. The diesel slip for the micro reactor is considerably higher than the one for the monolith reactor. This is most likely a consequence of uneven coating of the catalyst material inside the micro reactor. As a consequence of the high diesel slip for the micro reactor the FTIR measurements are less accurate. This is a problem originating from the FTIR instrument itself due to interference between diesel and the other analyzed species. Even though the micro reactor has approximately 85% fuel conversion at 800 C, the H2 yield at this temperature is close to the H2 yield of the monolith reactor at the same temperature (Fig. 8). At this temperature the monolith reactor has reached its maximum fuel conversion (95%) at current operating conditions. The H2 yield reached in the micro reactor is high even though a high diesel slip is observed. This is believed to arise due to higher steam reforming activity for the micro reactor compared to the monolith reactor. Regarding the CO2 selectivity there are some interesting differences between the two reactor designs (Fig. 9). In the micro reactor, the first CO is observed above 775 C, which would indicate that the CO formed from steam reforming and partial oxidation is oxidized to CO2. This is an indication of
Table 2 e Geometric data of the micro reactor. Number of foils Single foil length Reactor channel height Reactor channel width Reactor channels per foil Volume, reactor channel (without catalyst coating)
8 160 mm 0.5 mm 0.5 mm 27 8.64 cm3
enhanced WGS activity, which also would explain the high H2 yield, even though a high diesel slip is observed. The results might be an effect of the enhanced heat dissipation, avoiding elevated temperatures, which is beneficial for WGS activity. The heat dissipation of metallic micro reactors running reforming reactions and the caused isothermal temperature profile is described in previous works [11,17]. Due to the large difference in diesel slip between the reactor designs, the selectivity of CH4 is not straightforward to compare (Fig. 10). However, the pattern for the CH4 selectivity of the two reactor designs deviates. It can be observed that the selectivity for CH4 formation in the monolith reactor increases linearly with temperature. Meanwhile, the micro reactor shows a declining CH4 selectivity with temperature, leveling out at around 1%. Therefore the reaction path for CH4 formation in the two reactor designs can be assumed to be different. Since ethene is a known coke precursor it is essential to find a catalyst with low selectivity for the compound. Additionally it would be beneficial to have a reactor design lowering the ethene selectivity. As seen in Fig. 11 where the selectivity for ethene is shown, the micro reactor looks promising. In Fig. 12 the fuel conversion (a), diesel slip (b), H2 yield (c), and CO2 selectivity (d) are presented as a consequence of dosing air through the multiple air inlets. From these graphs it is clear that there is limited gain in dosing air along the reactor channels. The fresh air reacts rapidly with the produced H2 instead of hydrocarbons present in the reactant mixture. As a consequence of this the H2 yield is decreased as well as the fuel conversion. However, it cannot be ruled out that introduction of air at a later stage in a full-size OSR reactor without external heating would be beneficial. Since the O2, even though it is oxidizing the H2, would increase the temperature along the bed and even out the temperature gradient. A higher temperature in the endothermic part of the reactor would facilitate a faster steam reforming. Not only the freshly introduced air rapidly oxidizes the H2 but also the CO is oxidized. This is concluded from the increased CO2 selectivity with a higher ratio of air injected later along the reactor channels. During the experiments with the micro reactor an increased pressure was noted upstream the reactor. The pressure drop over the micro reactor was calculated to 1.1 bar
18042
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 5 e CFD simulations of O2 concentrations (vol.%) in the channels of the micro reactor (see Fig. 3 for a schematic picture of the reactor channels) a) O2 distribution when air is dosed through air inlet 2, b) O2 distribution when air is dosed in equal amounts through air inlet 2e4.
at 800 C and GHSV ~20,000 h1. This value is much higher than the estimated pressure drop of the monolith reactor [13]. The high pressure drop is a consequence of longer channels and the thick washcoat layer inside the micro reactor. Both these parameters need to be adapted for further development of the micro reactor. Even though the pressure drop was high for the micro reactor it did not increase with time. This, in conjunction with a fulfilled carbon balance, indicates that the observed phenomena of lower conversion in the bonded micro reactor compared to the monolith cannot be attributed to coke formation under low concentrations of O2.
Conclusions The micro reactor shows a high H2 yield despite low fuel conversion, which is assumed to be accomplished by an increased temperature in the steam reforming part.
There are no obvious advantages with dosing air at multiple inlets for a bench scale reactor with external heating. Since diluted reactant flow is employed in the micro reactor experiments, the temperature effects of the dosing air is not straightforward to measure. Additionally, the effect is that the produced H2 is oxidized by the freshly, introduced oxygen. However it cannot be concluded that multiple air feed in a fullscale reactor, without external heating, is not beneficial. When no external heating is supplied, the oxidation of hydrogen potentially generates heat, which increases the rate of the steam reforming reaction. Last but not least it remains that thermal stress of the catalyst under severe conditions (high GHSV) can cause severe deactivation [18] which might be avoided by multiple air dosing in the micro reactor. The largest disadvantage with the current design of the micro reactor is the uneven distribution of washcoat in the reactor channels, causing a high diesel slip. This could be avoided by changing the sequence of coating and bonding of
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 6 e Propane conversion and propane slip to estimate the homogeneity of the washcoat inside the micro reactor.
Fig. 7 e Comparison of the fuel conversion for the monolith reactor and the micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
Fig. 8 e Comparison of H2 yield for the monolith reactor and micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
18043
Fig. 9 e Comparison of the CO2 selectivity for the monolith reactor and the micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
Fig. 10 e Comparison of the CH4 selectivity for the monolith and micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
Fig. 11 e Comparison of the ethene selectivity of the monolith and micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
18044
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 12 e a) Fuel conversion, b) diesel slip, c) H2 yield and d) CO2 selectivity when dosing air at different ratios along the micro reactor channels. Reaction conditions: GHSV ~40,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The ratio between the inlets are denoted as the percentage of total air, inlet 1/inlet 2/inlet 3/inlet 4. The different air distributions are denoted as: (-) 100/0/0/0, ( ) 25/25/25/25 and ( ) 70/10/10/10.
the micro reactor. If the foils were coated before welded together, the distribution of washcoat would be significantly enhanced. Another large disadvantage with the current micro reactor design is the high pressure drop. This is especially valid for vehicle applications where everything is a compromise between advantages and disadvantages. Hence, an increased pressure drop requires a higher initial pressure. Extra pressure somewhere on the vehicles means a decrease in fuel-to-wheel efficiency and an increased fuel penalty. But the advantages with using a micro reactor with better heat integration and therefore a lower O2/C ratio could be employed, which means a higher H2 yield and lower fuel consumption. A micro reactor can also be designed in a more advanced way than a monolith reactor, which is another advantage. For large scale production (>10,000 units annually) of 500 Wel FC APUs, the production cost for the fuel reformer is estimated to approximately 60 V [19]. This cost would thereby be insignificant compared to the production cost for the fuel cell. Due to the previously addressed problems with the current micro reactor design, the focus in generation two will be focused on increased distribution of the washcoat inside the reactor channels. To solve the issues connected to the high pressure drop further work will be executed on the channel length and channel height, without compromising the enhanced mass and heat transfer of the micro reactor.
Acknowledgments The financial support from KICInnoEnergy in the project SynCon is gratefully acknowledged. The Swedish Foundation for Strategic Environmental Research (Mistra) is also acknowledged for the financial support. Thanks also to MEL Chemicals, Neste Oil, and Corning Inc. for the supply of the CeO2eZrO2 support, NExBTL fuel and cordierite monoliths, respectively.
references
[1] Huai T, Shah SD, Miller JW, Younglove T, Chernich DJ, Ayala A. Analysis of heavy-truck diesel activity and emission data. Atmos Environ 2006;40:2333e44. [2] Brodrick C-J, Lipman TE, Farshchi M, Lutsey NP, Dwyer HA, Sperling D, et al. Evaluation of fuel cell auxiliary power units for heavy-duty diesel trucks. Transp Res Part Transp Environ 2002;7:303e15. [3] Diesel Net, http://www.dieselnet.com, [last visited: 15-042014]. [4] Lutsey N, Brodrick C-J, Lipman T. Analysis of potential fuel consumption and emissions reductions from fuel cell auxiliary power units (APUs) in long-haul trucks. Energy 2007;32:2428e38. [5] Haynes DJ, Shekhawat D. Oxidative steam reforming. In: Shekhawat D, Spivey JJ, Berry D, editors. Fuel cells:
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
[6] [7] [8]
[9]
[10]
[11]
technologies for fuel processing. Amsterdam: Elsevier Science; 2011. Holladay JD, Hu J, King DL, Wang Y. An overview of hydrogen production technologies. Catal Today 2009;139:244e60. Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. Hoboken, N.J: Wiley; 2006. Beretta A, Donazzi A, Livio D, Maestri M, Groppi G, Tronconi E, et al. Optimal design of a CH4 CPO-reformer with honeycomb catalyst: combined effect of catalyst load and channel size on the surface temperature profile. Catal Today 2011;171:79e83. Thormann J, Pfeifer P, Schubert K, Kunz U. Reforming of diesel fuel in a micro reactor for APU systems. Chem Eng J 2008;135:S74e81. O'Connell M, Kolb G, Schelhaas KP, Schuerer J, Tiemann D, Ziogas A, et al. Development and evaluation of a microreactor for the reforming of diesel fuel in the kW range. Int J Hydrogen Energy 2009;34:6290e303. € rke O, Schubert K. Aartun I, Venvik HJ, Holmen A, Pfeifer P, Go Temperature profiles and residence time effects during catalytic partial oxidation and oxidative steam reforming of propane in metallic microchannel reactors. Catal Today 2005;110:98e107.
18045
[12] Thormann J, Pfeifer P, Kunz U. Dynamic performance of hexadecane steam reforming in a microstructured reactor. Chem Eng J 2012;191:410e5. [13] Heck RM, Farrauto RJ, Gulati ST. Catalytic air pollution control: commercial technology. 3rd ed. Hoboken, N.J: Wiley; 2009. [14] Neste Oil, http://www.nesteoil.com, [last visited: 24-02-2014]. [15] Granlund MZ, Jansson K, Nilsson M, Dawody J, Pettersson LJ. Evaluation of Co, La, and Mn promoted Rh catalysts for autothermal reforming of commercial diesel. Appl Catal B 2014;154e155:386e94. [16] Institute for Micro Process Engineering. Karlsruhe Institute of Technology, http://www.imvt.kit.edu, [last visited: 02-022014]. € rke O, Pfeifer P, Schubert K. Highly selective methanation [17] Go by the use of a microchannel reactor. Catal Today 2005;110:132e9. [18] Gunnarsson F, Granlund MZ, Englund M, Dawody J, Pettersson € relind H. Combining HC-SCR over Ag/Al2O3 and LJ, Ha hydrogen generation over Rh/CeO2-ZrO2 using biofuels: an integrated system approach for real applications. Submitted. € rke O, Pfeifer P. Abschlußbericht BMBF-Projekt [19] Go “Micromotive”. Karlsruhe; 2005.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Comparison between a micro reactor with multiple air inlets and a monolith reactor for oxidative steam reforming of diesel Moa Z. Granlund a,*, Oliver Go€rke b, Peter Pfeifer b, Lars J. Pettersson a a
KTH Royal Institute of Technology, Department of Chemical Engineering and Technology, SE-100 44 Stockholm, Sweden b Karlsruhe Institute of Technology, Institute for Micro Process Engineering (IMVT), D-76344 Eggenstein-Leopoldshafen, Germany
article info
abstract
Article history:
In order to lower the emission from idling heavy-duty trucks auxiliary power units can be
Available online 11 July 2014
implemented. Due to limited space available on-board the truck the units needs to be both efficient and compact. One alternative for these units is a fuel cell supplied with hydrogen
Keywords:
from a fuel reformer. Today, mostly monolithic reactors are used in the field of oxidative
Oxidative steam reforming
steam reforming of fuels, which has some challenges that need to be addressed before a
Diesel
possible breakthrough occurs on the market. One is the temperature gradient developed
Micro reactor
over the length of the monolith as a consequence of the sequential reactions. This could be
Monolith reactor
improved by using a metallic micro reactor with better heat integration between the re-
Multiple air inlets
action zones and further improving the integration with multiple air inlets along the catalytic bed. The aim with this study was to compare a conventional monolith reactor for oxidative steam reforming of fuel with a novel micro reactor design where air was dosed at four different positions along the reactor channels. The experiments were not necessarily conducted autothermal, i.e. a heating jacket was applied for operation. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction It is estimated that heavy-duty vehicles are responsible for about 30% of the total emissions of nitrogen oxides (NOx) and particulate matters (PM) [1]. In addition the engine is idling 30e40% of the time just to supply the vehicle with electricity [1,2]. The emission during idling is a considerable part of the total emission for the vehicle and there is an urgent need to
decrease those in order to meet future emission legislations [3]. One alternative to solve this could be to implement a fuel cell system where the hydrogen is produced via on-board reforming of diesel, a diesel fuel cell auxiliary power unit (FC APU) [4]. There are three different processes for reforming diesel to hydrogen-rich gas: partial oxidation, steam reforming and oxidative steam reforming (OSR); the latter could be operated self-sustaining as autothermal reforming (ATR). The three
* Corresponding author. Tel.: þ46 8 790 91 50. E-mail address: [email protected] (M.Z. Granlund). http://dx.doi.org/10.1016/j.ijhydene.2014.06.096 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
18038
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
processes have different advantages and disadvantages and for on-board fuel reforming OSR is the most suitable option [5]. The main reason is that the process is autothermal and therefore no external heating is needed. The autothermal characteristic is a consequence of the process being a combination of the exothermic partial oxidation and endothermic steam reforming. Other advantages are the short start-up time due to the fast thermal response and the relatively high hydrogen yield [6]. Due to the limitations of space onboard heavy-duty vehicles the design of the FC APU must be compact. An OSR reactor is normally based on a tubular reactor where the catalytic material is coated on a honeycomb cordierite monolith. This minimizes the pressure drop and also makes the reformer resistant to vibrations [7]. One of the most prominent disadvantages with this reactor design is the characteristic temperature gradient, developing along the monolith as a consequence of the combination of the two sequential reactions. The fast and exothermic partial oxidation reaction takes place at the entrance, causing a sharp increase in temperature and the slower, endothermic steam reforming reaction consumes the heat towards the outlet of reactor [8]. The heat transport from the initial partial oxidation reaction is transferred along the monolith only by convection and the interaction between the two reaction zones is poor. The initial raise in temperature can cause thermal deactivation of the catalytic material by the potential formation of hot spots. In order to limit the temperature increase in the beginning, better heat integration between the two reaction zones is necessary. One option to decrease the temperature gradient over the reactor and to increase the heat and mass transfer could be to use a micro reactor with better heat integration [9,10]. With a micro reactor it is possible to make a design closer connecting the exothermic and the endothermic part of the reactor. Another advantage with micro reactors is that metal is used which increases the heat conductivity of the reactor structure compared to ceramic monoliths [11]. One completely different possibility, already successfully tested at the Institute for Micro Process Engineering at Karlsruhe Institute of Technology, is to indirectly couple endothermic steam reforming with the exothermic combustion of anode-off gas in two different foil arrangements inside a metallic micro reactor [12]. In the current publication, however, the attempt was to further improve the heat integration between the two reaction zones in a direct coupling. Therefore, a micro reactor with multiple air dosing points axially distributed along the reaction channels was designed. The idea is to further improve the heat management by dosing the air axially along the reactor channels by adapting the kinetics between the partial oxidation and steam reforming reactions. With a better heat integration between the partial oxidation and steam reforming parts of the catalyst the temperature stress that usually occurs in the beginning of a monolith would be decreased. The integration could potentially also lead to higher temperatures in the steam reforming part, increasing the rate of the steam reforming reaction with a higher H2 yield as result. The scope of our work is to evaluate if micro structured reactors are a viable alternative to monolith reactors for on-
board production of hydrogen. The deactivation of the catalytic material caused by thermal aging, i.e. sintering and collapsing of the pore structure, might therefore be decreased as previously shown in Ref. [11].
Experimental Catalyst preparation The catalytic material used in the comparative study between the two different reactor designs was 3 wt.% Rh supported on high surface area CeO2eZrO2 (MEL Chemicals, 16.5% CeO2/ ZrO2) prepared by the incipient wetness method. As metal precursor, Rh nitrate solution was used (Rh(NO3)3 solution, 9.53 w/w, SigmaeAldrich). After the impregnation the catalytic material was calcined in air at 800 C for 3 h. Subsequently, the catalytic material was suspended in ethanol and ball milled for 24 h. For the monolith reactor the catalytic material was applied to a cordierite monolith (400 cpsi) by dip coating until the washcoat loading was approximately 20 wt.%. After the coating the monolith was calcined in air at 800 C for 3 h. The inserted catalyst mass was 1.15 g. The diameter and length of the monolith was 20 mm and 30 mm, respectively, resulting in an inner reaction volume of approximately 5.35 cm3 [13]. Before coating the micro structured reactor, coating tests with single metallic micro structured foils were conducted. Prior to coating the Nicrofer® foils were calcined at 800 C for 3 h in air to generate a thin oxide film on the surface. Subsequently the foils were coated dropwise with the catalyst slurry. The coated foils were then investigated with scanning electron microscope and microprobe analysis. The obtained coating showed an excellent adhesion on the micro channel foils and the catalytic material was homogeneously distributed on the foils, which can be seen in Fig. 1. For the micro reactor with multiple air inlets the micro structured foil stack was welded by diffusion bonding and air inlets were attached to the flow chambers by tungsten inertgas welding. The stack was calcined at 800 C for 3 h and afterwards flow-coated with the same slurry as it was used for dipcoating of the single micro structured foils. After drying and calcination of the catalyst for 3 h at 800 C in air, the flow chambers were welded onto the coated stack on the reaction side by tungsten inert gas welding. The inserted catalyst mass was also 1.15 g. The fuel used in the study was synthetic diesel of type biomass-to-liquid (NExBTL, Neste Oil). The main advantage with using synthetic diesel is its high content of alkanes, implying low content of aromatic and polyaromatic hydrocarbons that are prone to coke and lead to catalyst deactivation. Another benefit with synthetic diesel is its low sulfur content [14]. The characteristics of the synthetic diesel are summarized in Table 1.
Reactor set-up The OSR experiments were performed in a bench-scale setup where either the micro reactor or the monolith reactor was
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
18039
Fig. 1 e Scanning electron microscopy picture of a micro structured Nicrofer® foil coated with 3 wt.% Rh/CeO2eZrO2.
mounted. For the monolith reactor experiments a horizontally mounted high-temperature stainless steel (Sandvik 253 MA, Øinner ¼ 24.3 mm) reactor was used. This reactor configuration is described in detail elsewhere [15]. The integration of the micro reactor in the reactor setup can be seen in Fig. 2. The temperature measured on the outside of the micro reactor was assumed to be the reactor temperature, since the temperature gradient was negligible. Due to limitations of the external heating system of the micro reactor the maximum operating temperature was 800 C. The reformate was analyzed with an FTIR instrument, a mass spectrometer (V&F Inc., H-sense) and a gas chromatograph (Varian CP-3800). The FTIR instrument (MKS Multigas™ 230 HS) was equipped with a high speed spectrometer and was used to analyze the concentrations of H2O, CO, CO2, CH4, short-chain hydrocarbons and unconverted fuel online in the wet reformate. The mass spectrometer monitored the H2 concentrations in the dry reformate. The gas chromatograph was equipped with two sequentially packed columns, a Porapak followed by a molecular sieve (5 Å), and a thermal conductivity detector used to analyze the concentrations of H2, CO, CO2, O2 and N2 in the dry reformate. Prior to the experiments in the monolith reactor the catalyst was pre-treated in N2 at 950 C for 30 min to accelerate the thermal aging of the material. The reactor design evaluation experiments were performed in the temperature range of 700e900 C with steady state measurements at 50 C intervals. The experiments were performed at a gas hourly space velocity (GHSV) of 50,000 h1.
Table 1 e Characteristics of NExBTL [14]. Properties Sulfur content Min. boiling point Dist. 95% evap. Aromaticsa Carbon, C Hydrogen, H a
Value
Unit
<1 180 293 <0.3 84.9 15.1
ppm C C wt.% wt.% wt.%
Includes mono-, di-, and tri þ - aromatics.
To evaluate the heat integration capacity of the micro reactor, 100% of the air feed was dosed through the first air inlet (see Fig. 3). The influence of dosing air along the reactor channels, through the multiple air inlets (see Fig. 3) of the micro reactor was evaluated at 750 C and GHSV ~40,000 h1. During all experiments, both monolithic reactor and micro reactor, the operation parameters H2O/C and O2/C were kept constant at 2.5 and 0.45, respectively.
Micro reactor design The metallic micro reactor with multiple air inlets (fuel inlet/air inlet 1 þ 3 additional air inlets distributed axially along the reactor channels) was manufactured at the Institute for Micro Process Engineering (IMVT) at Karlsruhe Institute of Technology (KIT) [16]. The idea with the design is to combine three strategies into one micro reactor design, all to minimize the typical temperature gradient developing axially in the reactor channels of a OSR reactor. The three strategies are: Enhancing the heat transfer, both in axial and radial direction, by manufacturing the reactor in metal. Further improving the heat transfer between the exothermic partial oxidation and the endothermic steam reforming by dosing them in opposite directions. Dosing the air at multiple places axially along the reactor channels to distribute the partial oxidation of diesel. The micro reactor consists of multiple foils (each containing 27 reaction channels) welded together by diffusion bonding and mounted in the outer shell. A picture of the bonded foils is shown in Fig. 4. The fuel inlet/air inlet 1 and the outlet are part of the outer shell and are not included in the picture. The air fed in air inlet 1,2 and 3 is introduced in each reactor channel via holes positioned: 5 mm, 40 and 80 mm, downstream the fuel inlet/air inlet 1. After air inlet 3, the gas flow is redirected 180 and returned towards the outlet. A schematic picture of one reactor channel is shown in Fig. 4. The geometric data of the micro reactor is summarized in Table 2. When the reactor channels are coated with catalyst, both the height and width are reduced by approximately 150 mm (75 mm coating thickness).
18040
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 2 e Schematic picture of the experimental setup with the micro reactor installed. A detailed experimental setup with the monolith reactor installed can be found in Ref. [15].
Product analysis and calculations SHx Cy ¼ In the calculations the diesel is approximated to C14H26. Fuel conversion is defined as the amount of fuel converted to CO, CO2 or CH4 according to Eq. (1), where Fi denotes the respective molar flows. Xfuel ð%Þ ¼
FCO þ FCO2 þ FCH4 $100 14 Ffuel
(1)
The H2 yield is defined as the percentage of the theoretical, maximum amount H2 formed when assuming the only products are CO2 and H2, i.e. all CO is converted to CO2 through the water-gas shift reaction (WGS): YH2 ð%Þ ¼
FH2 þ FCO $100 FH2 max
(2)
The selectivity for CO2 compared to CO is calculated according to Eq. (3). SCO2 ð%Þ ¼
FCO2 $100 FCO2 þ FCO
The selectivity of hydrocarbon is shown in Eq. (4).
Fig. 3 e Schematic picture of one reactor channel in the micro reactor design.
(3)
y$FHx Cy FC1 ;in
(4)
Results and discussion The design of the micro reactor is based on results from previous experiments with ceramic monolith reactors. The designing of the micro reactor was performed in multiple steps. The first step was to ensure that the total flow is equally distributed even in the case when air dosing is performed downstream the reactor channels. The results from the CFD simulations can be seen in Fig. 5 where the concentration of O2 is correlated to the color. In Fig. 5a the flow is simulated when air is dosed through air inlet 2 while in Fig. 6b the air is dosed, equally divided, through air inlets 2e4. From the CFD simulations it was concluded that the flow distribution was unaffected by the introduction of additional air at a later stage in the reactor channels. The second step was to evaluate the design concept of the foils that were going to be used in the micro reactor. This was executed in a clamped reactor setup, where the foils were clamped together after they had been coated with catalyst (3 wt.% Rh supported on CeO2eZrO2). The step enabled evaluation of the substrate's effect on the catalyst and also to evaluate the coating technique. The clamped reactor was evaluated through OSR of propane. The results from the clamped reactor design showed that the catalyst was not affected by the change in substrate. The next step was to evaluate the distribution of catalytic washcoat inside the bonded micro reactor. An uneven coating would mean that some channels are not coated and no catalytic activity appears here. This also influences the pressure drop, which decreases in the uncoated channels, causing some diesel slip. The experiments were conducted with propane in the temperature interval 700e850 C. The propane conversion and slip are shown in Fig. 6. From these results it was concluded that the washcoat was not homogeneously distributed inside the micro reactor. There are possibly channels that are uncoated and others channels with higher washcoat loading, resulting in a relatively high fuel slip.
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
18041
Fig. 4 e The welded foils for the micro reactor prior to catalyst coating and mounting in the outer shell. The outer shell includes the fuel inlet/air inlet 1 and the outlet. The next step was to investigate the fuel reforming reaction in the bonded micro reactor and compare that to the monolith reactor. The fuel conversions for the two reactors are compared in Fig. 7. The conversions for the reactors look different, the monolith reactors leveling out around 750 C meanwhile the micro reactor seems to have a sharp increase in conversion around 800 C. Due to limitations in the external heating system for the micro reactor it was not possible to go higher in temperature to observe the behavior. The conversion is lower for the micro reactor, but this originates from the high fuel slip due to uneven coating. The diesel slip for the micro reactor is considerably higher than the one for the monolith reactor. This is most likely a consequence of uneven coating of the catalyst material inside the micro reactor. As a consequence of the high diesel slip for the micro reactor the FTIR measurements are less accurate. This is a problem originating from the FTIR instrument itself due to interference between diesel and the other analyzed species. Even though the micro reactor has approximately 85% fuel conversion at 800 C, the H2 yield at this temperature is close to the H2 yield of the monolith reactor at the same temperature (Fig. 8). At this temperature the monolith reactor has reached its maximum fuel conversion (95%) at current operating conditions. The H2 yield reached in the micro reactor is high even though a high diesel slip is observed. This is believed to arise due to higher steam reforming activity for the micro reactor compared to the monolith reactor. Regarding the CO2 selectivity there are some interesting differences between the two reactor designs (Fig. 9). In the micro reactor, the first CO is observed above 775 C, which would indicate that the CO formed from steam reforming and partial oxidation is oxidized to CO2. This is an indication of
Table 2 e Geometric data of the micro reactor. Number of foils Single foil length Reactor channel height Reactor channel width Reactor channels per foil Volume, reactor channel (without catalyst coating)
8 160 mm 0.5 mm 0.5 mm 27 8.64 cm3
enhanced WGS activity, which also would explain the high H2 yield, even though a high diesel slip is observed. The results might be an effect of the enhanced heat dissipation, avoiding elevated temperatures, which is beneficial for WGS activity. The heat dissipation of metallic micro reactors running reforming reactions and the caused isothermal temperature profile is described in previous works [11,17]. Due to the large difference in diesel slip between the reactor designs, the selectivity of CH4 is not straightforward to compare (Fig. 10). However, the pattern for the CH4 selectivity of the two reactor designs deviates. It can be observed that the selectivity for CH4 formation in the monolith reactor increases linearly with temperature. Meanwhile, the micro reactor shows a declining CH4 selectivity with temperature, leveling out at around 1%. Therefore the reaction path for CH4 formation in the two reactor designs can be assumed to be different. Since ethene is a known coke precursor it is essential to find a catalyst with low selectivity for the compound. Additionally it would be beneficial to have a reactor design lowering the ethene selectivity. As seen in Fig. 11 where the selectivity for ethene is shown, the micro reactor looks promising. In Fig. 12 the fuel conversion (a), diesel slip (b), H2 yield (c), and CO2 selectivity (d) are presented as a consequence of dosing air through the multiple air inlets. From these graphs it is clear that there is limited gain in dosing air along the reactor channels. The fresh air reacts rapidly with the produced H2 instead of hydrocarbons present in the reactant mixture. As a consequence of this the H2 yield is decreased as well as the fuel conversion. However, it cannot be ruled out that introduction of air at a later stage in a full-size OSR reactor without external heating would be beneficial. Since the O2, even though it is oxidizing the H2, would increase the temperature along the bed and even out the temperature gradient. A higher temperature in the endothermic part of the reactor would facilitate a faster steam reforming. Not only the freshly introduced air rapidly oxidizes the H2 but also the CO is oxidized. This is concluded from the increased CO2 selectivity with a higher ratio of air injected later along the reactor channels. During the experiments with the micro reactor an increased pressure was noted upstream the reactor. The pressure drop over the micro reactor was calculated to 1.1 bar
18042
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 5 e CFD simulations of O2 concentrations (vol.%) in the channels of the micro reactor (see Fig. 3 for a schematic picture of the reactor channels) a) O2 distribution when air is dosed through air inlet 2, b) O2 distribution when air is dosed in equal amounts through air inlet 2e4.
at 800 C and GHSV ~20,000 h1. This value is much higher than the estimated pressure drop of the monolith reactor [13]. The high pressure drop is a consequence of longer channels and the thick washcoat layer inside the micro reactor. Both these parameters need to be adapted for further development of the micro reactor. Even though the pressure drop was high for the micro reactor it did not increase with time. This, in conjunction with a fulfilled carbon balance, indicates that the observed phenomena of lower conversion in the bonded micro reactor compared to the monolith cannot be attributed to coke formation under low concentrations of O2.
Conclusions The micro reactor shows a high H2 yield despite low fuel conversion, which is assumed to be accomplished by an increased temperature in the steam reforming part.
There are no obvious advantages with dosing air at multiple inlets for a bench scale reactor with external heating. Since diluted reactant flow is employed in the micro reactor experiments, the temperature effects of the dosing air is not straightforward to measure. Additionally, the effect is that the produced H2 is oxidized by the freshly, introduced oxygen. However it cannot be concluded that multiple air feed in a fullscale reactor, without external heating, is not beneficial. When no external heating is supplied, the oxidation of hydrogen potentially generates heat, which increases the rate of the steam reforming reaction. Last but not least it remains that thermal stress of the catalyst under severe conditions (high GHSV) can cause severe deactivation [18] which might be avoided by multiple air dosing in the micro reactor. The largest disadvantage with the current design of the micro reactor is the uneven distribution of washcoat in the reactor channels, causing a high diesel slip. This could be avoided by changing the sequence of coating and bonding of
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 6 e Propane conversion and propane slip to estimate the homogeneity of the washcoat inside the micro reactor.
Fig. 7 e Comparison of the fuel conversion for the monolith reactor and the micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
Fig. 8 e Comparison of H2 yield for the monolith reactor and micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
18043
Fig. 9 e Comparison of the CO2 selectivity for the monolith reactor and the micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
Fig. 10 e Comparison of the CH4 selectivity for the monolith and micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
Fig. 11 e Comparison of the ethene selectivity of the monolith and micro reactor at GHSV ~50,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The reactor designs are denoted as: (-) monolith reactor and ( ) micro reactor.
18044
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
Fig. 12 e a) Fuel conversion, b) diesel slip, c) H2 yield and d) CO2 selectivity when dosing air at different ratios along the micro reactor channels. Reaction conditions: GHSV ~40,000 h¡1, O2/C ~0.45 and H2O/C ~2.5 with NExBTL. The ratio between the inlets are denoted as the percentage of total air, inlet 1/inlet 2/inlet 3/inlet 4. The different air distributions are denoted as: (-) 100/0/0/0, ( ) 25/25/25/25 and ( ) 70/10/10/10.
the micro reactor. If the foils were coated before welded together, the distribution of washcoat would be significantly enhanced. Another large disadvantage with the current micro reactor design is the high pressure drop. This is especially valid for vehicle applications where everything is a compromise between advantages and disadvantages. Hence, an increased pressure drop requires a higher initial pressure. Extra pressure somewhere on the vehicles means a decrease in fuel-to-wheel efficiency and an increased fuel penalty. But the advantages with using a micro reactor with better heat integration and therefore a lower O2/C ratio could be employed, which means a higher H2 yield and lower fuel consumption. A micro reactor can also be designed in a more advanced way than a monolith reactor, which is another advantage. For large scale production (>10,000 units annually) of 500 Wel FC APUs, the production cost for the fuel reformer is estimated to approximately 60 V [19]. This cost would thereby be insignificant compared to the production cost for the fuel cell. Due to the previously addressed problems with the current micro reactor design, the focus in generation two will be focused on increased distribution of the washcoat inside the reactor channels. To solve the issues connected to the high pressure drop further work will be executed on the channel length and channel height, without compromising the enhanced mass and heat transfer of the micro reactor.
Acknowledgments The financial support from KICInnoEnergy in the project SynCon is gratefully acknowledged. The Swedish Foundation for Strategic Environmental Research (Mistra) is also acknowledged for the financial support. Thanks also to MEL Chemicals, Neste Oil, and Corning Inc. for the supply of the CeO2eZrO2 support, NExBTL fuel and cordierite monoliths, respectively.
references
[1] Huai T, Shah SD, Miller JW, Younglove T, Chernich DJ, Ayala A. Analysis of heavy-truck diesel activity and emission data. Atmos Environ 2006;40:2333e44. [2] Brodrick C-J, Lipman TE, Farshchi M, Lutsey NP, Dwyer HA, Sperling D, et al. Evaluation of fuel cell auxiliary power units for heavy-duty diesel trucks. Transp Res Part Transp Environ 2002;7:303e15. [3] Diesel Net, http://www.dieselnet.com, [last visited: 15-042014]. [4] Lutsey N, Brodrick C-J, Lipman T. Analysis of potential fuel consumption and emissions reductions from fuel cell auxiliary power units (APUs) in long-haul trucks. Energy 2007;32:2428e38. [5] Haynes DJ, Shekhawat D. Oxidative steam reforming. In: Shekhawat D, Spivey JJ, Berry D, editors. Fuel cells:
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 3 9 ( 2 0 1 4 ) 1 8 0 3 7 e1 8 0 4 5
[6] [7] [8]
[9]
[10]
[11]
technologies for fuel processing. Amsterdam: Elsevier Science; 2011. Holladay JD, Hu J, King DL, Wang Y. An overview of hydrogen production technologies. Catal Today 2009;139:244e60. Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. Hoboken, N.J: Wiley; 2006. Beretta A, Donazzi A, Livio D, Maestri M, Groppi G, Tronconi E, et al. Optimal design of a CH4 CPO-reformer with honeycomb catalyst: combined effect of catalyst load and channel size on the surface temperature profile. Catal Today 2011;171:79e83. Thormann J, Pfeifer P, Schubert K, Kunz U. Reforming of diesel fuel in a micro reactor for APU systems. Chem Eng J 2008;135:S74e81. O'Connell M, Kolb G, Schelhaas KP, Schuerer J, Tiemann D, Ziogas A, et al. Development and evaluation of a microreactor for the reforming of diesel fuel in the kW range. Int J Hydrogen Energy 2009;34:6290e303. € rke O, Schubert K. Aartun I, Venvik HJ, Holmen A, Pfeifer P, Go Temperature profiles and residence time effects during catalytic partial oxidation and oxidative steam reforming of propane in metallic microchannel reactors. Catal Today 2005;110:98e107.
18045
[12] Thormann J, Pfeifer P, Kunz U. Dynamic performance of hexadecane steam reforming in a microstructured reactor. Chem Eng J 2012;191:410e5. [13] Heck RM, Farrauto RJ, Gulati ST. Catalytic air pollution control: commercial technology. 3rd ed. Hoboken, N.J: Wiley; 2009. [14] Neste Oil, http://www.nesteoil.com, [last visited: 24-02-2014]. [15] Granlund MZ, Jansson K, Nilsson M, Dawody J, Pettersson LJ. Evaluation of Co, La, and Mn promoted Rh catalysts for autothermal reforming of commercial diesel. Appl Catal B 2014;154e155:386e94. [16] Institute for Micro Process Engineering. Karlsruhe Institute of Technology, http://www.imvt.kit.edu, [last visited: 02-022014]. € rke O, Pfeifer P, Schubert K. Highly selective methanation [17] Go by the use of a microchannel reactor. Catal Today 2005;110:132e9. [18] Gunnarsson F, Granlund MZ, Englund M, Dawody J, Pettersson € relind H. Combining HC-SCR over Ag/Al2O3 and LJ, Ha hydrogen generation over Rh/CeO2-ZrO2 using biofuels: an integrated system approach for real applications. Submitted. € rke O, Pfeifer P. Abschlußbericht BMBF-Projekt [19] Go “Micromotive”. Karlsruhe; 2005.