ENERGY INTEGRATION OF THE STEAM REFORMING PROCESS OF BIOETHANOL

Document Sample
ENERGY INTEGRATION OF THE STEAM REFORMING PROCESS OF BIOETHANOL Powered By Docstoc
					  ENERGY INTEGRATION OF THE STEAM REFORMING PROCESS
                              OF BIOETHANOL AND FUEL CELLS
                                 J.A. Francesconi, M.C. Mussati, R.O. Mato, P.A. Aguirre*.
                                         Instituto de Desarrollo y Diseño, INGAR (CONICET)


         Abstract. Fuel cells are being considered a promising alternative to the propulsors in transport vehicles in
         the next future since they operate at higher efficiencies and produce low environmental impact. The aim of
         this work is to investigate the synthesis and energy integration of a fuel processor for hydrogen production.
         Steam reforming of bioethanol and hydrogen purification process are coupled to a polymeric fuel cell
         (PEMFC). The hydrogen processors consist mainly of three fixed-bed catalytic reactors; the first is a fuel
         reformer and the other ones are used for reducing impurities, mainly CO, which affect the system operation
         and deteriorate the cell. A commercial simulator was used to solve the mass and energy balance, and to
         compute the operative conditions for the process units. The highest energy consumption is demanded by the
         reforming reactor, which operates at the highest thermal level. It is needed to preheat, evaporate and re-heat
         the feed stream, and to provide the reaction heat. To obtain an efficient integration, the heat exchanged
         between the reformer outcoming streams of higher thermal level (reforming and combustion gases) and the
         feed stream, should be maximized. The operation at high temperatures diminishes the processor efficiencies.
         Another process variable that affects the process efficiency is the water-to-fuel ratio fed to the reformer. The
         water excess must be evaporated and re-heated consuming additional fuel in the reformer, diminishing then
         the system efficiency.


      Keywords: bioethanol reforming; hydrogen processor; fuel cells; process integration.



1. Introduction

      Fuel cells have received increased attention in recent years for power generation owing to their potential
higher thermal efficiency and lower CO2 emissions per unit of power produced. Fuel cells are intrinsically much
more efficient and could achieve 70-80% system efficiency (Fuel Cell Handbook, 2004).
      Many research and development projects have been conducting in both the fuel cell itself and the fuel
processors for generating hydrogen. There exist several routes for hydrogen production from the primary fuels.
Ethanol presents several advantages related to natural availability, storage and handling safety, ethanol can be
produced renewably from several biomass sources. Besides the bioethanol-to-hydrogen system has the
significant advantage of being nearly CO2 neutral, since the produced carbon dioxide is consumed for biomass
growth, thus offering a nearly closed carbon loop.
      The aim of this work is to investigate the synthesis and energy integration of a processor of hydrogen
produced by steam reforming of bioethanol coupled to a polymeric fuel cell. Proton exchange membrane (PEM)
fuel cells for stationary applications are highly integrated systems including fuel processing, fuel cell itself and
post-combustion units (Godat and Marechal, 2003; Ratnamala et al.,2005). Process design has therefore a
large impact on the performances of the total system. Our goal here has been analyze and improve by
simulation techniques the performances of a fuel processor and a proton exchange membrane fuel cell (PEMFC).
As a starting point for the design, the Little (1994) hypothetical system has been used as a reference case, in
order to analyze the sensitivity of major decision parameters and to determine the major integration bottlenecks.



  *
   To whom all correspondence should be addressed.
  Address: Instituto de Desarrollo y Diseño - INGAR, Avellaneda 3657, (CP) S3000GJC, Santa Fe - Argentina
  E-mail: paguir@ceride.gov.ar
This step aims also at understanding the influence of constraints and computing their importance with respect to
the system performances.
   The system includes a steam bioethanol reforming (SR), a water gas shift (WGS) and a preferential
oxidation (PROX) reactors for the fuel processing feeding a proton exchange membrane cell stack (PEMFC) and
a post-combustion unit. The heat exchanger networks were modeled using the HYSYS operation LNG
exchanger model. This has been used to identify the best heat exchange opportunities and to define the optimal
operating conditions of the reforming system in order to provide the best overall efficiency considering the
balance of plant. The model developed will serve as a basis for further process design including the optimal
structure determination and the thermo-economic optimization.




                                     Fig. 1. Fuel processing process flow diagram.


2. The Bioethanol Processing System

   The fuel processing process has been built according to Little (1994). The following configuration can be
identified: The fuel processing system that includes, the steam bioethanol reformer (SR), the water gas shift
(WGS) reactors and the preferential oxidation reactor (PROX). The proton exchange membrane fuel cell
(PEMFC); and the post-combustion system.
   The model has been implemented using commercial modular software, HYSYS. As this simulation case is
developed for system design, the pressure drops have been neglected and the operating pressure has been fixed at
3 atm.
   The inlet flows presented in Fig. 1 are the following: The stream #1 is the bioethanol flow provided at 3 atm
and 25C. The stream #2 is the water flow required for the steam reformer whose flowrate is controlled by the
water to ethanol molar ratio. The stream #11 is a air flow required for the PROX. In the integrated system,
the air compression work will be produced by expanding the post combustion gases in an expander. The
compressor and expander isentropic efficiencies are 70%. The inlet air conditions are 1 atm, 25C.
2.1. Steam Reformer of Bioethanol
   Thermodynamic studies (García and Laborde,1991; Vasudeva et al. 1996) has shown that the steam
reforming of ethanol is feasible for temperature higher than 230 C, being methane, carbon oxides and H2 the
main products. The steam reforming of alcohols for hydrogen production involves a complex multiple reaction
system, the purity of a hydrogen product being affected by many undesirable side reactions. Therefore, the yield
of hydrogen depends in a complex manner on the process variables such as pressure, temperature, reactants ratio,
etc, and on the catalyst used.
   The steam bioethanol reformer performs the following reactions:
   The endothermic bioethanol decomposition reaction:
   C2H5OH ↔ CO + H2 + CH4,                 = 49.05 kJ/mol
   The endothermic steam methane reforming:
   CH4 + H2O ↔ 3H2 + CO,               = 206.11 kJ/mol
   The exothermic water gas shift (WGS):
   CO + H2O ↔ CO2 + H2,             = -41.1 kJ/mol
    The three reactions are supposed to be at equilibrium. The reactor is supposed to be isothermal, meaning
that heat has to be supplied to maintain the temperature in the reactor from an external energy source. In
order to define appropriately the energy requirement of the heat exchange, we have assumed that the inlet
streams will be preheated up to the reaction temperature. In this study, the temperature of the reformer and
the water to ethanol molar ratio (W/E) have been considered as being the major decision variables for
the analysis.




                Fig. 2. Output molar fraction for reforming unit as function of W/E molar ratio and temperature


Fig. 2 shows H2, CO, CO2, CH4 output molar fraction from the reformer unit as function of water/ethanol molar
ratio and temperature. H2 production is favored at low molar ratio and higher temperatures; however optimal
operating condition will be determined by a global efficiency analysis.
2.2. Water Gas Shift Reactors
   CO produced from reforming reactions must be brought down to the ppm level because it gets adsorbed on
the noble catalyst of the PEMFC and poisons it. The CO level is reduced by water gas shift reactors, these units
proceeds with the exothermal WGS reaction that is supposed to be at the equilibrium. This step also leads to
additional H2 generation. The water gas shift reaction is usually carried out in two adiabatic shift reactors in
series with an inter-cooler in between to remove the heat of reaction for the exothermic water gas shift reaction.
The fist-stage reactor typically operates at 350-450 ºC and is called the high-temperature shift (HTS) reactor.
The HTS reactor uses a chromiun-promoted iron oxide catalyst. The second stage is a low-temperature shift
(LTS) reactor which operates at 150-250 ºC, using a copper-zinc catalyst supported on alumina. The LTS is
capable of achieving a residual CO concentration on the order of 0.5-1.5 dry vol%. For integration with PEM
fuel cells, the reformate needs to be processed in additional process modules to reduce the CO content to ppm
level. We consider the WGS system configuration propose by Litle (1994), with two adiabatic reactors, HTS and
LTS, the input temperature of the first reactor has been fixed at 500 ºC, and 150 ºC for the second one.

2.3. Preferential Oxidation Reactor
   The preferential oxidation reactor is used to eliminate the CO that has not been converted in the WGS
reactors. This reactor is required to reach very low level of CO content in the fuel stream at the fuel cell inlet in
order to avoid poisoning of the membrane. The preferential oxidation is one of the possible technologies for this
task. It is using oxygen to proceed with the following reaction: CO + O2 → CO2 .
   Unfortunately, the selectivity of the catalyst will not avoid the combustion of some hydrogen in the gas
stream with the following reaction: H2 + ½ O2 → H2O. In the model, the CO content at the reactor outlet is
fixed at 20 ppm. The air flow rate (stream #11) is computed as a function of the CO flow rate assuming two mole
of O2 per mole of CO. An adiabatic operation at input temperature of 237 ºC has been considered for the PROX
reactor. The reactor inlet is supposed to be cooled down to the reactor temperature before entering the reactor.

2.4. Proton Exchange Membrane Fuel Cell
     The target power for the PEMFC unit was the 1 kW, the required amount of hydrogen was fixed at 33.3
gmol/h (Choi and Stenger,2003). Fuel utilization is considered at 80%.The PEMFC is supposed to be isothermal
and isobar. The cell temperature will be fixed at 80 C. Water management strategies have not been considered in
the model.
The overall efficiency of the system is computed by
               Pelec
η=                 FC       add                                                                                   (1)
     HHV fuel ( f fuel + f fuel )

where   Pelec is a generated electric power (1 kW), HHV fuel the higher heating value of fuel (1358.5 kJ/gmol),
    FC                                                                         add
 f fuel the fuel flow rate processed in the fuel processing system (#1) and f fuel is the additional flow rate required
in the post-combustion system


2.5. Post-Combustion System
   The depleted fuel of the PEMFC, formed by cathode and anode outlets, is burned in the post combustion
system. The generated heat will be used to balance the energy requirement of the fuel processing section.
Supplementary firing of ethanol will be considered if the energy content of the depleted fuel is not sufficient to
satisfy the balance. The supplementary fuel requirement will be computed to achieve a minimum approach
temperature of ∆T=100 ºC in the cold side of the reformer. Complete and stoechiometric combustion for the
burner unit has been assumed.
     The flue gases will exchange heat with the other units before being expanded in a turbine coupled to the air
compressor. The inlet temperature in the turbine will be computed to balance the air compression work
requirement. Additional ethanol burning will be considered, if is necessary, to satisfy this balance.


2.6. Heat Exchange Model
     Intentionally, the heat exchangers have not been explicitly considered in the simulation model described
above because the heat recovery system has been considered as being unknown apriori. We have
therefore used the LNG unit to model the integrated heat exchange system without having to impose a
heat exchange network structure. The LNG exchanger model is a HYSYS operation that solves heat and
material balances for multi-stream heat exchangers and heat exchanger networks. The LNG calculations are
based on energy balances for the hot and cold fluids. In this approach, the results of the simulation model will
characterize the hot and cold streams of the system. Table 1 summarized the thermal requirement for the fuel
processor streams connected to the LNG unit.


                                 Table 1. Fuel Processor Streams connected to the LNG unit.
N0.      Stream    Description                                                                Tin (ºC)   Tout (ºC)
 1        Cold     Water and ethanol preheating for the SR reaction                               80        Tref
 2        Cold     Water and ethanol preheating for the SR reaction                               25        Tref
 3         Hot     Stream between the SR and HTS                                                 Tref       500

 4         Hot     Stream between the HTS and LTS                                                 out
                                                                                                 THTS       150

 5         Hot     Stream between the HTS and PROX                                                 out
                                                                                                 TLTS       237

 6         Hot     Stream between the PROX and the PEMFC inlet                                    out
                                                                                                TPROX       80

 7        Cold     Anode and cathode gases from FC, preheating for the combustion                 80        500
 8        Cold     Air preheating for the combustion                                              out
                                                                                                TCOMP       300

 9         Hot     Combustion gases after SR to turbine                                           Tcg         in
                                                                                                            Ttur
 10       Cold     Extra ethanol preheating for the combustion                                    25        300


     The LNG unit permits analyze the system energy integration by process integration method (or pinch
technology) (Linnhoff et al., 1985). Process integration studies start with the definition of a list of hot and cold
streams. The hot and cold streams define, respectively, heat sources and heat requirements of the system that are
characterized by a heat-temperature diagram usually defined by a heat load, an inlet and a target temperature.
The heat sources are then composed to compute the hot composite curve that represents the heat availability in
the system as a function of the temperature. The same procedure is applied for the cold streams to draw the cold
composite curve. Considering that the heat exchange will be technically feasible if the temperature
difference between the hot and cold streams will always be superior to a predefine ∆Tmin , the maximum
heat recovery by heat exchange between the hot and the cold streams will be obtained when the ∆Tmin constraint
will be activated. This point is called the system pinch point. By heat balance, one may then compute the
minimum energy requirement of the system and the minimum heat losses to be evacuated from the system.
Using the composite curve calculation together with a simulation model, we define the simulation of the
heat exchanges network without knowing about its configuration. With this approach, we will be able to compute
the influence of the decision variables to later on define the best system configuration.

3. Results and Discussions

   The integrated process model has been used to determine the optimal operating conditions to be considered in
the system. The influence of the water/ethanol molar ratio and reforming temperature has been studied.

   Fixed the Water/Ethanol molar ratio and the reforming temperature ( Tref ), the simulation model compute:

the ethanol flow necessary to produce 33.3 gmol/h of H2 for the PEMFC inlet. This condition determines the

temperature of combustion gases ( Tcg ) subsequent to the energy exchange with the reformer unit. If this

temperature is lower than Tref + ∆T , an additional ethanol burning is computed until achieve this thermal

requirement. We considered ∆T =100º C. Flue gases will exchange heat with the other units before being
expanded in a turbine. If the flue gases temperature, after this heat exchange, is lower that the inlet temperature
                   in
in the turbine ( Ttur ) necessary to balance the air compression work, a supplementary firing of ethanol is

computed.

3.2. The Influence of the Water/Ethanol Molar Ratio




                    .
                        Fig. 3 The influence of the water/ethanol molar ratio over system efficiency
    Fig 3 shows efficiency for different water/ethanol molar ratio at Tref =750 ºC. Smaller molar ratios present

lower efficiency because the CO generation is superior, and the PROX unit consumes more H2 to reduce the CO
level to 20 ppm. This suggests the necessity to optimize the input temperatures in the WGS reactors to achieve a
better efficiency. At higher W/E molar ratio, the water excess must be evaporated and re-heated consuming
additional fuel in the reformer, diminishing then the system efficiency. From the figure we can conclude that the
optimal value for these reforming conditions is near a water/ethanol molar ratio of 5. This agrees with practical
use, where a water excess respect to the stoichiometric values prevents coke formation and deposition onto the
catalyst.

3.1. The Influence of the Reforming Temperature




                       Fig. 4. The influence of the reforming temperature over system efficiency

   The developed model gives information about the optimal reforming temperature. Fig. 4 shows efficiency
versus reforming temperature. The operation at high temperatures diminishes the processor efficiencies.
   The optimum water/ethanol molar ratio and temperature for the reformer unit (Fig. 2) and for the global
efficiency (Fig 3-4) do not coincide indicating the importance of adopting an integrated approach rather then
separated optimizations.
   In Fig. 5, we present a 3D graph efficiency as function of the water/ethanol molar ratio and reforming

temperature, the maximum efficiency occur at Tref =700 ºC and W/E=4.
               Fig. 5. System efficiency as a function of reforming temperature and water/ethanol molar ratio

3.3. Composite Curves
   The interest of the process integration techniques stands in the fact that it is able to identify heat exchange
bottlenecks in the system (pinch points) and to propose process modifications that will increase the overall
efficiency of the system. Using the LNG operation we can build the composite curves from the system. Fig. 6

shows the composite curves of system for the optimal operating conditions of Tref =700 ºC and W/E=4.




                                  Fig. 6. Composite curves for optimal Tref=700, W/E=4


   From the analysis of the composite curves of the system, to improve the efficiency is necessary to reduce the
plateau at the water and ethanol vaporization level. In this way, the cold and hot composites will be closer
meaning that both the required energy and energy losses will be minimized. The way in that this improvement
could be achieved as well as the analysis of different alternative configurations to improve the process
integration, will be analyzed in future works.

4. Conclusions

   A simulation model has been developed to analyze the performances of a proton exchange membrane fuel
cell system for stationary or mobile applications. The model implemented the heat exchanger network without
having to consider its configuration. This approach reveals to be very interesting in order to study the impact of
the major decisions parameters (water/ethanol molar ratio, reforming temperature).
   The solutions obtained from the simulation TPEMFC system model permit to achieve a total efficiency as
high as 38% that are much higher than the one of the original system design.
   The results of this preliminary analysis will be used to design the heat exchange network system and perform
a more accurate thermo-economic optimization. The present study can serve as a basis for the development of an
integrated PEMFC power pack for small-scale, e.g, household, or automotive applications.


References
Choi, Y., Stenger, H. G. (2004). Kinetics, simulation and insights for CO selective oxidation in fuel cell
   applications. Journal of Power Sources, 129, 246.
EG&G Technical Services, Inc. (2004), Fuel Cell Handbook, 7th ed.; U.S. Department of Energy, Washigton
García, E. Y., Laborde, M. A. (1991). Hydrogen production by the steam reforming of ethanol: Thermodynamic
   analysis. Int. J. Hydrogen Energy, 16, 307
Godat, J., Marechal, F. (2003). Optimization of a fuel cell system using process integration techniques. Journal
   of Power Sources, 118, 411
HYSYS, AEA Technology (2001). Calgary, Canada: Hyprotech Ltd.
Linnhoff, B., Townsend, D., Boland, D. W., Hewitt, G. F., Thomoas, B. E. A., Guy, A. R., Marsland, R. H. A
   (1985). User Guide on Process Integration for the Efficient Use of Energy; Institute of Chemical Engineers:
   Rugby, U.K.
Little, A. D. (1994). Multi-Fuel Reformers for Fuel Cells Used in Transportation-Multi-Fuel Reformers. Phase I.
   Final Report, Report #DOE/CE/50343-2, prepared for U.S. Department of Energy, Office of Transportation
   Technologies. Arthur D. Little, Inc., Cambridge.
Ratnamala, G. M., Shah, N., Mehta, V., Rao, P. V., Devotta, S. (2005). Integrated Fuel Cell Processor for a 5-
   kW Proton-Exchange Membrane Fuel Cell. Ind. Eng. Chem. Res., 44, 1535
Vasudeva, K., Mitra, N., Umasankar, P., Dhingra, S. C. (1996). Steam reforming of ethanol for hydrogen
   production: Thermodynamic analysis. Int. J. Hydrogen Energy, 21,13


Acknowledgements
   The authors want to thank for the financial support from CONICET (Consejo Nacional de Investigaciones
Científicas y Técnicas), ANPCYT (Agencia Nacional de Promoción Científica y Técnica), UNL (Universidad
Nacional del Litoral) from Argentina.