ThermalFluidDynamic Simulation of a PEM Fuel Cell Using a Hierarchical 3D1D Approach

1
Thermal-Fluid-Dynamic Simulation of a PEM Fuel
Cell Using a Hierarchical 3D-1D Approach

• Stefano Cordiner, Vincenzo Mulone, Fabio
  Romanelli
•  Department of Mechanical Engineering
• Università di Roma Tor Vergata

2
Index

• Introduction
• Background - Challenges
• Modelling Approaches
• Results
• Conclusions

3
Background - PEMFC

• Fuel cells are electrochemical devices that
  convert the chemical energy of a reaction
  directly into electrical energy.
• Unlike batteries, which store the energy, fuel
  cells operate continuously as long as they are
  provided with reactant gases.
• As by-products of this transformation water and
  heat are produced.
• These features make FCs ideal candidates for
  substitution of standard power generation systems
  both for stationary and transport applications.

4
Background Challenges
from www.ballard.com
5
Background Challenges

• A fuel cell system may be constituted
• the fuel cell stack
• a fuel processor and a fuel clean up
• an air management system
• an electric power conditioner
• heat and water management systems

6
System Schematic
7
Background Challenges

• A fuel cell system may be constituted
• the fuel cell stack
• a fuel processor and a fuel clean up
• an air management system
• an electric power conditioner
• heat and water management systems
• Each component needs to be properly optimized
• A proper integration of these component is then
  fundamental

8
Simulation Potential Target

• Simulation could represent a powerful tool to
  address some of the major issues
• System integration
• Control
• A more detailed comprehension of what actually
  occurs inside the stack
• Component design and optimisation

9
Modelling System level

• A system level approach consists of an integrated
  analysis tool capable of predicting overall
  system performance taking into account the
  individual components behavior
• it may be used also for
• lay out design
• transient analysis
• components matching

10
System level Layout architecture

• The system level simulation allows to evaluate
  the dynamic response of a given configuration
• Parallel architecture modelling
• FC stack
• Super-capacitors
• Load
• Current step 0-20 A

11
Modelling Stack Level

• At stack level the model could be used to
  describe phenomena as the heat exchange or the
  flow management
• still a simplified approach
• deeper insight capabilities
• At stack level the behavior of small scale
  effects could be either represented using
  semi-empirical relationships or could be derived
  from more detailed simulations

12
Modelling Cell Level
13
Modelling Cell Level

• Geometries
• straight channel
• complete cell

14
Modelling Cell level

• PEM Fuel cell modeling requires a proper
  description of
• Fluid flow and heat and mass transfer in both gas
  channels and porous electrodes
• Electrochemical reactions
• Multiphase flow with phase change
• Transport of current and potential field in
  porous media and solid conducting regions

15
Mixed 3D-1D approach - Cell level

• In our model we use a mixed 3D -1D approach
• The fluid dynamics is solved by a full 3D
  description
• The membrane physical/electrochemical behavior is
  described by a 1D approach
• Fluent 6.2 is used as the basis for CFD
  calculation
• The membrane behavior is described by specific UDF

16
Modelling Cell levelSchematic

• Bipolar plate
• Gas channels
• Porous gas diffuser
• Catalyst layer

17
Mixed 3D-1D approach - Cell level

• 1D approach across the electrolyte surface

18
3D Basic assumption - Cell level

• Main hypothesys of the 3D model are (Hu et al.,
  2004)
• Steady conditions
• Laminar flow in the channels
• Isotropic porous media
• Infinite electrodes conductivity
• Isothermal operation
• During the fuel cell operation liquid water may
  be formed resulting in a two phase transport
  phenomena
• Multiple phase mixture approach (Wang and Chen,
  1996)
• Hu et al. Energy Conv. Manag. 45, pp 1861-1882
  (2004)
• Wang and Cheng Int. Journal of heat and mass
  transfer 39 pp. 3619-3632 (1996)

19
Cell level 3D Model detailsGas channels and
Diffusion Layers

• Mass, Liquid water formation

• Momentum

• Species

20
Cell level 3D Model details Electrolyte

• 1D approach across the electrolyte
• Nernst potential
• Operational cell potential

21
Cell level 3D Model details Electrolyte

• 1D approach across the electrolyte
• Ohmic losses
• The membrane (Nafion 117) conductivity as a
  function of ? (water content) is expressed for ?
  gt1 by 1
• for ?lt1 we assume ?3.14E-3 (?cm)-1
• Computed current j is modified by jj(1-s) when
  liquid water exists on the electrolyte surface

22
Cell level 3D Model details Electrolyte

• 1D approach across the electrolyte
• Cathode overpotential from Tafel expression 1,
  2
• T. E. Springer, T.A. Zawodzinski, S. Gottesfeld,
  Polymer Electrolyte Fuel Cell Model, J.
  Electrochem. Soc., Vol. 138, No. 8, August 1991
• B.R. Siverstsen, N. Djilali, CFD-based modelling
  of proton exchange membrane fuel cells JPS 141
  (2005) 65-78

23
Cell level 3D Model detailsElectrolyte

• 1D approach across the electrolyte (UDF test)

24
Cell level 3D Model details Electrolyte

• 1D approach across the electrolyte
• Water transport

25
Cell level 3D Model details Electrolyte

• Mass sources

26
Cell level 3D Model details Solution

• Local 3D flow field data are sources for the
  local 1D
• Electrochemical data are calculated by the 1D and
  passed back to the 3D
• In this way the problem is completely and
  univocally defined for both models

27
Analysis of results
28
Analysis of results

• Model validation
• Fuel cell characteristics
• n of cells 8
• active area 50cm2
• membrane thickness 2.e-4 m
• gas diffuser porosity 0.5
• Gas feeding
• anode pure hydrogen st. 3 (min)
• cathode air st. 4 (min)
• Humidification
• anode dry
• cathode wet

29
Model Validation

• Comparison with experimental data

30
Cell level 3D Results

• Oxygen distribution

31
Cell level 3D Results

• Hydrogen distribution

32
Cell level 3D ResultsOxygen distribution

• 0.8 V
• 0.03 A/cm2

0.7 V 0.08 A/cm2
0.6 V 0.17 A/cm2
33
Cell level 3D ResultsLiquid water transport

• Gas (left) and Liquid (right) phase velocities
• I 0,7 A/cm2, V 0,4, u 50, T343 K

34
Cell level 3D ResultsLiquid water transport

• Inter-phases mass transfer rate (gas to liquid)

35
Cell Level 3D ResultsWater Transport through
the Electrolyte

• Straight channel, counter-flow configuration, dry
  fluxes
• Produced water at cathode side can diffuse
  through the membrane when current is low, at high
  current the drag effect overcomes the diffusion

36
Cell level 3D ResultsBipolar Plate Simulation
37
Cell level 3D Results

• Current distribution on the electrolyte

O2 distribution on the electrolyte
38
Cell level 3D Results

• H2 distribution on the electrolyte

39
Cell level 3D Results

• H2O (cathode side) and (anode side)

40
Conclusions

• A mixed 3D-1D simulation model has been presented
  as part of a complete framework of simulations
  tools which can be used both for components and
  system analysis of PEM FC based energy generation
  systems
• The model has proved to be able to reproduce
  experimental data
• The model can be used to deeply analyze fluid
  dynamic processes in the gas channel and in the
  GDL
• The model is also able to describe the multiphase
  flow and reproduce the water formation phenomena
  at high specific currents

41
Next steps

• The analysis of different configurations as
  cross channel and interdigitated
• The link of the process with an automatic
  optimisation procedure to speed the find of
  optimal configurations

42
Thermal-Fluid-Dynamic Simulation of a PEM Fuel
Cell Using a Hierarchical 3D-1D Approach

• Stefano Cordiner, Vincenzo Mulone, Fabio
  Romanelli
•  Department of Mechanical Engineering
• Università di Roma Tor Vergata