Bond Graph Model of a PEM Fuel Cell
G. Fontes, R. Saïsset, C. Turpin, S. Astier
Laboratoire d’Electrotechnique et d’Electronique Industrielle, LEEI, UMR CNRS/INPT 5828
2 rue Camichel - 31071 Toulouse cedex 7 - France
Phone: (+33) 5 61 58 82 08, Fax: (+33) 5 61 63 88 75, E-mail: name@leei.enseeiht.fr
Keywords: fuel cell, modelling, experimental
validation
Abstract – The PEM fuel cell is itself a
multidisciplinary system: hydraulics, chemistry,
electrochemistry, thermal, and electrical engineering
are concerned. All these fields are strongly coupled.
The authors chose an energy approach to model the
PEM fuel cell applied through the Bond Graph
representation. In a first part, the fuel cell operation is
described. In a second part, the PEM fuel cell is
proposed. And at last, experimental validations are
presented.
I. INTRODUCTION
Today, Bond Graph modelling is not a common approach
to model electrochemical components (fuel cells,
accumulators, batteries…). However, Bond Graph
modelling appears well suited to these multi-disciplinary
components: chemistry, electrochemistry, thermal and
electrical engineering are at least concerned. It is an
original approach, especially for the electrochemistry
community. The proposed model of a PEM fuel cell was
developed in collaboration with electrochemists from the
Laboratoire Genie Chimique Toulouse (LGCT). The
initial goal was not to develop a model in order to
optimize the internal design of a PEM fuel cell, but to
develop a model the most physical as possible (an
intermediary model between local models which solve
differential equations, and models, such as “black boxes”,
which are efficient for control system) in order to simulate
systems including a PEM fuel cell.
Our approach allowed to create a generic model (here
applied to a PEM fuel cell) which can be applied to others
electrochemical components (Li-ion accumulators, Leadacid
accumulators…)
II. OPERATION PRINCIPLE AND IMPLEMENTATION OF A
PEM FUEL CELL
[LARMINIE, 2000] [HIRSCHENHOFER et al., 1998]
The operation principle of a Proton Exchange Membrane
(PEM) fuel cell consists in making react hydrogen and
oxygen to produce water, electricity and heat. It is an
oxidation-reduction reaction in the presence of platinum,
opposite of the water electrolysis reaction:
- + + ® e H H 4 4 2 2 (anode) (1)
O H e H O 2 2 2 4 4 ® + + - + (cathode) (2)
heat lectricity e O H O H + + ® + 2 2 2 2
1
(total reaction)
(3)
Its implementation consists in stacking elementary cells in
series (Fig. 1) to obtain exploitable voltages with a very
good efficiency. Each cell includes two electrodes (anode
and cathode) separated by an electrolyte which is
electronic insulator and ionic conductor.
positive
end plate
negative
end plate
bipolar
plate
calalyst layer
diffusion layer
electrolyte
(membrane)
anode cathode
ee-
bipolar
plate
H+
H2O
e-
H2
channels
O2
channels
H2 O2
MEA
(Membrane Electrode Assembly)
e-
LOAD
I
H2 in
O2
O2 + H 20
out
H2 out
Fig. 1: PEM fuel cell fed by hydrogen and O2 (stack of 3 cells in series
electrically; parallel gas supplies)
The bipolar plates provide the electrical connection
between two adjoining cells, supply gases on two faces
(O2 for the cathode of a cell, H2 for the anode of the
juxtaposed cell), evacuate excess gas, produced heat and
water. Graphite answers all these criteria and is generally
used. The typical thickness of the bipolar plates is one
centimeter.
The electrodes must be porous to allow the gas diffusion
.....