Method and device for the stacking of fuel cells
The present invention relates to fuel cell stacks. A fuel cell has a structure with sets of electrochemical elements including anode, membrane and cathode, arranged between end plates. More particularly, the invention relates to certain sub-components that are assembled together to build up a fuel cell stack and relates to the process to assemble or to repair a fuel cell.
In a fuel cell, there is an anode situated between the electrolyte and one end plate, and a cathode situated between the electrolyte and the other end plate. Solid electrolytes are known; accordingly, the electrolyte may itself possess the necessary rigidity to be fitted in the cell. In the same way the operating temperatures vary considerably, differing from the ambient temperature by up to several hundred degrees Celsius or more. Joining individual fuel cells together to form a fuel cell stack, in order to obtain the desired operating voltage through the series connection of a suitable number of cells is also well known and is made routinely.
In addition to the individual fuel cells, then, a fuel cell stack contains special end plates on either side of the stack, which terminate the stack, and in front of each of the said end plates a connection plate (for example, a gold-coated copper plate) with the connections, in order that the stack can be connected to the current conductors leading to the power consumer. The said elements are stacked until the fuel cell stack comprises the desired number of fuel cells, and are then clamped including the connection plates and end plates. This can be done by means of tie rods, which run right through the stack and are bolted to the end plates.
Assembling such a stack is costly, since all the elements must be laid one on top of another, piece by piece, and held in the correct seating position until finally bolted. The connection plates, for example, are moreover expensive to manufacture.
It is possible to simplify assembly by the use of so-called bipolar plates. A bipolar plate is an assembly including the plates of two adjacent fuel cells arranged in one part, with the advantage that in assembling multiple fuel cells to form a fuel cell stack the number of parts to be assembled can be reduced. Additional cooling ducts can be arranged in the bipolar plates, which leads to a substantial improvement in the heat balance of the fuel cell stack. During operation of the fuel cell stack, current flows in a bipolar plate, since this constitutes the electrical connection between the anode of one adjacent fuel cell and the cathode of the other adjacent fuel cell.
It is equally possible to bond adjacent end plates of the individual fuel cells to form a bipolar plate. Instead of clamping, individual elements or all elements of the fuel cell or the entire fuel cell stack could be adhesively bonded together, which provides a functional solution. This process is likewise expensive.
Prior inspecting the parts is required to avoid bonding damaged parts such as parts featuring scratches etc; then assembling implies uniform application of the adhesive, drying on, dissipation of the solvent vapours, bonding under pressure and heat, cleaning off excess adhesive (blockage of gas supply and cooling ducts etc), checking that all media such as air, hydrogen, water etc are properly separated from one another etc. All the above mentioned operations are obstacles to easy, rapid assembly of the elements to form a fuel cell or a fuel cell stack. This is especially non productive since a fuel cell stack may contain many fuel cells. For example, a stack of fuel cells having a polymer membrane as electrolyte with an output of 7kW and weighing approximately 20 kg has around 100 cells.
There is also the question of testing a newly manufactured fuel cell stack, and of repair and maintenance; individual defective cells must be withdrawn from the stack, repaired or exchanged before they can be reinserted. Although this is basically not impossible in the case of a bonded stack, the effort involved is prohibitive. Even with a clamped stack the effort is still considerable: the clamping must be released, which demands particular care when handling the cells which are not to be replaced. These should not be separated from one another and in particular should not be dismantled as a unit, so as not to disturb the original seating of the individual elements.
A further disadvantage lies in the use of bipolar plates: replacement of the selected cells must include the end plates of the adjacent cells. The adjacent cells must therefore be separated; electrodes and electrolyte lose their original seating. The advantage accruing in the manufacture of the fuel cell stack becomes a disadvantage in the further life of the stack.
It has become known practice, therefore, to form sets of two end plates and to seal them along their common edge using a corded seal in the nature of an O-ring, thereby creating a unit identical in function to a bipolar plate. In so doing, the precise machining of the adjoining surfaces of the two end plates is critical, since the contact resistance must remain small (otherwise the fuel cell stack loses efficiency) and cooling ducts should remain tightly sealed. The run and sealing
property of the cord seal is just as critical, since depending on the construction of the fuel cell stack the gas supply ducts have to be sealed off from one another (one duct carries oxygen, for example, and the other duct hydrogen). Fitting the expensive cord seal (which has to be individually manufactured for specific end plates) is difficult and therefore costly.
The disadvantages cited above also apply to the manufacture and assembly of the connection and end plates of the stack with the associated seals. Although they are not formed in one piece, for the sake of descriptive simplicity, sets of two joined end plates resembling the bipolar plates will likewise be referred to as bipolar plates, since these sets have the same function in the fuel cell stack as the bipolar plates formed in one piece.
Consequently, it is possible on the one hand to use one-piece bipolar plates, with the disadvantage that when detaching a fuel cell from the fuel cell stack it is impossible to avoid separating adjacent cells. The alternative is to use separable, multipart bipolar plates as described above, which, however, entail considerable cost in manufacturing and joining together.
No design construction has hitherto been disclosed, which affords the sound functionality afforded by the one-piece bipolar plate, for example, and which facilitates assembly and handling whilst also reducing the cost of manufacture, assembly and maintenance of the ends of the fuel cell stack with their end and end plates that also have to be kept fluid-tight.
Accordingly it is an object of the present invention to overcome these disadvantages and to provide an improved design construction.
The invention proposes a plate assembly for a fuel cell having a structure with an electrolyte zone, said plate assembly containing at least one duct for supplying a fluid to the electrolyte zone, said plate assembly contacting said electrolyte zone, the plate assembly comprising at least two plates, at least one of which including said at least one duct, a flexible intermediate layer being removably arranged between said two plates.
The invention also proposes a bipolar plate assembly for a stack of a plurality of fuel cells, said fuel cell stack comprising at least two cells connected in series, each having a structure with an electrolyte zone, said bipolar plate assembly joining the electrolyte zones of two adjacent cells, said bipolar plate assembly comprising a plate assembly as proposed hereabove.
The invention also proposes an end plate assembly for a single fuel cell or for an end cell of a stack containing a plurality of fuel cells, said cell(s) having a structure with an electrolyte zone, said end plate assembly joining the electrolyte zone of a cell and to a current collector, said end plate assembly comprising a plate assembly as proposed hereabove.
The invention also extend to a fuel cell stack having a plate assembly or a bipolar plate assembly or an end plate assemply as proposed hereabove, and to a flexible intermediate layer for a plate assembly or for a bipolar plate assembly or for an end plate assembly as proposed hereabove.
Additionally, the invention proposes a method of changing at least one fuel cell or fuel cells in a fuel cell stack having intermediate layers wherein, having released the clamping of the stack, the end plates of adjacent fuel cells forming bipolar plates are separated from one another, and the cells to be changed with their associated end plates are removed together with the intermediate layers thereby released, and new intermediate layers are used in assembling intact cells.
The use of a conductive intermediate layer firstly dispenses with the direct, electrically conductive contact between the end plates, which also allows the intermediate layer to be designed as a flat element extending over the dimensions of the end plates, thereby providing an easily handled element for the assembly of a fuel cell stack or the replacement of fuel cells in a fuel cell stack.
In addition to achieving the above mentioned advantages, a compressible intermediate layer furthermore allows it to bear flexibly against the adjoining surfaces of the adjacent end plates, with the result that less stringent demands can be placed on the production tolerances for the end plates. This applies not only to the local surface quality but also to the design form of the surface over its entire dimension, whether this be of convex or inclined design, for example, instead of a flat design. A deformable intermediate layer is capable of offsetting such machining defects, so that the machining itself can be simpler and hence much less expensive.
In order to assemble a fuel cell stack, the individual elements are first placed together and the clamping then applied. With the use of the intermediate layer according to the invention, all end plates or bipolar plates can be aligned in the stack during clamping with deformation of the
intermediate layers. There is no overstressing of the electrolyte, for example, or of the carrier structure (Polymer Exchange Membrane or PEM) in the finally assembled stack.
In one particular embodiment the effort of assembly is significantly further reduced if the intermediate layer is fixed to one end plate, by adhesive bonding, for example. Instead of three elements to be assembled, the bipolar plate according to the invention then only has two.
It is also possible to manufacture a bipolar plate according to the present invention as a unit by bonding both end plates to the intermediate layer and to fit this in assembling the fuel cell stack, which reduces the assembly cost. However, the unavoidable disadvantage of having to separate the adjacent fuel cells when replacing a cell, as described above, then has to be accepted into the bargain.
Since the end plates may have a thickness in the order of a millimetre and longitudinal and lateral dimensions in the order of 10 cm or more, for example, the cooling ducts must be designed as an open labyrinth on the intermediate layer side, so that the labyrinth is covered by the intermediate layer and the cooling ducts are sealed. In contrast to the known one piece bipolar plates, a fully bonded bipolar plate according to this invention allows cooling ducts to be provided in the plates easily.
The intermediate layer according to the invention is advantageously impervious to the ambient air and to the media used, such as hydrogen or oxygen and a cooling medium. This makes it possible to achieve a compact stack construction.
Flexible graphite possesses the stated characteristics. The invention is not limited to this material, however, but encompasses all materials that are conductive and can be used as intermediate layer.
Equipping the intermediate layer with an electrically conductive structure obviates the need to use a hitherto indispensable plate for the connection of the electrical conductors to the fuel cell stack. Since the copper plates hitherto used had to be provided with corrosion protection (e.g. gold coating), the present invention not only simplifies assembly but also brings a significant reduction in manufacturing costs.
The invention will be explained in more detail with reference to the drawings, in which:
Fig 1 shows a schematic diagram of the structure of a stack of fuel cells according to the prior art, Fig 2 shows a schematic diagram of the structure of a stack of fuel cells according to a first application of this invention,
Fig 3 shows a schematic diagram of the structure of a bipolar plate using an intermediate layer according to the invention, Fig 4 shows in more details a view of the first embodiment of an intermediate layer according to the invention, Fig 5 shows a schematic diagram of the structure of a second embodiment of an intermediate layer according to the invention, Fig 6 shows a schematic diagram of the structure of a stack of fuel cells using the first embodiment of an intermediate layer according to the invention, Fig 7 shows a schematic diagram of the structure of another stack of fuel cells, using the second embodiment of an intermediate layer according to the invention,
Fig 8 shows a section into an intermediate layer according to one of the above first embodiment, Fig 9 shows a section into a third embodiment of an intermediate layer according to the present invention, Fig 10 shows a detail of the third embodiment of an intermediate layer according to the invention, Fig 11 show a general layout of an intermediate layer according to the third embodiment.
Fig. 1 is a schematic representation showing an example of a stack IP of fuel cells formed from the two PEM fuel cells n and n+1, as known in the prior art. Such an arrangement will be familiar to the person skilled in the art. For the sake of simplicity, the stack IP is shown with only two fuel cells n and n+1; but it may contain any number of cells. In a common application, for example, one hundred cells are arranged in the stack. Each fuel cell n and n+1 comprises end plates 2 and a part of a so-called bipolar plate 3, the cells n and n+1 being joined as already known to form the bipolar plate 3. Between the plates 2 and 3 is a polymer membrane 4, and between the membrane 4 and each of the plates 2, 3 an anode 5 or a cathode 6 respectively. Ducts 7 serve to deliver gas, H2 to the anode side and O2 or ambient air to the cathode side (other possible reactants will also be known to the person skilled in the art). The ducts are designed in such a way that the gases can be delivered as uniformly as possible to the entire membrane
surface. In order not to clutter the diagram, gas supply and removal channels which run along the side faces 20, 21 of the stack IP and which serve the ducts 7, are not shown. Cooling ducts in the bipolar plate 3 are likewise omitted.
The layout of the channels and the cooling ducts will be familiar to the person skilled in the art. The channels basically run the entire length of the stack IP and terminate blindly at one end of the stack IP, whilst at the other end of the stack IP they generally pass through its end plate, so that they can be connected to an external supply. The same applies to the coolant supply for the stack IP. An electrical load 8 is connected by a conductor 81 to one anode 5 and to one cathode 6. The figure 1 in schematic form shows the direct connection of the conductor 81 to the electrodes.
Fig. 2 shows a fuel cell stack 1A having a separable bipolar plate 31 according to the present invention. The structure of the stack 1A corresponds to the stack IP of Fig. 1, except for the separable bipolar plate 31 formed from plates 2' and the intermediate layer 30. The plates 2' are very similar (they could be identical) to the end plates 2.
The general structure of the stack 1A according to Fig. 2 is, as stated, be familiar to the person skilled in the art. The geometry of the intermediate layer 30 is not limited to the shape represented in Fig. 3; such is also the case for any other element of the stack. The intermediate layer 30 is preferably composed of flexible graphite, like that marketed in Germany by SGL Carbon AG under the name "Sigraflex". The intermediate layer may be made in the form of a film with a thickness ranging from 1 mm to 0J mm. The thickness is preferably less than 1 mm, in the region of between 0.5 mm and 0.2 mm.
The above mentioned material conforms flexibly to the surface contour of the plates 2' and allows current to be transmitted over the entire contact surface with minimal transmission resistance. At the same time, the flexible conformity allows the machining tolerance in the production of the plates 2' to be relaxed, which reduces the scrap rate for defectively produced plates, thereby saving costs. If, for example, the contact surface of a plate 2' to the intermediate layer 30 is not aligned exactly parallel with the opposing surface in contact with the anode or cathode (the plate 2' then has a wedge shape), this results in an inclined position of the other plate 2'. Without the benefit of the invention, this inclination leads to stresses in the stack IP, e.g. to an inadmissible, local loading of the membrane 4, which reduces the efficiency of the fuel cell affected and hence of the entire stack IP. The deformable intermediate layer 30 eliminates this risk, since the
relative mutual position of the plates 2' can be compensated for by the deformation of the layer 30.
Fig. 3 shows another application in which the use of an intermediate layer 30 according to the first embodiment allows to dispense with the use of a cord seal. One can see a bipolar plate 3 IB in an exploded view similar to Fig. 2, with the adjacent anode 5 and cathode 6. The plate 2a belongs to the fuel cell n and the plate 2b to the fuel cell n+1. A flexible intermediate layer 30 is situated between the plates 2a and 2b. Cooling ducts 9 are arranged in the end plate 2a and are sealed off by the intermediate layer 30, thus equipped with the appropriate number of holes like holes 10 shown in Fig. 3, to provide channels supplying and recycling cooling fluid in and from ducts 9.
Fig. 4 shows in more details the intermediate layer 30 according to the invention, having a body 11 surrounded with edge sections 12 having holes 10. Each hole 10 is a small section of one of the channels supplying one of the required fluids to the stack 1A. As well known by those having a working skill in the art, similar holes and edge sections also are provided to the surrounding part of each plate 2', at least to the polymer membrane 4, and possibly to the anode 5 and cathode 6 (see Fig. 2), all those elements 2 and 4 having a layout similar to what is shown in Fig. 3 for the intermediate layer 30, for the elements 5 and 6 other layouts are possible too. The edge sections 12 of the intermediate layer 30 provide a fluid-tight seal for, so that the aligned holes 10 constitute gas supply and removal channels which run along the outsides 20, 21 of the stack 1 and feed the ducts 7 (Fig. 3) running transversely through the stack 1A.
The intermediate layer 30 furthermore has edge sections 13 forming a fluid-tight seal for the edges of the end plates 2'. This prevents any undesirable contact of the various media. Also shown are recesses 14 for tie elements, which mechanically hold the stack 1A together. The geometry of the intermediate layer 30 thereby corresponds to the geometry of the adjacent elements in the stack 1A, whether these are end plates 2 or plates 2' that are part of the bipolar plate 31 or any anode or cathode for current connection.
Fig. 6 shows a stack 1C with individual fuel cells n-x, n and n+x. The ends 40, 50 of the stack 1C have end plates 41 (the edges of plates 41 seals lateral channels, openings being optionally provided for connecting the channels to the external supply and recycling of fluids). Also shown are connection plates 42 with connecting poles 53 for the connection to an outside conductor (not
shown). A cooling plate 44 in the form of a structured graphite plate is also shown in the end 40 and adjacent to the end 50. An unstructured plate 54 is inserted in the end 50. The plates 44, 54 prevent any cross-flow of the media, the plate 44, as stated, being additionally provided with cooling ducts 9.
The fuel cells n-x, n and n+x have the structure described for Fig. 1, but the cooling ducts 9 running transversely through the bipolar plates 3. A cord seal SD, which prevents undesirable contact of the various media circulating in the stack, is provided between all the above-mentioned elements (end plates 41, connection plates 42, cooling plate 44 and plate 54 and in the bipolar plates 3). The connection plates 42 comprise a second embodiment of an intermediate layer 70 according to this invention.
Fig. 5 shows the form of the intermediate layer 70 according to this second embodiment. A copper part 71 (good conductor for electricity) for the connecting of poles 53 and which draws off the current generated by the stack 1C to a outside conductor (not shown but similar to conductors 81), has been inserted. The copper part 71 has an outline contour such that a positively interlocking seat is obtained in the intermediate layer 70, and has the largest possible surface area for low contact resistance between the copper part 71 and the intermediate layer 70. The intermediate layer 70 shown in the figure is preferably of a three-sub-layer structure: two outer layers of flexible graphite have the configuration corresponding to the outline contour of the hatched area in the figure, so that the copper part 71 can be inserted into the free space and is covered by the outer layers.
In a variation of this embodiment, a conductor lattice can be sunk into the intermediate layer 70 and connected to the poles 53. The conductor lattice can also be inserted between sub-layers; the capacity for deformation displayed by the flexible graphite or some other substance with the same characteristics produces a more secure seating of the lattice and uninterrupted current flow. The contact resistance diminishes the more close-meshed the lattice.
It would be equally possible to use a conductor in the shape of a stranded cable, and to insert the strands between the sub-layers or to sink them into an intermediate layer 70. The manner in which current is led off from the stack 1C by means of an intermediate layer 70 may be modified, as required, by the person skilled in the art, but always without departing from the scope of the
invention. Thanks to this second embodiment of the invention, a special connection plate with gold coating can be dispensed with.
The above example shows that the use of an intermediate layer according to the invention can be mixed with conventional configuration of several cells in a fuel cell stack. Obviously, as shown in Fig. 7 representing a stack ID, the intermediate layer according to this invention can be used in all and every cells of the stack. The fuel cell stack ID shown in Fig. 7 comprises intermediate layers 70 and 30. The material of the intermediate layers 70 and 30 conforms flexibly to the surface contour of the plates 2a, 2b.
A third embodiment is explained below with Figs. 8 to 11. Where the application of the fuel cell stack requires a great compactness of the stack, it is proposed a third embodiment of an intermediate layer 60 with integrated cooling ducts. It achieves a radical saving of construction volume and weight due to an integration of functions. This is especially useful for technical applications, where the arrangement of series of cells usually uses bipolar plates 3. The third embodiment proposes an intermediate layer 60 of another design. Indeed, if the use of an intermediate layer 30 providing a separable bipolar plate has several advantages regarding to costs, electric conduction and manufacturing compared to gluing or making otherwise integral bipolar plates, however the additional graphite layer increases the construction volume of the bipolar plate. The third embodiment of this invention proposes a design of the gasket that uses the thickness of the intermediate layer to embed at least in part the cooling ducts. Accordingly, the total thickness of the bipolar plate can be lowered.
Fig. 8 shows a bipolar plate 31C where plates 2C have on one face ducts 7 supplying hydrogen or oxygen or other suitable reactants, and have on the opposite face cooling ducts 9 for circulating a cooling fluid. The total thickness of plates 2c is dp, and total thickness of the intermediate layer 30 is dg.
Figure 9 shows a bipolar plate 3 IE comprising one plate 2C, as above, and one plate 2'. It could be two plates 2C or two plates 2'. The bipolar plate 3 IE comprises a intermediate layer 60 which includes ducts 61 for circulating a cooling fluid. The comparison of Fig. 8 and Fig. 9 allows to understand that this design allows to increase the section of cooling ducts all other things being unchanged, with the benefit of an improved cooling. Alternately, this design allows to decrease the thickness of the bipolar plate, if the desired cooling performance is unchanged, with the
benefit of an improved compactness. In this case, the total thickness of plates 2' is less than the total thickness of plate 2c.
In order to keep the intermediate layer 60 as a single self sustaining piece, it is necessary to discontinue the cooling ducts 61 at places such as in the vicinity 62 (see Fig. 11) of holes 10E supplying the cooling fluid from outside channels. It is preferable to reinforce the intermediate layer 60 by discontinuing the cooling ducts 61 in some other zones such as in the central zone 63. Consequently, to retain the continuity of the cooling ducts, it is required to bridges the vicinity 62 and any similar zone like the central zone 63 with bridging grooves provided in the adjoining face of adjacent plates 2C and or 2'. As shown in Fig. 10, in order to maintain the total cross-section of the cooling ducts in the bridges while designing very thin grooves for bridging purposes, it is possible to increase the width of the bridging grooves, or to provide two or more bridging grooves 29 (Fig. 11) without impairing the robustness of adjacent plates 2C or 2' (because the grooves are very local).
The intermediate layer 60 can be manufactured for instance by punching an intermediate layer 30 with the appropriate tool, featuring the design of the ducts 61.
The intermediate layer 60 e.g. has a thickness of 300 μm in the compressed condition, that allows to diminish the thickness of one plate by this amount comparing with embodiment 30. Since the thickness of adjacent plates 2C is roughly 3,3 mm, the construction volume can be decreased about 10 percent by using only plates 2'. Also, the use of the third embodiment design allows for reducing the weight of the bipolar plate as well as for the intermediate layer, about 10 percent.
In all the embodiments shown, the passage of current through the bipolar plate 3 or an end plate according to the invention varies as a function of the dc resistance of the intermediate layer 30, 60, 70 itself and of the contact resistance between the intermediate layer and the adjacent plate. It has emerged from test measurements that the contact resistance from end plate 2a to end plate 2b -as is known, for example, where a cord seal according to the prior art is used- is equal to the sum of the resistances (contact and dc resistance) when an intermediate layer of flexible graphite with a thickness of 0,5 mm is inserted in a bipolar plate 3 according to the invention under the pressure prevailing in the usual clamping of a fuel cell stack.
This means that the use of an intermediate layer according to the invention of correctly selected thickness affords a significant advantage in the transmission of current through the bipolar plate, regardless of the other advantages such as more cost-effective production of the end plates, greater ease of assembly etc.