"A FUEL CELL STACK" FIELD OF THE INVENTION

This invention relates to a fuel cell. In particular, the invention relates to a fuel cell stack which utilises an improved heat exchanger. BACKGROUND OF THE INVENTION

In recent years, fuel cell stacks have become popular. This is largely due to their environmental friendliness as the fuel that they use is typically in the form of hydrogen or other environmentally friendly fluids. However, fuel cell stacks have inherent problems which .reduce their appeal.

A fuel cell stack is composed of a number of fuel cells. Each fuel cell is typically made up of a cathode and an anode which sandwich an electrolyte membrane. A popular choice for the production of the cathode and anode is the use of bipolar plates. Bipolar plates are conductive plates in a fuel cell stack that can act as an anode for one cell and a cathode for the next cell. Bipolar plates can be made of graphite, metal or conductive composite polymers (possibly incorporating carbon). Graphite is popular due to its electrical and thermal conductivity.

Each fuel cell of a fuel cell stack must be maintained at a relatively constant temperature in order to prevent destruction of the fuel cell through thermal loading. This is particularly challenging as the reaction within the fuel cell is exothermic so a large quantity of heat is generated. In order to manage the thermal energy of a fuel cell stack, heat exchangers are normally used. Heat exchangers for fuel stacks generally use two types of coolants, namely air or water. This is typically due to their availability; As a significant part of each thermodynamic cycle of each fuel cell stack, heat exchangers have been applied to fuel cell stacks to remove the heat released by each single cell to provide a constant operational temperature.

Heat exchangers are normally produced by machining channels in an outwardly facing side of both the cathode and the anode. These channels allow coolant to flow adjacent to the cathode and anode removing heat from the fuel cells. Air-cooled heat exchangers pass air through the channels and are generally an open system. Water-cooled heat exchangers pass water through the channels and require a closed system in order to contain the coolant.

In order to enable the channels to be machined into an outwardly facing side of both the cathode and the anode, a thicker anode and a thicker cathode are required. This dramatically increases the price of the cathode and the anode as substantially more material is required.

Another inherent problem with fuel cell stacks is that the cathodes and anodes must be durable, resistant to corrosion, and offer low contact resistance. However, to remove heat generated within the fuel cell stack, the coolant should be distributed evenly between the two adjacent fuel cells. Therefore, and as stated above, very tiny channels are machined in an outwardly facing side of both the cathode and the anode. This increases the contact resistance between adjacent fuel cells and leads to a high electrical resistance for a fuel cell stack.

OBJECT OF THE INVENTION It is an object of the invention to overcome and/or alleviate one or more of the above disadvantages and/or provide the consumer with a useful or commercial choice. SUMMARY OF THE INVENTION

In one form, the invention resides in a fuel cell stack comprising:

a plurality of fuel cells for producing electrical power;

a heat exchanger material located between adjacent fuel cells; and wherein the heat exchanger material is a metal foam.

Normally, each fuel cell has a cathode and an anode which sandwich an electrolyte membrane. The cathode is normally in the form of a plate. The anode is also normally in the form of a plate. Preferably, the cathode and anode plates are in the form of bipolar plates.

The cathode may be made from any suitable material. Preferably, the cathode is made from graphite. However, other materials such as stainless steel, aluminium, platinum, and gold may be used to produce the cathode. The anode may be made from any suitable material. Preferably, the anode is made from graphite. However, other materials such as stainless steel, aluminium, platinum, and gold may be used to produce the anode. At least one cathode may have an outwardly facing side which is substantially planar. Normally, all of the cathodes have an outwardly facing side which is substantially planar.

At least one anode may have an outwardly facing side which is substantially planar. Normally, all of the anodes have an outwardly facing side which is substantially planar.

The stack can be made of proton exchange fuel cells or solid oxide fuel cells, consuming hydrogen, butane, methanol, and other petroleum products as ^' the fuels. The heat exchanger material generally connects an anode of one fuel cell to a cathode of an adjacent fuel cell. The heat exchanger material is typically in the form of a metal foam plate. The metal foam may be made from suitable metals such as copper, aluminium, stainless steel, gold and their alloys.

Preferably, the metal foam is made from aluminium. The weight of a metal foam plate is normally 30-80% lower than a cathode/anode bipolar plate of the same dimensions.

Preferably, the metal foam has a thermal conductivity of 20-30 W K ^_1 m ^"1 , being 200 times higher than that of air, and 20 times higher than that of water.

Preferably, the metal foam plate has an electrical conductivity at least one order of magnitude higher than a graphite cathode/anode bipolar plate of the same dimensions.

Preferably, the metal foam plate has 50% lower electrical and thermal contact resistance than a graphite cathode/anode bipolar plate of the same dimensions. Preferably, the metal foam plate provides a sealing means in order to seal the gas flow paths between each single cell. In one embodiment, the sealing means is a silicon gasket.

Preferably, the metal foam has porosity above 0.9. Suitably, the metal foam has a PPI (Pore Per Inch) value of 5-50. Preferably, the cell size of the metal foam can be in the range of 0.5 - 3 mm.

Preferably, a coolant is passed through the metal foam heat exchanger. The coolant is preferably air or water.

BRIEF DESCRIPTION OF THE DRAWINGS

An embodiment of the invention will be described with reference to the accompanying drawings in which:

FIG. 1 shows a top view of a fuel cell stack according to an embodiment of the invention; and

FIG. 2 shows an exploded perspective view of a fuel cell stack according to FIG. 1.

FIG. 3 shows ^' a photograph of a testing rig used to test properties of the metal foam heat exchanger; FIG. 4. shows a photograph of metal foam heat exchangers within the test rig; ^{■ ' ~}

FIG. 5. shows a further photograph of metal foam heat exchangers within the test rig;

FIG. 6 shows a graph which represents heat transfer rate Vs. air mass flow rate of the metal foam heat exchanger; and

FIG. 7 shows a graph which represents pressure drop Vs. air mass flow rate of the metal foam heat exchanger DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

FIGS. 1 to 2 show an air cooled fuel cell stack 10 for generating power. The fuel cell stack 10 shown has six fuel cells 20 located between seven heat exchanger plates 30. It should be appreciated that the fuel cell stack 10 may have a larger number of fuel cells 20 with a heat exchanger plate 30 located between respective adjacent fuel cells 20 with additional heat exchanger plates 30 also located at ends. For example, a fuel cell stack 10 which has ten fuel cells 20 will have eleven heat exchanger plates 30. That is, nine heat exchanger plates 30 located between each respective pair of fuel cells 20 with two additional heat exchanger plates 30 located at ^' the ends. Alternatively, there may be nine heat exchanger plates 30 located between each respective pair of fuel cells 20 without plates at the ends. Each of the fuel cells 20 is formed from a cathode plate 40 and an anode plate 45 which sandwiches an electrolyte membrane 50. Both the cathode plate 40 and the anode plate 45 can be formed from graphite. However, it should be appreciated that other suitable materials such as composite polymers, aluminium, steel, copper, titanium and gold can be used to form the cathode plate 40 and anode plate 45. Both the cathode plate 40 and the anode plate 45 have an outer face which is substantially flat. Both the cathode plate 40 and the anode plate 45 have gas flow holes 60 with associated gas flow gaskets 70. Although the cathode plate 40 and the anode plate 45 are indicated as being separate structures in FIGS. 1 and 2 it will be appreciated that the use of bipolar plates to act as cathode/anode for alternate cell stacks is also considered.

The electrolyte membrane 50 is a standard membrane assembly 50, such as for example a polymer electrolyte membrane, that is known to people skilled in the art. It should be appreciated that the electrolyte membrane 50 may be suitable for any type of a fuel cell including a proton exchange fuel cell or a solid oxide fuel cell. The heat exchanger plates 30 are preferably made from aluminium metal foam but, alternatively, may be made from foamed alloys of aluminium, copper, and/or steel as the base material for the metal. Accordingly, each heat exchanger plate 30 has a series of interconnected apertures which form tortuous pathways in which fluid (such as air or water) is able to flow through. The apertures can range in size from about 0.5 to about 3.0 mm including 0.5, 1.0, 1.5, 2.0, 2.5 and 3.0 mm. The weight of the metal foam heat exchanger plate 30 is normally 30-80% lower than a graphite cathode/anode bipolar plate of the same dimensions. In one embodiment the weight of the metal foam heat exchanger plate 30 is 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or 80% lower than a graphite cathode/anode bipolar plate of the same dimensions.

Preferably, the' metal foam has a thermal conductivity of 20-30 W K ^~1 m ^~1 , being 200 times higher than that of air, and 20 times higher than that of water. This includes the metal foam having a thermal conductivity of 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29 or 30 W K ^"1 m ^"1 ,

Suitably, the metal foam has a PPI (Pore Per Inch) value of 5-50. This includes values of 5, 10, 15, 20, 25, 30, 35, 40, 45 and 50 PPI.

Preferably, the cell size of the metal foam is in the range of 0.5 - 3 mm. This range includes values of 0.5, 1.0, 1.5, 2:0, 2.5 and 3.0 mm.

A preferred metal foam suitable for use with the present invention will present one or more, preferably substantially all, of the following characteristics as set out in Table 1 : Table 1 : Preferred physical properties of suitable metal foams

The heat exchanger plates 30 are structural bodies. That is, they are able to carry a load. As with the cathode plates 40 and the anode plates 45, the heat exchanger plates 30 have gas flow holes.

In order to assemble the fuel cell stack 10, heat exchanger plates 30 are located between and connect cathode plates 40 to anode plates 45 of adjacent fuel cells 20. Additional heat exchanger plates 30 are located adjacent the end cathode plate 40 and end anode plate 45. When assembly occurs, the gas flow holes are aligned as is standard practice. The fuel cells 20 and heat exchanger plates 30 are then compressed together using supporting structures and placed in a housing (not shown) ready for use. In use, a pump (not shown) is used to suck or pump cold air through the heat exchanger plates 30 which are in contact with the hot cathode plate 40 and anode plate 45. The cold air is used as a heat sink and removes the heat from the graphite plates and leaves the heat exchanger plate 30 at a relatively higher temperature, which thermal energy may be used in other applications.

It should be appreciated that a water loop may be used instead of air as the heat sink. It would be evident to a person skilled in the art how to accomplish a closed fuel cell stack 10 that uses water as the coolant. The heat exchanger plate 30 being made of metal foam provides a number of advantages. Firstly, the anode plates 45 and cathode places 40 (which may take the form of bipolar plates) can be made from thinner graphite plates as machining of the outer face of the graphite plate is not required. That is, flat graphite plates are able to be used. This also reduces the material cost of the cathodes plates 40 and anode plates 45. Further, as the heat exchanger plates 30 are thinner than the grooves that are normally machined on the graphite plates in prior art devices, the overall length of the fuel cell stack 10 is reduced. Still further, the overall weight of the fuel cell stack 10 is reduced due to the weight of the heat exchanger plate 30 being less than the graphite used to create the channels.

Other advantages are created by the use of the metal, preferably aluminium or copper, foam heat exchanger plates 30 due to the fact that they replace the traditional cooling channels. The aluminium foam heat exchanger plates 30 conduct electricity and make efficient use of the graphite separate anode/cathode or bipolar plate surface area when compared to existing technology. The graphite plates that sandwich the aluminium foam heat exchanger plate 30 (or potentially the entire cell) can be machined as a block leading to greatly reduced contact resistances. This enables simpler sealing of the fuel cell stack 10. Further, contact resistance is reduced when compared to traditional design.

Experimental Data

A test rig 100 was produced and consists of an open air wind tunnel and a heating system, where the measurements are performed on both sides of the metal foam heat exchangers samples. FIG 3 shows the test rig 100. A Elmo-G 2BH1 400 vacuum pump 101 sucks air through a nozzle 102 (5% uncertainty). To determine the air mass flow rate the pressure over the nozzle is measured, using a 202IP Digitron manometer 103■(± 0 13 mbar). The mass flow rate is controlled by a calibrated gauge valve 104 (5% uncertainty), downsizing the cross-section area. The downstream and upstream air temperatures are measured by two Go-Temp! thermocouples 105 (± o.i °C), located in the middle of the channel height. Two Adixen ASD 2001 capacitance gauges 106 (± 0 ^, 33· mbar) are used to measure the pressure drop between two sides of the metal foam heat exchanger samples. To inject the heat and keep the graphite surface temperature constant three Type HAP 200 heating plates 107 (see Fig. 4) and Hillesheim GmbH controllers 108 are applied. To consider the temperature distribution through the graphite plates, more thermocouples 109 (± o.i ⁶ C) are located in the depth of graphite plates (see the square dot lines in Fig. 5). Temperatures are monitored and logged by a laptop 110, in which the cable connection is provided between thermocouples and the laptop. The test section 1 1 1 is perfectly designed to seal the channels connections and provide enough compressing load on samples to reduce the contact thermal resistances, see Fig. 3 and 5. An uncertainty analysis is performed and results in a 5% total measurement uncertainties.

Two type of metal foams were examined, characterized in Table 2, where PPI is number of pores per inch, is form drag, and K is the permeability of the foam. Aluminum foam plates 112 with 5 mm thickness were sandwiched between two high-strength lightweight carbon fiber plates (very promising candidate for PEMFC bipolar plates) 1 13, as shown in Fig. 4. To reduce the electrical and thermal contact resistances foam plates were bonded to the graphite plates at both sides.

Table 2. Properties of applied metal foam samples

# Porosity PPI f ^■ K ( ^* 10 ^" ' m")

1 0.95 20 0.093 1,3

2 0.93 40 0.089 0.61 Results and discussions

Heat transfer rate and pressure drop

PEM fuel cell stacks have been widely investigated and the results of common applied water-cooled heat exchangers have been considered. A comparison will be made between the design of a metal foam heat exchanger and water-cooled heat exchanger, as explained earlier.

According to FIG. 6 by increasing the mass flow are of air, the magnitude of removed heat from graphite plates augments. Considering both trends of applied metal foams (20 and 40 PPI), the increase of removed heat linearly occurs as it is expected. To compare the performance of metal foam heat exchanger with that of water-cooled heat exchanger, the range of removed heat for PEMFC stacks has been calculated between 5 kW/m ² to 10 kW/m ² . To dump the same amount of heat with metal foam heat exchanger, the required mass flow rate is about 0.01 kg/s in which the stream velocity in the channel is about 1 m/s for both types of considered metal foams. It is interesting to note that 20PPI metal foam has the potential to dump about 20 kW/m ² heat (two times higher than what PEMFC and micro PEMFC system requires) while the mass flow rate needs to be about 0.02 kg/s.

While the heat performance of metal foam heat exchanger shows a satisfactory trend, the pressure drop of that would be considered to estimate the total performance and compare it with conventional designs. As Fig. 7. illustrates, increasing the mass flow rate of air augments the pressure drop, and 40 PPI sample reasonably shows higher pressure drop as it makes more blockage effects, compared to 20 PPI sample. One should note that 0.01 kg/s mass flow rate to dump maximum 10 kW/m ² generates about 2 kPa pressure drop for 2,0 PPI sample and 5 kPa pressure drop for 40 PPI sample. Calculating pressure drop for a range of water channels, designed to cool down the fuel cell system, water-cooled heat exchangers generate 30-60 kPa (almost 6-30 times higher than that of metal foam heat exchangers)to dump the same magnitude of heat.

Calculating the pumping power of both heat exchangers, the water-cooled heat exchanger uses about 30-60W, however a metal foam heat exchanger. consumes about 17- 1W to remove the same amount of heat from graphite plates within a common PE FC system. Therefore, applying metal foam heat exchangers easily reduce the internal power consumption by 30% (this stimulates into a 30% increase in net power generation). Electrical conductivity and contact resistance

The innovation of the metal foam heat exchanger is partly linked with the fact that it replaces the cooling channels with an electrically conducting material,. The newly proposed design uses a thin metal foam plate (see FIG. 5) which conducts electricity and makes efficient use of the graphite plate surface area compared to the existing technology. More interestingly, the whole system (graphite plate, metal foam, and the next cell) can be machined as a block leading to extremely lower contact resistances. A measuring test was conducted to estimate the electrical and contact resistance to a sandwiched metal foam plate with two graphite plates. The electrical resistance of the each graphite plates was 90 micro ohms/m and interestingly the sandwiched metal foam had almost the same electrical resistance. This approves a perfect contact with minor electrical contact resistance. It should also be noted that the cooling channels, in the conventional design, not only disconnect the two electrically conducting plates but also lead to a very high contact resistance in the current practice. Nonetheless, implementing the new design, this drawback will be eliminated from the system. It is worth mentioning that the measured electrical conductivity of a 90% porous aluminum metal foam was one order of magnitude higher than that of graphite. Downsize systems

Being lightweight and having high strength puts metal foams in a better position for application in fuel cells following their successful application in aerospace industry where size and weight are among the most important parameters. The density of 90% porous Aluminum metal foams is almost 6 times less than that of graphite. Hence, a straight forward mathematical calculation shows that, even replacing half of the graphite used in a fuel cell stack by such metal foams leads to a 20% lighter fuel cell stack; a significant improvement vital for fuel cell vehicles. This could be more noticeable when one observes that graphite contributes to at least 80% of the total weight of a fuel cell stack. Also, being highly efficient in heat transfer, they require less area to transfer the same amount of heat compared to the existing technologies and significantly reduce the size and weight of fuel cell systems.

Another evidence of the novelty of the new design is that it removes the separate water loop from the water-cooled systems and thus leads to smaller fuel cell stacks. Compared to the existing air-cooled fuel cells, as mentioned earlier, the expected increase in the heat transfer rate is notable based on a rough-and-ready estimate (almost 10 times). Besides, the air that removes heat form the graphite plate can directly be used to add heat and humidity to the fuel so that a part of the waste heat is recovered without the need to a separate heat exchanger that increases the cost and heat transfer resistance.

It should be appreciated that various other changes and modifications may be made to the embodiment described without departing from the spirit or scope of the invention.