52 Fuel Reforming

5.2.1 Fuel cell requirements

Fuel reforming is the process of taking the delivered fuel, such as gasoline or propane, and converting it to a form suitable for the PEM fuel cell. This will never involve simply converting it to pure hydrogen, there will always be other substances present, particular carbon compounds.

A particular problem with fuel reformers and PEM fuel cells is the presence of carbon monoxide. This has very severe consequences for this type of fuel cell. It 'poisons' the catalyst on the electrode, and its concentration must be kept lower than about 10 parts per million. Carbon dioxide will always be present in the output of a reformer, and this poses no particular problems, except that it dilutes the fuel gas, and slightly reduces the output voltage. Steam will also be present, but as we have seen in the last chapter, this is advantageous for PEM fuel cells.

There is a very important problem that the presence of carbon dioxide in the fuel gas imposes on a fuel cell system. This is that it becomes impossible to use absolutely all of the hydrogen in the fuel cell. If the hydrogen is pure, then it can be simply connected to a fuel cell, and it will be drawn into the cell as needed. Nothing need ever come out of the fuel side of the system. When the fuel gas is impure, then it will need to be circulated through the system, with the hydrogen being used as it goes through, and with virtually 100% carbon dioxide gas at the exit. This makes for another feature of the cell that needs careful control. It also makes it important that there is still some hydrogen gas, even at the exit, otherwise the cells near the exit of the fuel flow path will not work well, as the hydrogen will be too dilute. This means that the systems described in this section will never have 100% fuel utilisation, some of it will always have to pass straight through the fuel cell stack.

5.2.2 Steam reforming

Steam reforming is a mature technology, practised industrially on a large scale for hydrogen production. The basic reforming reactions for methane and octane CsH18 are:

The reforming reactions (5.1) and (5.2), and the associated 'water-gas shift reaction' (5.3) are carried out normally over a supported nickel catalyst at elevated temperatures, typically above 500°C. Over a catalyst that is active for reactions (5.1) or (5.2), reaction (5.3) nearly always occurs as well. The combination of the two reactions taking place means that the overall product gas is a mixture of carbon monoxide, carbon dioxide and hydrogen, together with unconverted fuel and steam. The actual composition of the product from the reformer is then governed by the temperature of the reactor (actually the outlet temperature), the operating pressure, the composition of the fuel, and the proportion of steam fed to the reactor. Graphs and computer models using thermodynamic data are available to determine the composition of the equilibrium product gas for different operating conditions. Figure 5.3 is an example, showing the composition of the output at 1 bar, with methane as the fuel.

It can be seen that in the case of reaction (5.1), three molecules of carbon monoxide and one molecule of hydrogen are produced for every molecule of methane reacted. Le Chatelier's principle therefore tells us that the equilibrium will be moved to the right (i.e. in favour of hydrogen) if the pressure in the reactor is kept low. Increasing the pressure will favour formation of methane, since moving to the left of the equilibrium reduces the number of molecules.


Figure 5.3 Equilibrium concentration of steam reformation reactant gases as a function of temperature. Note that at the temperature for optimum hydrogen production, considerably quantities of carbon monoxide are also produced


Figure 5.3 Equilibrium concentration of steam reformation reactant gases as a function of temperature. Note that at the temperature for optimum hydrogen production, considerably quantities of carbon monoxide are also produced

Another feature of reactions (5.1) and (5.2) is that they are usually endothermic which means that heat needs to be supplied to the reaction to drive it forward to produce hydrogen and carbon monoxide. Higher temperatures (up to 700°C) therefore favour hydrogen formation, as shown in Figure 5.3.

It is important to note at this stage that although the shift reaction (5.3) does occur at the same time as steam reforming, at the high temperatures needed for hydrogen generation, the equilibrium point for the reaction is well to the left of the equation. The result is that by no means all the carbon monoxide will be converted to carbon dioxide. For fuel cell systems that require low levels of CO, further processing will be required. These reactions are the basis of the great majority of industrial hydrogen production, using natural gas (mainly methane) as the fuel.

Hydrocarbons such as methane are not the only fuels suitable for steam reforming. Alcohols will also react in a steam reforming reaction, for example methanol:

CH3OH + H2O-> 3H2 + CO2 [AH = 49.7kJmol-1] (5.4)

The mildly endothermic steam reforming of methanol is one of the reasons why methanol is finding favour with vehicle manufacturers as a possible fuel for fuel cell vehicles, a point which is considered further in Section 5.4.2 below. Little heat needs to be supplied to sustain the reaction, which will readily occur at modest temperatures (e.g. 250°C) over catalysts of mild activity such as copper supported on zinc oxide. Notice also that carbon monoxide does not feature as a principal product of methanol reforming. This makes methanol reformate particularly suited to PEM fuel cells, where carbon monoxide, even at the ppm level, can cause substantial losses in performance due to poisoning of the platinum catalyst. However, it is important to note that although carbon monoxide does not feature in reaction (5.4), this does not mean that it is not produced at all. The water gas shift reaction of (5.3) is reversible, and carbon monoxide is produced in small quantities. The result is that the carbon monoxide removal methods described below are still needed with a methanol reformer used with a PEM fuel cell.

5.2.3 Partial oxidation and autothermal reforming

As an alternative to steam reforming, methane and other hydrocarbons may be converted to hydrogen for fuel cells via partial oxidation (POX):

Partial oxidation can be carried out at high temperatures (typically 1200 to 1500° C) without a catalyst, but this is not practical in small mobile systems. If the temperature is reduced, and a catalyst employed then the process becomes known as Catalytic Partial Oxidation (CPO). Catalysts for CPO tend to be supported platinum-metal or nickel based.

It should be noted that reactions (5.5) and (5.6) produce less hydrogen per molecule of fuel than reactions (5.1) or (5.2). This means that partial oxidation (either non-catalytic or catalysed) is less efficient than steam reforming for fuel cell applications. Another disadvantage of partial oxidation occurs when air is used to supply the oxygen. This results in a lowering of the partial pressure of hydrogen at the fuel cell, because of the presence of the nitrogen, which further dilutes the hydrogen fuel. This in turn results in a lowering of the cell voltage, again resulting in a lowering of system efficiency. To offset these negative aspects, a key advantage of partial oxidation is that it does not require steam.

Autothermal reforming is another commonly used term in fuel processing. This usually describes a process in which both steam and oxidant (oxygen, or more normally air) are fed with the fuel to a catalytic reactor. It can therefore be considered as a combination of POX and the steam reforming processes already described. The basic idea of autothermal reforming is that both the endothermic steam reforming reaction (5.1) or (5.2) and the exothermic POX reaction of (5.5) or (5.6) occur together, so that no heat needs to be supplied or removed from the system. However, there is some confusion in the literature between the terms partial oxidation and autothermal reforming. Joensen and Rostrup-Nielsen (2002) have published a review which explains the issues in some detail.

The advantages of autothermal reforming and CPO are that less steam is needed than with conventional reforming and that all of the heat for the reforming reaction is provided by partial combustion of the fuel. This means that no complex heat management engineering is required, resulting in a simple system design. This is particularly attractive for mobile applications.

5.2.4 Further fuel processing: carbon monoxide removal

A steam reformer reactor running on natural gas and operating at atmospheric pressure with an outlet temperature of 800°C produces a gas comprising some 75% hydrogen, 15% carbon monoxide and 10% carbon monoxide on a dry basis. For the PEM fuel cell the carbon monoxide content must be reduced to much lower levels. Similarly, even the product from a methanol reformer operating at about 200°C will have at least 0.1% carbon monoxide content, depending on pressure and water content. The problem of reducing the carbon monoxide content of reformed gas streams is thus very important.

We have seen that the water gas shift reaction:

takes place at the same time as the basic steam reforming reaction. However, the thermodynamics of the reaction are such that higher temperatures favour the production of carbon monoxide, and shift the equilibrium to the left. The first approach is thus to cool the product gas from the steam reformer and pass it through a reactor containing catalyst, which promotes the shift reaction. This has the effect of converting carbon monoxide into carbon dioxide. Depending on the reformate composition more than one shift reactor may be needed, and two reactors is the norm. Such systems will give a carbon monoxide concentration of about 2500-5000 ppm, which exceeds the limit for PEM fuel cells by a factor of about 100. It is similar to the CO content in the product from a methanol reformer.

For PEM fuel cells, further carbon monoxide removal is essential after the shift reactors. This is usually done in one of four ways.

In the selective oxidation reactor a small amount of air (typically around 2%) is added to the fuel stream, which then passes over a precious metal catalyst. This catalyst preferentially absorbs the carbon monoxide, rather than the hydrogen, where it reacts with the oxygen in the air. As well as the obvious problem of cost, these units need to be very carefully controlled. There is the presence of hydrogen, carbon monoxide and oxygen, at an elevated temperature, with a noble metal catalyst. Measures must be taken to ensure that an explosive mixture is not produced. This is a special problem in cases where the flowrate of the gas is highly variable, such as with a PEMFC on a vehicle.

The methanation of the carbon monoxide is an approach that reduces the danger of producing explosive gas mixtures. The reaction is the opposite of the steam reformation reaction of equation (5.1):

This method has the obvious disadvantage that hydrogen is being consumed, and so the efficiency is reduced. However, the quantities involved are small; we are reducing the carbon monoxide content from about 0.25%. The methane does not poison the fuel cell, but simply acts as a diluent. Catalysts are available which will promote this reaction so that at about 200°C the carbon monoxide levels will be less than 10 ppm. The catalysts will also ensure that any unconverted methanol is reacted to methane, hydrogen or carbon dioxide.

Palladium/platinum membranes can be used to separate and purify the hydrogen. This is a mature technology that has been used for many years to produce hydrogen of exceptional purity. However, these devices are expensive.

Pressure swing absorption (PSA): in this process, the reformer product gas passed into a reactor containing absorbent material. Hydrogen gas is preferentially absorbed on this material. After a set time the reactor is isolated and the feed gas is diverted into a parallel reactor. At this stage the first reactor is depressurised, allowing pure hydrogen to desorb from the material. The process is repeated and the two reactors are alternately pressurised and depressurised. This process can be made to work well, but adds considerably to the bulk, cost and control problems of the system.

Currently none of these systems has established itself as the preferred option. They have the common feature that they add considerably to the cost and complexity of the fuel processing systems.

5.2.5 Practical fuel processing for mobile applications

The special features of onboard fuel processors for mobile applications are that they need:

• to be capable of starting up quickly;

• to be able to follow demand rapidly and operate efficiently over a wide operating range;

• to be capable of delivering low-CO content gas to the PEM stack;

• to emit very low levels of pollutants.

Over the past few years, research and development of fuel processing for mobile applications, as well as small scale stationary applications, has mushroomed. Many organisations are developing proprietary technology, but almost all of them are based on the options outlined above, namely steam reforming, CPO, or autothermal reforming.

Companies such as Arthur D. Little have been developing reformers aimed at utilising gasoline type hydrocarbons (Teagan et al. 1998). The company felt that the adoption of gasoline as a fuel for FCVs would be likely to find favour amongst oil companies, since the present distribution systems can be used. Indeed Shell have demonstrated their own CPO technology on gasoline and ExxonMobil in collaboration with GM have also been developing a gasoline fuel processor. Arthur D. Little spun out its reformer development into Epyx which later teamed up with the Italian company De Nora, to form the fuel cell company Nuvera. In the Nuvera fuel processing system the required heat of reaction for the reforming is provided by in situ oxidising a fraction of the feedstock in a combustion (POX) zone. A nickel-based catalyst bed following the POX zone is the key to achieving full fuel conversion for high efficiency. The POX section operates at relatively high temperatures (1100-1500° C) whereas the catalytic reforming operates in the temperature range 800-1000°C. The separation of the POX and catalytic zones allows a relatively pure gas to enter the reformer, permitting the system to accommodate a variety of fuels. Shift reactors (high and low temperature) convert the product gas from the reformer so that the exit concentration of CO is less than 1%. As described earlier, an additional CO-removal stage is therefore needed to achieve the CO levels necessary for a PEM fuel cell. When designed for gasoline, the fuel processor also includes a compact desulphurisation bed integrated within the reactor vessel prior to the low temperature shift.

Johnson Matthey have demonstrated their HotSpot reactor on reformulated gasoline (Ellis et al. 2001). They built a 10 kW fuel processor which met their technical targets, but they also addressed issues relating to mass manufacture; their work has identified areas that will require further work to enable gasoline reforming to become a commercial reality. These included:

• hydrogen storage for start-up and transients;

• an intrinsically safe afterburner design with internal temperature control and heat exchange that can cope with transients;

• effect of additives on fuels;

• better understanding of the issues relating to sulphur removal from fuels at source;

• improved sulphur trapping and regeneration strategies.

Johnson Matthey are now engaged in a commercialisation programme for their technology. The pace of development is now such that in April 2001, GM demonstrated their own gasoline fuel processor in a Chevrolet 2-10 pickup truck, billed as the world's first gasoline-fed fuel cell electric vehicle. With the rapid developments being made in this area it remains to be seen which of the various fuel processing systems will become economically viable in the future.

One way to side-step all of the problems associated with onboard fuel processing is to make the fuel processing plant stationary, and to store the hydrogen produced, which can be loaded onto the mobile system as required. In fact, this is may well be the preferred option for some applications, such as buses. However, as ever, solving one problem creates others, and the problems of storing hydrogen are quite severe. These are dealt with in Sections 5.3 and 5.4 below.

Electric Car Craze

Electric Car Craze

Electric Cars Are A Savior From Pollution And Can Be A Laudable Investment For You. Find Out How Electric Car Conversion Shrinks Your Driving Expenses.

Get My Free Ebook

Post a comment