WO2006035210A1

WO2006035210A1 - Hydrogen generator for fuel cell

Info

Publication number: WO2006035210A1
Application number: PCT/GB2005/003698
Authority: WO
Inventors: Mark Christopher Turpin; Janet Marie Turpin
Original assignee: Voller Energy Limited
Priority date: 2004-09-30
Filing date: 2005-09-22
Publication date: 2006-04-06

Abstract

This invention relates to a device for the production of hydrogen from hydrocarbon fuels wherein the fuel is pyrolysed into carbon and hydrogen and subsequently the carbon is oxidised in a secondary reaction. The pyrolyser may be periodically cycled to an oxidising input to eliminate carbon build-up and generate fuel to heat the pyrolyser and also to provide water and/or space heating.

Description

Hydrogen Generator for Fuel Cell

Background

Fuel cells offer the potential to generate electricity directly from fuel gases with hitherto impossible quietness and efficiency. For simplicity of construction it is desirable to fuel cell manufacturers to develop systems that operate at low temperatures, this being particularly important for portable and low power systems. A number of low temperature fuel cell systems are produced commercially and these include proton exchange membrane fuel cells, alkaline fuel cells and phosphoric acid fuel cells. Each of these systems is preferably fuelled using pure hydrogen and is adversely affected if the hydrogen contains oxides of carbon.

The proton exchange membrane fuel cell utilises a proton exchange membrane closely coupled to a catalyst that is able to dissociate hydrogen fuel into protons and electrons. This type of fuel cell represents one of the simplest fuel cell systems to manufacture and operate but is beset by the problems associated with delivering the hydrogen fuel. Hydrogen is a permanent gas that is not amenable to liquefaction other than under extreme cryogenic conditions. This factor makes delivery of hydrogen as a fuel problematic, as the energy density of compressed gas, stored in heavy metal or composite cylinders, is untenably low. There is clear commercial need for a system to allow hydrogen to be generated from commercially available fuels such as propane, butane, petroleum spirit and paraffin using a simple, low cost apparatus. Moreover, it is highly desirable that the system should deliver hydrogen free of contamination by carbon monoxide or carbon dioxide both of which gases have a deleterious effect on fuel cell operation. Nomenclature

Throughout this application the process of thermolytic cracking of a hydrocarbon feedstock into carbon and hydrogen is referred to as pyrolysis. The device for achieving such reaction is referred to as a pyrolyser. For the avoidance of doubt pyrolysis may be taken, within the context of this specification, to be synonymous with 'thermolytic cracking', 'thermolysis' and 'cracking'. Prior art

The problem of liberating hydrogen from a hydrocarbon fuel has been the subject of considerable research. The approach usually adopted is the use of a reformer system wherein the hydrocarbon is partially oxidised. In such systems the fuel gas is mixed with an oxidant, typically air or steam, and is passed over a heated catalyst. The object of this process is to render the fuel into a form wherein the carbon content is converted to carbon monoxide and the hydrogen content of the fuel is liberated as either hydrogen or water vapour. The reformate, thus produced, is cooled and then passed through a separate catalytic reactor to perform a so-called water gas shift reaction. In this process the carbon monoxide is reacted with water vapour to produce carbon dioxide and hydrogen. The latter reaction is significantly exothermic and is adversely driven to reactants by high temperature. It is therefore common for the shift reactor to have multiple stages with cooling between each stage. This can make the shift reactor expensive to produce and control. Once the carbon monoxide has been safely converted to carbon dioxide it is necessary to remove the carbon dioxide via some purification system. Typically oxidative and steam reformers employ a pressure swing apparatus (PSA) to effect such purification A schematic of such a system is included in Figure 1 and a chemical representation of the processes is included below. C_nH_n+2 +^ O₂ ^Catal^^mP > nCO + (0.5n + l) H₂

Partial oxidation reformer reaction

C_nH_n+2 + nH₂O ^^y^^p > nCO + (1.5n + l) H₂

Steam reformer reaction _{co + H2}Q Catalyst/Temp _> ^ ₊ p^ gtøβ _reaction

An example of a steam reforming (SR) apparatus is to be found in US Pat. No. 5,932,181 to Kim et al. The apparatus is designed to produce a continuous supply of high purity hydrogen from natural gas and comprises a desulphurisation unit to removed the sulphur bearing odorant from the gas, a steam reformer to liberate the hydrogen from the incoming gas feed, a water gas shift reactor to oxidise carbon monoxide to carbon dioxide and a PSA to effect final purification. An example of a catalytic partial oxidation reactor (CPOX) is to be found in US patent No.6,221,280 to Anumakonda et al. In this system a heavy hydrocarbon, having in excess of 6 carbon atoms, is volatilised and mixed with air before contacting a rhodium/alumina catalyst maintained at 1050°C. The hydrocarbon is partially oxidised to a mixture of hydrogen and carbon monoxide whilst any sulphur containing materials in the fuel are converted to hydrogen sulphide. As previously the product gas must be purified to remove all traces of sulphur and carbon monoxide before it is suitable for use in a low temperature fuel cell.

An example of a fuel processor that combines both steam reforming and partial oxidation is described in US Patent No 6,555,259 to Carpenter et al. In this system a fuel is mixed with air and steam before contacting a rhodium/zirconia catalyst. The advantage of such a system is that the SR reaction is endothermic whilst the POX reaction is exothermic. Control of the relative intensity of each reaction allows the processor to operate autothermally (self control of temperature). This class of fuel processor has become widely known as an autothermal processor (AP). It can be appreciated from the above examples that the principle difficulty with existing fuel processors for the production of hydrogen for fuel cells is that all produce a mixture of various gases some of which are unsuitable for introduction to a fuel cell system. Particularly problematic are carbon oxides which are damaging to the performance and endurance of the cells. Carbon dioxide is a significant problem within alkaline fuel cells, wherein it reacts with the electrolyte to create a precipitate. In other cell systems carbon dioxide is less damaging, but can indirectly cause catalyst poisoning due to carbon monoxide, which can be produced via a reverse shift reaction occurring within the cell. Carbon monoxide is a much more significant threat to fuel cell operability, as it is a prolific poison for the catalysts used to dissociate hydrogen within the cell. It is therefore necessary to ensure that the output from the reformer system is below 50ppm carbon monoxide and is preferably below 5ppm carbon monoxide. Fuel processor systems capable of meeting this purity target are currently i) bulky due to the need for a multi-stage water gas shift reactor and pressure swing apparatus, ii) expensive due to the use of platinum group metal catalysts, iii) unreliable due to the tendency of the fuel cell catalyst to perform a reverse shift reaction in the presence of trace amounts of carbon dioxide. In addition, the water gas shift reaction has an equilibrium condition favouring low temperature, that is to say the lower the temperature the lower the concentration of carbon monoxide in the gas produced. The low temperature required to achieve satisfactory gas quality means that the rate of the shift reaction is slow and the reactor must be large to allow the gas sufficient catalyst contact time. The optimisation between reactor size/weight, operational temperature and gas quality is complex, requiring complex and expensive control facilities.

As a result of these issues it is fair to say that the production of hydrogen for fuel cells has become one of the most significant impediments to the uptake of this technology. An alternative route to the generation of hydrogen from hydrocarbon fuels is via a pyrolysis reaction in a pyrolyser. A pyrolyser is an entropy driven device that breaks down the fuel at high temperature into its elemental constituents, in the case of a hydrocarbon fuel, hydrogen and carbon. A pyrolyser can typically be used to generate hydrogen from any hydrocarbon fuel that when reduced to its elemental composition will generate more molecules of hydrogen than there were molecules of the fuel. Examples of such fuels are methane, ethane, propane, butane, petroleum spirit and paraffin.

The pyrolysis reaction is generally endothermic (requires external energy) but is driven at high temperatures by the increased entropy of the gases produced. In general, this gives pyrolysers a unique advantage over oxidative reformers in that the entropy change is maximum when the fuel is of high molecular weight and saturated. This means that the system is extremely beneficial when used with high carbon alkanes typical of those found in paraffin, petroleum, butane and propane. For comparative purposes the entropy change during the pyrolysis of methane is +82 J.molΛK^'1, whilst for butane the equivalent value is +369 J.mol^'^K^"1.

The use of a simple pyrolyser for hydrogen production has been described in US Patent Nos. 6,653,005 and 6,670,058 to Muradov. In 6,653,005 a system is described wherein the hydrocarbon gas is decomposed over a carbon or metal catalyst to yield a deposited carbon material and a hydrogen rich exhaust gas. The problem with the system described is that excessive maintenance would be required to remove the build up of carbon and maintain the pyrolyser in an operational mode. In US Patent No 6670058 a system is described wherein the pyrolysis of a hydrocarbon fuel is conducted in the presence of a carbon material, preferably activated carbon. The process is used for the continuous production of hydrogen and carbon, the latter being separated, activated and partially recycled to the reaction chamber. International Patent WOO 1/20703 to Manikowski and Noland describes a pyrolyser system having two pyrolysis chambers. One chamber is used for active pyrolysis whilst the other is oxidised using air to yield two output streams, one of which is hydrogen rich and is used to power a fuel cell and the other of which is carbon monoxide rich and is used to fuel a heat engine and generator. After some period of operation the gas feeds to the two pyrolysers are reversed.

The problem with WOO 1/20703 is that the period of switch-over can cause carbon monoxide contamination to enter the hydrogen feed for the fuel cell and such contamination will, over time, result in the failure of the fuel cell system. Each of the systems described previously for hydrogen production face difficulties when applied to low power or portable systems. The SR and POX systems are in general too large and heavy for this type of application, whilst those systems which have been described as operating via the pyrolysis of carbonaceous fuels do not adequately deal with the necessity for continuous operation that is a clear requirement of many fuel cell applications. The aim of this invention is:- To provide a means of supplying hydrogen from commonly available hydrocarbon fuels, of a quality suitable for low temperature fuel cell systems, particularly but not exclusively PEM fuel cell systems.

To provide a low capital cost plant for achieving the above.

To provide a hydrogen supply of sufficient compactness as to be suitable for low power and portable fuel cell systems in the range 100-5000 Watts.

Reciprocating Pyrolyser

The approach adopted in this invention is the use of a pyrolyser system having a reciprocating gas supply, whereby the pyrolyser operation is separated into three distinct operations. The pyrolyser is initially supplied with carbonaceous gas to produce relatively pure hydrogen contaminated with light hydrocarbons. This gas is satisfactory for the operation of a fuel cell as the light hydrocarbons do not interact with the fuel cell system.

After some period of operation the pyrolyser becomes clogged by carbon deposits and the pressure required to force gas through the pyrolysis chamber begins to increase.

At this point the pyrolyser input is switched to an oxidising gas stream, typically air or steam. The output from the pyrolyser in this mode is either carbon monoxide or water gas depending on the oxidant used, and is now switched away from the fuel cell and stored for use as a combustion fuel. For most applications the use of water as oxidising agent is highly preferential. However, for applications where relatively pure carbon monoxide/nitrogen mixture is useful, the use of air can simplify the system design. The fuel gas produced in this way is recycled to heat the pyrolyser and to supply heat for hot water or space heating.

Once the carbon deposit within the pyrolyser is exhausted, the inlet valve is cycled to supply a purge gas, which may simply be the fuel gas. The outlet from the pyrolyser remains directed towards the combustion fuel storage tank until all of the oxide species within the pyrolyser have been removed. After the pyrolyser has been purged in this way the outlet valve is cycled back to supplying the fuel cell.

Preferred Embodiments

This invention will now be described by reference to 2 preferred embodiments, one comprising the operation of a single pyrolysis chamber and one comprising a more complex system comprising 3 pyrolysis chambers coupled to auxiliary systems designed to supply heat, hot water and electrical power. The purpose of the preferred embodiments is to allow detailed description of the operation and utility of the invention and is in no way intended to limit the invention to those systems herein described.

Preferred Embodiment Example 1 - Single Pyrolyser

The pyrolyser of this invention is shown in Figure 2. The pyrolyser consists of an

enclosed tubular chamber (1) maintained at 600-1200°C in a gas fired oven (2),

utilising a catalysed fibre burner (6). Gases can be passed through the pyrolyser via inlet (3) and outlet (4) pipework. The chamber is filled with a support which may be a catalytic support (5) that encourages the deposition of carbon via the pyrolysis reaction, equation 1, which is illustrative for the fate of an alkane fuel but is not intended to limit the utility of the current invention to alkane fuels.

_QJi_2n+2 High Temperature _> ^ ₊ tø + J)!^ Equation 1

Because of the high temperature of use, the containment of the pyrolyser was produced from a high temperature nickel alloy (Inconel 901) although the choice of material depends markedly on the temperature of operation of the device. The internal support material can be produced from a suitable oxide refractory that is stable to hydrogen at the operating temperature and can be doped with a catalyst such as nickel or chromium. Alternatively, supports made from sintered coarse metal powders, open celled metal foams or wire wool are all equally successful. Indeed any support may be used that is stable at the operational temperature, promotes the deposition of carbon and is stable in both oxidising and reducing atmospheres.

This embodiment employs a single pyrolyser chamber having a reciprocating gas feed and outfall. The design is shown schematically in figure 3. The system comprises a single pyrolyser tube contained within a high temperature oven. The tube has an internal construction consistent with Figure 2 and is heated by means of a burner arrangement that is preferably of a catalysed ceramic fibre type such as that described in US4,491,640, and widely used in portable domestic heating appliances designed to run on propane as described in US4,177,168. The inlet to the pyrolyser is equipped with servo valves to allow the admittance of hydrocarbon fuel gas via V3 or an oxidant. The inlet oxidant is preferably water supplied from a precision piston pump, Pl. The outlet from the pyrolyser is equipped with a heat exchanger to cool the outflow gases and servo valves to allow the output to be directed either to a heating gas tank, T2, or to a fuel cell gas tank, Tl, where both .tanks are preferably of the sliding cylinder gasometer type. The pyrolyser chamber is equipped with a water cooled heat exchanger coil, E3, to allow the amount of reject heat extracted from the chamber to be controlled. Flow sensors Fl, F2 and F3 allow the ratio of incoming gas volume to outflow gas volume to be monitored by an external mechanism incorporating a control algorithm such as a personal computer. Mass flow controllers MFl and MF2 allow the stoichiometry of gas supplied to external fuel cell systems to be controlled based on the current demand. The size of the pyrolyser and the gas storage tanks need to be matched to the electrical demand of the fuel cell. This is best determined experimentally as it depends markedly on the density and reactivity of the support used within the pyrolyser. For optimal operation it is desirable but not essential that the cycle time of the system should be longer than 15 minutes but less than 120 minutes. At long cycle times the volumes of gases that must be stored become large whilst at short cycle times contamination of the fuel cell feed gas with carbon monoxide can become a problem due to ineffective purging.

For simplicity the description of the device operation will be specific to the use of propane as a fuel, although this should be interpreted to be representative of any hydrocarbon fuel.

For simplicity the description of the device operation will be specific to the use of water/steam as the oxidant, although this should be interpreted to be representative of any suitable oxidant.

For simplicity the description of the device operation will be specific to the use of propane as the purge gas, although this should be interpreted to be representative of any gas not containing oxygen, either as free oxygen or in compound form.

The operational cycle of the device will now be described.

On start-up the pyrolyser is heated to an operational temperature in the range 600-

1200⁰C using propane supplied via valve Vl and control restriction Rl. The pyrolyser heater can utilise any burner design, however, as the burner needs to burn very different fuel gases in different modes of operation, the use of a catalysed ceramic fibre burner is highly beneficial.

Once the pyrolyser reaches operational temperature, pump P2 is started and valves V3 and V4 are opened. After a short period of operation, associated with the purging of air and moisture from the pyrolyser, valve V5 is opened and V4 closed. The product gas obtained from the pyrolyser is rich in hydrogen, being typically 80 vol% hydrogen with a balance of light hydrocarbons, largely methane. The gas is cooled in a heat exchanger and stored in the hydrogen gas store Tl. Tl is preferably a sliding partition gasometer such that the pressure of the gas contained is largely independent of the volume stored. Alternatively, a compressor can be used to compress the product gas into a high pressure containment of lower volume.

Once gas is available in Tl the fuel cell can be used to generate electrical energy. The gas from the hydrogen store is passed through the fuel cell at a stoichiometry¹ above that normally used in fuel cell systems. For this apparatus a stoichiometry in the range 1.4 to 2 is desirable. Because of the need to control the flow of gas to the fuel cell system in ratio to the current, it is preferable (but not essential) to employ some form of electronic mass flow control apparatus MFl. Excess fuel gas from the fuel cell is recycled to the catalytic burner via a non-return valve V8.

During operation the hot water and space heating functions run continuously via pump P2. The hot water system removes heat from the gases exiting the pyrolyser in heat exchanges El and E2 and also removes heat directly from the pyrolyser chamber via E3. The flow of water in E3 is controlled via a restrictor valve R2 which is primarily used to control the hot water cistern temperature. However, the control system must also be cognisant of the amount of stored water gas in the system and will close R2 to force water through the pyrolysis chamber and reject heat through the - forced convection space heater Sl in the event that T2 approaches a full condition. By this mechanism the rate of combustion of gas from T2 will be increased to maintain the pyrolyser temperature.

¹ Stoichiometry is used here in the sense of the ratio hydrogen delivered to the current drawn As pyrolysis proceeds the pyrolyser becomes clogged with carbon and the flow rate of fuel measured via Fl declines. At some preset flow rate value the operation of the pyrolyser switches and V3 and V5 close, whilst V4 opens and water is slowly injected into the pyrolyser chamber via the piston pump Pl . The water forms steam within the pyrolyser and converts the deposited carbon to water gas. The product gas is cooled via heat exchanger E2 and stored in the gasometer T2. During this phase of operation the proton exchange membrane fuel cell Cl continues to produce electricity from hydrogen stored in Tl. However, if desirable the system can incorporate a second fuel cell, C2, of a CO tolerant type (solid oxide fuel cell for example) to run from the gas in T2. Alternatively the gas in T2 is transferred via a non-return valve to the catalytic burner that heats the pyrolyser.

The control of the water input into the pyrolyser has multiple facets. Firstly the delivery should be slow and controlled with constant feedback from the pressure at the injector nozzle downstream from Pl. The use of a multi-cylinder piston pump adapted from a pressure washer proved a satisfactory method to achieve this. Secondly to determine the point at which all of the carbon has been consumed requires an algorithm to monitor water input and gas output. When carbon is available each incoming mole of water will produce 1 mole of carbon monoxide and 1 mole of hydrogen, hence 18 grammes of water will produce approximately 50 litres of water gas. As the carbon is exhausted the incoming water will simply volatilise in the pyrolyser and re-condense in the heat exchanger and the high volumes of outflow gas will no longer be observed. By monitoring outflow volume as a function of input water, the point at which all of the carbon has been consumed can be readily identified. Once all of the carbon is consumed Pl is shut down and V3 opened and the cycle returned to the starting configuration.

It can be appreciated from the description of the operation of the pyrolyser that the pyrolyser operates in 3 distinct modes.

1. Pyroly sis. In which carbon is deposited and hydrogen generated

2. Oxidation. In which the carbon is oxidised to carbon monoxide

3. Purge. A short period in which the pyrolyser is returned to pyrolysis but the outfall gases are contaminated by carbon monoxide and are therefore routed via the heating gas storage facility, T2.

The stages for the pyrolysis of propane with oxidation using steam are included in Table 1. In this mode of operation the overall reaction performed by the pyrolyser is the same as that performed by a steam reformer. However, the reaction is conducted step-wise to yield predominantly hydrogen fuel stream for the fuel cell and a high carbon monoxide fuel stream for combustion.

Table 1 The processes occurring during the 3 cycles of the pyrolyser Preferred Embodiment Example 2 - Triple Pyrolyser

It can be appreciated from the foregoing description that the pyrolyser system of this invention is ideally operated in a cyclical manner having three modes of operation. An alternative mode of operation is the use of three separate pyrolysers, such that each pyrolyser, at any one time, is performing one of the three operational modes. In this way a device is constructed which eliminates the intermittent gas delivery which is a problem with preferred embodiment example 1.

This embodiment of the invention will be described by reference to the schematic diagram in figure 4. It should be noted that this embodiment is for example only, and is not intended to be restrictive in any way.

For the purpose of this description ganged valves are groups of valves that are all opened or closed by a common operation. Ganged valves may be physically ganged whereby one servo unit operates all of the valves or functionally ganged whereby the ganged valves are separate but have a common control.

The core of this preferred embodiment is a pyrolysis oven containing three of the pyrolysers described earlier. The pyrolysers are connected to the external environment via three way manifolds leading to servo valves ganged in groups of 3. The inlet and outlet valve gang arrangements are shown with more clarity in the schematic in Figure 5. The inlet valves are labelled VGl, VG2 and VG3, whilst the outlet valves are labelled VG4, VG5 and VG6 in Figures 4 and 5. Each inlet valve gang is supplied with propane from the fuel store (T3), hydrogen rich gas from the hydrogen store (Tl) and air from the compressor (P4). The outlet valve gangs are connected to a hydrogen store (Tl), a carbon monoxide store (T2), and a direct connection to the burner system within the pyrolyser chamber. Electronic connections to the valve gangs ensure that only one inlet and outlet gang are open at any one time.

Outfall gases from the pyrolyser pass through one of three water cooled heat exchangers (El, E2 and E3) before entering the valve gangs. Hot water generated by cooling the outfall gases is stored in the hot water tank (T4) and is also circulated through a space heater (Sl). Hot water is actively pumped through the outfall gas coolant circuit via a pump (Pl) and is supplied directly for external use via a pump (P2) leading to a faucet (Fl). The temperature of the coolant water is maintained below 5O⁰C by activation of the fan in Sl. An additional cooling circuit (E4) is provided within the pyrolyser chamber and can be controlled via a restrictor valve (R3) to allow the extraction of excess heat from the chamber.

The hydrogen store (Tl) is connected via a small compressor (P3) to a low temperature fuel cell system (Cl) such as a proton exchange membrane fuel cell. The compressor circulates the hydrogen rich gas from the hydrogen store (Tl), through the fuel cell (Cl) and back to the store (Tl). Dependent upon the electrical demand, a variable proportion of the return gas is diverted using a diverter valve (V 8) to the fuel store (T3) and is recycled through the pyrolyser. A dryer (Dl) is incorporated into the recycle line to eliminate any water from the gas passed to the fuel store. A secondary supply of hydrogen rich gas is drawn from the hydrogen store (Tl) via a restrictor valve (Rl) and is used as the gas supply for the purge cycle. This gas is also dried via an in line drier (D2).

The carbon monoxide storage tank (T2) stores carbon monoxide from the oxidation of carbon deposits within the pyrolyser. The carbon-monoxide/Nitrogen mix that is the outcome of air oxidation is used as the principal heating fuel for the pyrolysis chamber. It should be noted that the use of air as the oxidant is illustrative and should not be interpreted as restrictive. The oxidant input to the system could equally easily be steam and the carbon monoxide store would instead fill with water gas. Either fuel can be used to heat the pyrolyser. Empirically it was found necessary to incorporate a small compressor (P5) into the inlet to T2 in order to ensure that the store gas pressure was suitable for the burner design utilised within the pyrolyser. This component may not be necessary for other designs of burner.

The pyrolyser heater can be of any design capable of burning both carbon monoxide

(approx 30% in nitrogen) and propane. However, the use of a catalytic burner such as has been described previously is found most beneficial.

Each pyrolyser in the system is equipped with an electronic mass flow controller to allow monitoring and control of gas flow into the pyrolyser. Each pyrolyser is also equipped with a zirconia based oxygen sensor on the outfall to determine the point at which oxygen breakthrough occurs during the oxidation cycle. The use of a high temperature sensor, of the type used in automotive exhaust systems, has been found beneficial.

The operation of the system from a cold start will now be described.

Start-up

AU valves are initially closed. V7 is opened and propane flows to the burner system to heat the pyrolyser chamber. The burner power output is controlled via the restrictor valve R2. The pyrolyser tubes are sealed during start-up and so must be able to either withstand 5 bar internal pressure at 1200°C or be equipped with a ήon-return vent somewhere on the cool part of the pipe work. Once the pyrolyser reaches a preset temperature in the range 600-1200°Cthe pyrolysis process is able to start. Running

Valve V7 is closed and valve gangs VGl and VG4 are set to open. In this condition pyrolyser 1 is passing air, pyrolyser 2 is passing propane and pyrolyser 3 is passing hydrogen rich gas from the hydrogen store, Tl. Carbon within pyrolyser 1 (residual from an earlier cycle) is oxidised to carbon monoxide and the output is routed via VG4 to the carbon monoxide store (T2). Pyrolyser 2 is pyrolysing the incoming propane to deposit carbon and then passing the output hydrogen rich gas via VG4 to the hydrogen store (Tl). Pyrolyser 3 is passing a hydrogen rich gas at a considerably reduced flow rate when compared to pyrolysers 1 and 2; this gas is designed to purge all oxidising species from pyrolyser 3. The outfall from pyrolyser 3 is routed via VG4 to the pyrolysis chamber burner. The gas flows into pyrolysers 1 and 2 are matched using a computerised algorithm to control the mass flow controllers Fl and F2 such that the rate of carbon oxidation is equal to the rate of carbon deposition. After some period of operation the carbon build-up in pyrolyser 2 becomes significant and the carbon availability in pyrolyser 1 becomes depleted. This point is conveniently detected using the oxygen sensor, Sl, to detect oxygen breakthrough in pyrolyser 1. At this point, the first phase of the first cycle is terminated by the closure of VGl and VG4. The second phase is initiated by opening VG2 and VG5. Pyrolyser 1 is now purged whilst pyrolyser 2 switches to oxidation and pyrolyser 3 starts active pyrolysis. The outfall gases are routed via VG5 to the correct destinations. The second phase continues, as before, until the oxygen sensor, S2, on pyrolyser 2 registers oxygen breakthrough, at which point the sequence moves on to phase three via the closure of VG2 and VG5 followed by the opening of VG3 and VG6. Pyrolyser 1 is now in pyrolysis mode, pyrolyser 2 is purged and pyrolyser 3 is oxidised. When the oxygen sensor on pyrolyser 3, S3, indicates an oxygen breakthrough the system finishes the first cycle by the closure of VG3 and VG6 and the cycle restarts with the opening of VGl and VG4. It can be appreciated that the valve sequencing is such that each pyrolyser is subjected to a purge followed by a period of pyrolysis followed by a period of oxidation. The outfalls from the pyrolysers are similarly sequenced such that the pyrolysis product, which is rich in hydrogen, is passed to the hydrogen store; the oxidation product, which is rich in carbon monoxide, is passed to the carbon monoxide store; whilst the purge outfall is burned in the burner system. Once the system has started all of the fuel gas for the pyrolysis chamber is drawn from the carbon monoxide store and the purge line, which together have a significant surplus of calorific energy relative to the thermal requirement to heat the chamber. The restrictor valve R2 is not primarily used to control the temperature in the chamber but rather to ensure that a constant volume of gas is maintained in the carbon monoxide store. Temperature control of the chamber is achieved via the cooling circuit E4 the flow within which is controlled by the restrictor valve R3. Reject heat from E4 and also from the outfall gas heat exchangers El, E2 and E3 is used to heat the water cistern T4, this can supply hot water to a faucet and space heating appliance.

The hydrogen rich gas produced by the system is typically of approximately 80 vol% hydrogen with a balance of light hydrocarbons. The latter are benign to the operation of a fuel cell and this gas is ideally suited to use as fuel for a PEM fuel cell. Hydrogen rich gas is recirculated through the fuel cell via a compressor (P3). In order to utilise residual hydrocarbon in the gas from the hydrogen store (Tl), a proportion of the recycle is drawn back to the incoming fuel store (T3) via a diverter valve (V8). The proportion of gas diverted in this way is controlled by reference to the current demand on the fuel cell system. At high current demand a large proportion of the hydrogen in the fuel cell feed gas is consumed electrochemically and in consequence all of the recycle gas is passed to the fuel store (T3). To the contrary, when current demand is low the use of hydrogen is similarly low and up to 90% of the recycle can be fed back to the hydrogen store (Tl) with just 10% being transferred to the fuel store (T3). The algorithm used to control the recycle valve (V8) can be adapted to maintain the concentration of hydrogen in the hydrogen store (Tl) at a user selected level. The storage tanks used in this embodiment of the invention can be much smaller than those used in preferred embodiment example 1. As the system is continually manufacturing hydrogen, carbon monoxide and electricity the tanks purpose, rather than being a storage facility, can be to offer a buffer to provide supply during valve switches. This factor allows this preferred embodiment to offer a substantially smaller plant for the delivery of a given volume of hydrogen.

Claims

ClaimsI claim

1. A device for the production of hydrogen from hydrocarbon fuels wherein the fuel is pyrolysed into carbon and hydrogen and subsequently the carbon oxidised in a secondary reaction.

2. An invention as in claim 1 whereby the pyrolyser is periodically cycled to an oxidising input to eliminate carbon build-up and generate fuel used to heat the pyrolyser and also to provide water and/or space heating.

3. A device as in claims 1 and 2 whereby subsequent to oxidation the pyrolyser is purged with a non oxidising gas, prior to the start of the next pyrolysis cycle, to prevent contamination of the hydrogen feed with oxidising species.

4. A device as in claims 1 to 3 wherein the pyrolyser operates in three modes, being pyrolysis of the fuel to carbon and hydrogen, oxidation of the carbon deposit and purging of the pyrolyser with an oxygen free gas ready for the next pyrolysis cycle.

5. A device as in any of claims 1 to 4 whereby the system comprises at least 3 pyrolysers such that pyrolysis, oxidation and purging are conducted simultaneously on different pyrolysers to give continuous gas outputs.

6. A device as described in any of claims 1 to 5 wherein the oxidising species is water or steam and the heating fuel output is a mixture of hydrogen and carbon monoxide commonly known as 'water gas'

7. A device as described in any of claims 1 to 5 wherein the oxidising species is air and the heating fuel output is a mixture of carbon monoxide and nitrogen.

8. A device as described in any of claims 1 to 7 whereby the device delivers hydrogen uncontaminated by oxides of carbon and suitable for use in a low temperature fuel cell without further purification.

9. A device as described in claim 8 wherein the device is coupled to a proton exchange membrane fuel cell

10. A device as described in claim 8 wherein the device is coupled to an alkaline fuel cell

11. A device as described in claim 8 wherein the device is coupled to a phosphoric acid fuel cell

12. A device as described in claim 8 wherein the device is coupled to a molten carbonate fuel cell

13. A device as described in any of claims 1 to 12 wherein the gas supplies and outputs are reciprocated automatically according to a predefined control algorithm which uses as inputs data gathered from mass flow controllers on the pyrolyser inputs and oxygen sensors on the pyrolyser outfalls.

14. A device as described in any of claims 1 to 13 wherein the pyrolyser is heated using a catalytic combustor capable of handling widely varying fuels

15. A device as described in any of claims 1 to 14 configured to deliver electricity, hot water and space heating outputs.

16. A device as described in any of claims 1 to 15 wherein the system is configured with two fuel cell systems; one fuel cell running on hydrogen and being preferably a proton exchange membrane fuel cell, and one fuel cell which is highly tolerant of carbon monoxide and running on the water gas or carbon monoxide/nitrogen mixture produced by the apparatus.