EP1423331A1

EP1423331A1 - Method and device for producing a gas rich in hydrogen by thermal pyrolysis of hydrocarbons

Info

Publication number: EP1423331A1
Application number: EP02787145A
Authority: EP
Inventors: Claude Etievant; Fanny Gaillard; Dominique De Lapparent; Karine Pointet
Original assignee: D'etudes Des Technologies De L'hydrogene (ceth) Cie; Compagnie Europeenne des Technologies de lHydrogene SA
Current assignee: Elogen SAS
Priority date: 2001-07-17
Filing date: 2002-07-09
Publication date: 2004-06-02
Also published as: CA2453841A1; US7537623B2; US20040265223A1; FR2827591B1; WO2003008328A1; AU2002333962A1; FR2827591A1

Abstract

The invention concerns a method for producing a gas rich in hydrogen by thermal pyrolysis of hydrocarbons which consists in carrying out, in a reactor (R) a catalyst-free thermal cracking to pyrolyze a fuel selected so as to produce either a gas rich in hydrogen and free of carbon monoxide, or a gas rich in hydrogen and containing carbon monoxide and in using said gas effluents during pyrolysis and inert with respect to the cell as fuel at the burner (B) to heat the reactor so as to bring it to a reaction temperature, and which consists, subsequently, in burning the powder carbon produced in the reactor (R) during the pyrolysis reaction either to produce carbon monoxide or to produce carbon dioxide. The invention is useful in particular for supplying hydrogen to fuel cells and for producing synthesis gas.

Description

METHOD AND DEVICE FOR PRODUCING A GAS RICH IN

HYDROGEN BY THERMAL PYROLYSIS

HYDROCARBONS

The invention relates to a process for the pyrolysis of hydrocarbons or oxygenated fuels (alcohols, ETB, MTB, ...) intended to produce a gas rich in hydrogen and possibly for certain applications of carbon monoxide (CO). It applies, in particular but not exclusively, to the conversion of a hydrocarbon into a hydrogen-rich gas intended to supply fuel cells (PAC) either at low temperature of the proton exchange membrane type (abbreviated as PEM for “Proton Exchange Membrane ”), either at high temperature of the molten carbonate type (abbreviated as MCFC for“ Molten Carbonate Fuel Cell ”) or of the solid oxide type (abbreviated as SOFC for“ Solid Oxide Fuel Cell ”).

The expression “lower calorific value” or PCI is defined below. The calorific value is defined as the quantity of heat released by the complete combustion of the fuel unit considered. The lower calorific value excludes from the heat given off, the heat of condensation of the water supposed to have remained in the vapor state after combustion.

In general, we know that fuel cells are electrochemical systems that directly convert chemical energy into electrical energy. The theoretical yield is very high and the by-products very polluting. In addition, PACs have demonstrated, originally in space missions, their qualities of efficiency, reliability and longevity. These qualities confirm the interest of heat pumps for the production of portable (a few hundred watts to a few kilowatts), fixed or on-board (2 kilowatts to 200 kilowatts) electricity. Heat pumps therefore represent a possible alternative to heat engines in many of their uses. They can also be used for the production of co-generation boilers.

The hydrogen fuel can come from a pressurized or cryogenic tank. However, given the safety problems inherent in the storage of hydrogen, it is wise to obtain it from a fuel (hydrocarbon or alcohol), the hydrogen of which will be released as and when required.

For this purpose, different chemical reactions can be implemented: steam reforming, partial oxidation and pyrolysis, each of these reactions can be activated thermally and / or using a catalyst.

The pyrolysis of hydrocarbons releases not only hydrogen but also carbon and other products whose calorific value is appreciable. It follows that the production of hydrogen by pyrolysis will be done with an interesting energy yield only if the energy content of carbon and other co-products is valued. This situation is different from that of steam reforming since this process releases the hydrogen contained in the fuel as well as the hydrogen contained in the water. The same is true for partial oxidation when it is coupled to a shift stage. The reactions are written:

Noreforming: C ₃ H ₈ + 6 H ₂ 0 → 3 C0 ₂ + 10 H ₂

Partial oxidation C ₃ H ₈ + 3/2 0 ₂ → 3 CO + 4 H ₂

+ shift: CO + H ₂ 0 → C0 ₂ + H ₂ (x 3)

C ₃ H ₈ + 3/2 0 ₂ + 3 H ₂ 0 → 3 C0 ₂ + 7 H ₂ Pyrolysis: C ₃ H ₈ → 3 C + 4 H ₂

However, the experiments carried out by exploiting these different processes show that the formation of carbon monoxide in a non-negligible amount is inevitable since oxygen is introduced either directly or in the form of water vapor.

Thus, reformers using partial oxidation or steam reforming as a process generally include a high or low temperature “shift” unit (CO recycling), a steam generator, and a drying unit to remove excess steam. CO conversion is therefore delicate, expensive and cumbersome.

The invention starts from the observation that pyrolysis overcomes these stages since it takes place in the absence of any source of oxygen thus preventing the formation of CO. In addition, the partial pressure of hydrogen in the gas formed is higher compared to other methods given, for example, the absence of nitrogen from the air used in partial oxidation. In order to exploit the advantages of the process linked to the absence of an oxygen molecule, the fuel of choice is a hydrocarbon (methane, propane, butane) or a mixture of hydrocarbons. On the other hand, an alcohol can be envisaged at the choice for the cases where the production of synthesis gas, mixture of H ₂ and CO, is required.

Different methods of using hydrocarbon pyrolysis reactions have been proposed by several authors.

Thus in US patent 5,899,175, Manikowski et al. propose a hybrid system composed of a catalytic pyrolysis reactor producing a hydrogen-rich gas and a mixture of combustible residues. The hydrogen-rich gas powers a fuel cell to generate electricity and mixture of combustible residues is burned in a combustion engine to produce mechanical power.

This process treats various liquid fuels derived from petroleum. These fuels can be linear or branched alkanes with at least five carbon atoms but also all types of commercial fuels such as gasoline, kerosene ...

The operating conditions chosen are such that only 20% of the hydrogen contained in the fuel is converted into the form of dihydrogen.

Poirier et al. propose a pyrolysis process carrying out the catalytic decomposition of natural gas into a gas rich in hydrogen and carbon (MG Poirier, C. Sapundzhiev, Catalytic decomposition of natural gas to hydrogen for fuel cell applications, Int. J. Hydrogen Energy, vol. 22, No. 4, 1997, 429-433). The authors suggest the use of hydrogen-rich gas to power a PEM fuel cell. The catalytic bed on which the carbon formed during the pyrolysis reaction is deposited is then regenerated by burning the carbon with air. In order to operate in continuous mode despite the alternating sequences of pyrolysis and regeneration of the catalytic bed, the authors propose a concept based on the use of two reactors operating alternately. The first operates in a pyrolysis regime to produce a hydrogen-rich gas while the second regenerates the catalytic bed by oxidation of carbon. The structure of the catalytic bed is designed to leave a dead volume sufficient to allow the accumulation of a large amount of carbon. Note that 45% of the lower calorific value (PCI) of natural gas remains in carbon. Furthermore, the pyrolysis of methane, the main constituent of natural gas, is endothermic and requires an input of approximately 12% of the PCI of natural gas. The authors therefore propose to use the energy released by the combustion of carbon to provide the heat necessary for the decomposition of natural gas. In addition, it should be mentioned that the catalytic bed generates parasitic side reactions. Indeed, a production of CO is observed during the pyrolysis phase while no oxygen supply is carried out. This emission probably comes from the partial reduction of oxides present in the catalyst and formed during the regeneration phase.

However, if one wants to supply a PEM type fuel cell with the hydrogen-rich gas produced by this system, it is necessary to eliminate the CO because more than 10 ppm of CO prevents the functioning of the cell, the anodic catalyst at base of platinum which it contains being very sensitive to this pollutant. This purification can be carried out by a catalytic methanation process. A methanation reactor must then be placed upstream of the PEM type heat pump on its hydrogen supply circuit.

A propane pyrolysis system very similar to that of Poirier et al. was proposed by the German team of Ledjeff-Hey. It differs from it mainly by the nature of the hydrocarbon treated. (K. Ledjeff-Hey, N. Formanski, Th. Kalk, J. Roes; Compact Hydrogen Production Systems for Solid Polymer Fuel Cells, J. Power Sources, 71, 1998, 199-207) (K. Ledjeff- Hey, Th Kalk, J. Roes; Catalytic cracking of propane for hydrogen production for PEM fuel cells, 1998 Fuel Cell Seminar, Palm Springs, California 1998).

The pyrolysis reactors described above have serious shortcomings and shortcomings.

1) In terms of energy balances:

Indeed, the pyrolysis processes reveal an intrinsic difficulty: the lower calorific value (PCI) of the hydrogen produced is generally of the same order, or even lower than that of carbon and the other residues of pyrolysis. There is therefore a problem of management and exploitation of the energy available in carbon and in the other residues of pyrolysis. If this problem is not properly treated, the overall energy balance of the system can only be very low and therefore unacceptable.

2) At the level of the heat sources necessary for pyrolysis: The heat available to heat and dissociate the fuel within the pyrolysis reactor can come from three distinct sources:

- the combustion of carbon during the regeneration sequence,

- the combustion of other residues from pyrolysis,

- the combustion of unburned gases leaving the fuel cell. These three combustions take place in different chambers: the oxidation of solid carbon takes place inside the reactor during the regeneration phase. The combustion of the gaseous effluents can take place in a combustion chamber. It is therefore necessary to provide very efficient heat exchange structures to ensure the efficiency of heat transfers between the three reaction chambers: reactor undergoing pyrolysis, reactor undergoing regeneration and burner.

3) At the level of the catalyst:

The use of catalytic beds in the two reactors (pyrolysis and regeneration) poses many problems: reduction of efficiency and aging of the catalyst, thermal inertia of the reactor, generation of CO during the pyrolysis phase, cost of the system ...

4) In terms of uses: The systems described above have been designed for the following applications:

- hybrid generation of electricity by means of a fuel cell and mechanical energy by means of a combustion engine (US 5899175),

- generation of electricity in a fuel cell (MG Poirier, C. Sapundzhiev, Catalytic decomposition of natural gas to hydrogen for fuel cell applications, Int. J. Hydrogen Energy, vol. 22, N ° 4, 1997, 429-433), (K. Ledjeff-Hey, N. Formanski, Th. Kalk, J. Roes; Compact Hydrogen Production Systems for Solid Polymer Fuel Cells, J. Power Sources, 71, 1998, 199-207), (K. Ledjeff-Hey, Th. Kalk, J. Roes; Catalytic cracking of propane for hydrogen production for PEM fuel cells, 1998 Fuel Cell Seminar, Palm

Springs, California 1998). The invention aims to eliminate these drawbacks. To this end, it offers a solution allowing the use of the pyrolysis mechanisms for other applications such as: - Pre-reforming for high temperature batteries of the molten carbonates (MCFC) and solid oxides (SOFC) type. - The co-generation of heat and electricity for housing either by coupling with a PEM type heat pump, or by coupling with a SOFC heat pump. - The production of synthesis gas for petrochemicals.

To achieve these results, it proposes a process for producing a hydrogen-rich gas usable in a fuel cell, this process consisting in carrying out, in a reactor, thermal cracking, without catalyst, for pyrolyzing a fuel chosen so to produce either a gas rich in hydrogen and free of carbon monoxide, or a gas rich in hydrogen and containing carbon monoxide. The effluent gases produced during the pyrolysis and inert with respect to the cell are used as fuel at the level of the burner for heating the reactor with a view to bringing it to the reaction temperature.

Advantageously, the pulverulent carbon produced during the pyrolysis reaction can be burned either so as to produce carbon monoxide or CO, at least in part, or so as to produce carbon dioxide or C0 ₂ (heat) to complete the heating reactions and possibly providing a related heating system. According to another characteristic of the invention, the reactor can comprise two or more pyrolysis chambers used alternately. In this case, the chamber not used for pyrolysis is regenerated by the introduction of air which causes, at the right temperature, the combustion of the carbon deposited during the pyrolysis phase.

According to other characteristics of the invention:

- The combustion chamber is placed at the very heart of the reactor (s); - The composition and the nature of the mixture feeding the burner are variable over time;

- The initiation of combustion reactions in the burner takes place by means of a plasma;

- Efficient heat transfer structures between the burner and the pyrolysis chamber are provided;

- The catalytic bed is eliminated;

- The extraction and purification of hydrogen can be carried out using a permeable and selective hydrogen membrane;

- The trapping in the pyrolysis reactor of solid carbon particles is done by means of filters of refractory materials;

- The production of powdery carbon becomes an asset;

- The pyrolysis reactions are catalyzed by means of a plasma.

The combustion chamber is placed at the very heart of the reactor or reactors in order to efficiently heat the reactive mixture at high temperature.

The composition and the nature of the mixture feeding the burner are variable over time. In fact, the priming will be done using the fuel used for the pyrolysis, then once the latter has started, the resulting co-products will be recycled to maintain the reaction in the burner. The carbon monoxide produced during the combustion of powdered carbon can also fuel the burner.

A device for initiating combustion is provided in the burner by means of a plasma. Different means of plasma production are possible. The simplest method is to generate sparks between the electrodes of a spark plug similar to those used in the engines of motor cars. When starting combustion in the burner, it is necessary to involve this ignition system. The latter can be stopped in steady state as soon as the temperature of the burner has reached a value allowing the auto-ignition of the combustible mixture.

Efficient heat transfer structures between the burner and the pyrolysis chamber are provided. First of all, an excellent heat transfer must be ensured between the hot gases resulting from combustion in the burner and the very wall of the burner. This transfer is here made more efficient by increasing the contact surface of the metal with the hot gases by means of a metallic structure. It is shaped, in particular but not exclusively, fins, honeycomb or foam. This structure is arranged inside the burner and the contact with the internal walls of the latter is carried out so as to minimize the thermal resistance at this level. Likewise, the heat transfer between the burner enclosure and the reactive gas circulating inside the pyrolysis reactor is optimized by means of a metal structure similar to the previous one, but arranged in contact with the external wall of the burner. .

It has been found that several drawbacks result from the presence of the catalytic bed in the pyrolysis reactor. Laboratory work carried out by the applicant has shown that, in many cases, the use of a catalyst was not necessary to improve the. reaction yield. The removal of the catalytic bed simplifies the system without reduce its efficiency, reduce its cost and minimize parasitic, uncontrolled CO production.

An optional method of operating the process consists in extracting the hydrogen from the reactor by means of hydrogen-permeable and selective membranes. Besides the advantage of producing a very pure hydrogen, this mode makes it possible to increase the yield of the process since the extraction of the hydrogen displaces the chemical equilibrium in the direction of a more complete reaction. Several ways of introducing hydrogen permeable and selective membranes into the process are described in the text below.

The dehydrogenation reactions considered lead to the production of pulverulent solid carbon. In order to prevent this carbon from going to be deposited in the pipes of the hydrogen circuit and to recover this carbon by oxidation reactions, it is necessary to trap the solid particles of carbon by means of a filter arranged at inside the pyrolysis reactor. The filter is made of a porous refractory material. One possible way of making this filter is to form a tampon made of refractory fibers, made of alumina wool for example.

The production of powdery carbon turns out to be an asset here. Indeed, several options are possible for its valuation. For example, it can, in small power plants, be collected for use as is in industry. It can also be oxidized in situ so as to constitute a source of heat via the production of carbon dioxide or electricity via the production of carbon monoxide.

Carbon dioxide is produced by the combustion reaction: C + 0 ₂ → C0 ₂ ΔH ₂₉₈ = - 393.51 kJ

Carbon monoxide is produced by controlled oxidation, ie: C + y 0 ₂ → CO ΔH ₂₉₈ = - 110.53 kJ The CO produced can then be used as fuel in a SOFC type heat pump, thereby increasing the electricity yield.

Optionally, a plasma generator can be incorporated into the pyrolysis reactor. Indeed, the plasma generates chemically very reactive radicals and plays a role comparable to that of a catalyst. The effect of this effect is in particular to accelerate the dehydrogenation reactions of different hydrocarbons. Several types of plasma generators can be used for this application, mainly "barrier" discharge and microwave discharge generators. Such devices have been described in the prior patent WO 98/28223.

Embodiments will be described below, by way of nonlimiting examples, with reference to the appended drawings in which:

- Figure 1 is a pyrolysis device according to the invention intended to supply a low temperature PEM type heat pump;

- Figure 2 shows a pyrolysis device according to the invention intended to supply a high temperature heat pump of the SOFC type;

- Figure 3 shows a device comprising two reactors;

- Figure 4 shows a complete circuit incorporating the device of Figure 3;

- Figure 5 shows a device using a polymer membrane operating at low temperature intended to purify the mixture rich in H ₂ obtained; FIG. 6 represents a device using a metal membrane operating at high temperature intended to purify the mixture rich in H ₂ obtained

- Figure 7 shows a device characterized in that the metal membrane is placed inside the pyrolysis reactor.

FIG. 1 represents an example of a pyrolysis device according to the invention. This device here supplies a low temperature PEM fuel cell 1. In its simplest form, this device comprises a single cylindrical reactor R heated by a cylindrical burner B incorporated at its heart so as to ensure excellent heat transfer, the reactor-burner assembly being placed in a heat-insulated cylindrical envelope 2 intended for limit thermal losses from the system. The reactor R is delimited by the cylindrical wall of the burner and by an external cylindrical wall coaxial with the burner. It is closed by a bottom FD in the form of a spherical cap and by an annular top DA located around the top of the burner. In this configuration, the pyrolysis reactor operates cyclically. It is indeed successively the seat of pyrolysis reactions which produce a gas rich in hydrogen and carbon oxidation reactions which make it possible to regenerate the reactor.

To supply a PEM type heat pump, it is imperative to avoid introducing CO into the hydrogen-rich gas. Consequently for this application, the pyrolysis of an oxygenated fuel (alcohol, ETB, MTB ...) will be avoided in favor of a hydrocarbon such as methane or propane.

Regarding the pyrolysis phase, the reactor R is heated by means of the burner B to a temperature allowing the cracking reactions of the hydrocarbon used. This temperature is around 550-650 ° C for propane and 700-800 ° C for methane. The fuel, after a possible desulfurization, is introduced into the reactor R by means of the pipe. 3 located at the top of reactor R. Cracking by pyrolysis creates a gas rich in hydrogen and powdery solid carbon which is deposited in the reactor R. The alumina wool filter 4, located in the bottom of the reactor R, retains the carbon particles in the reactor and removes them from the hydrogen-rich gas extracted by the conduit 5 located on the other side of the filter 4. Before being introduced into the anode compartment of the heat pump 1, the hydrogen-rich gas is cooled, by means of a heat exchanger 6, to a temperature compatible with this type of battery, about 50 ° C. At the outlet of the anode compartment, the mixture of residual gases, mainly unburned hydrogen and methane, is recycled to burner B by means of line 7.

The burner B is a combustion chamber supplied at its top with fuel through the pipe 7 and with air through the pipe 8. An additional supply of fuel can be provided to provide additional heat. The combustion, when the system is cold, is initiated by means of a plasma produced for example by an electric discharge between the electrodes of a spark plug of a combustion engine 9 situated at the top of the burner B. When the temperature of the burner B has reached a sufficient level, combustion self-ignition occurs and the plasma is no longer required. In order to increase the efficiency of the heat transfer between the hot gases circulating in the burner B and the hydrocarbon to be cracked in the pyrolysis chamber R, metallic structures 10 for example of the fin, honeycomb or metallic foam type are arranged on either side of the wall of the burner. The hot gases resulting from combustion in the burner escape through the pipe 11 located in the bottom of the burner B. The useful heat contained in the exhaust gases is recovered in a heat exchanger 12. The duration of the sequence of Pyrolysis is limited by the accumulation of powdery carbon in reactor R. This duration is variable depending on the parameters of the system. It can typically be 15 to 30 minutes. When the reactor R is full of carbon, it is necessary to pass to the regeneration phase. Regarding the regeneration phase, a simple way to remove the carbon accumulated in the reactor R is to oxidize it to form a mixture of CO and C0 ₂ . For this, a suitable and heated air flow through the heat exchanger 6 is introduced, at the top of the reactor R, by means of the conduit 13. The conduit 3 is then closed. The reactions of carbon with oxygen in the air are written:

C + 0 ₂ → C0 ₂ ΔH ₂₉₈ = - 393.51 kJ

C + ^l 0 ₂ → CO ΔH ₂₉₈ = - 110.53 kJ

C + C0 ₂ → 2CO ΔH ₂₉₈ = +172.45 kJ The CO + C0 ₂ mixture thus formed is evacuated via line 5 and conveyed to the burner through line 7. During this regeneration phase, the fuel cell 1 of PEM type must not receive CO, for this purpose, it is isolated by means of the solenoid valves 14. It should be noted that EN solenoid valves placed on the conduits, controlled by an electrical control circuit regulate the various gas inflows. The conversion of CO to C0 _{2 is completed} by combustion of the gases in the burner.

The heat released is recovered in the heat exchanger 6 before entering the burner B then the excess heat not transmitted through the walls of the burner is recovered by the heat exchanger 12 via the exhaust gases.

If we consider the pyrolysis of methane or that of propane, with the device shown in Figure 1, the ideal reactions are written:

CH ₄ → C + 2 H ₂ ΔH ₂₉₈ = + 74.6 kJ

C ₃ H ₈ → 3 C + 4 H ₂ ΔH ₂₉₈ = + 103.85 kJ Pyrolysis therefore makes it possible to extract a maximum of 2 moles of hydrogen per mole of methane and 4 moles of hydrogen per mole of propane. As shown diagrammatically in FIG. 1, the process which is the subject of the invention makes it possible to carry out a co-production of heat and electricity from hydrocarbons such as natural gas or propane. The heat is recovered by the two exchangers 6 and 12. The production of electricity is here ensured by a PEM type fuel cell 1 which is powered by hydrogen from pyrolysis.

If the efficiency of the PEM type heat pump is 50%, this device can produce a maximum of 241 kJ of electricity per mole of methane, or 30% of the methane's PCI. The thermal energy recoverable from the exchangers is then 247 kJ. In co-generation, the maximum value of the overall PCI yield of heat + electricity production is therefore 61%. In the case of propane, an electricity production of 482 kJ per mole of propane is obtained, ie 23.6% of the PCI of propane. The thermal energy recoverable from the exchangers would be 1180 kJ per mole of propane. The maximum value of the overall PCI output of heat + electricity production is then 81%. This example is given as an indication in order to define an order of magnitude of the powers released and the yields.

FIG. 2 represents a pyrolysis device according to the invention coupled with a high temperature fuel cell 15 of the SOFC type. Its function is then to transform the fuel into synthesis gas (CO + H ₂ ) directly usable by the PAC 15. This conversion upstream of the cell will be called pre-reforming. It is well known that the conversion efficiency of PACs of the SOFC type is improved when they are supplied with synthesis gas (CO + H ₂ ) rather than directly with a hydrocarbon.

In addition to the advantage of improving efficiency, there is another advantage linked to the operating time of the SOFC type heat pump. Indeed, attempting to avoid external reforming would lead to introducing the hydrocarbon into the anode compartment of the PAC 15 and to carry out an internal steam reforming using the water formed at the anode. This very elegant solution nevertheless encounters a major difficulty linked to the deposition of carbon in the PAC 15. In fact, the pyrolysis reactions of the hydrocarbon cannot be avoided at the working temperatures of the PACs of the SOFC type. These reactions produce solid carbon which accumulates in the PAC 15 which it disrupts the functioning. To avoid this annoyance, there is every interest in having a pre-reformer upstream of the PAC 15; in this case, the reactor R will play this role. Indeed, it produces H ₂ during the pyrolysis phase and CO during the regeneration phase. The device has many similarities with the case presented above for a PEM type heat pump except on the following points:

- The heat exchanger 6 located on the outlet 5 is no longer useful since the gases from the pyrolyzer can be introduced at high temperature into the anode compartment of the heat pump 15. - The solenoid valves 14 have been eliminated because the heat pump 15 accepts CO and therefore does not need to be isolated during the regeneration phase.

- The air flow entering the cathode compartment of the PAC 15 comes out very hot and is recycled in two directions. A solenoid valve EN1 makes it possible to convey the hot air to the burner B through the conduit 8 to maintain the combustion, or to the reactor R through the conduit 13 for the regeneration sequence.

The operation of the process during the pyrolysis phase is substantially identical to that described in the example in FIG. 1 when the pyrolyzer fed a PEM type heat pump. Note the following differences:

- The PAC 15 accepts very well to be supplied with a mixture of H ₂ + CO gas. The constraint of producing a gas rich in hydrogen and totally devoid of CO is no longer necessary in the present situation. It therefore becomes possible to widen the choice of fuel to be pyrolyzed and to extend it to ethanol or to other oxygenated fuels.

- During the pyrolysis phase, a mixture of hydrogen-rich gases is produced, possibly with a CO content. This gas mixture is extracted from reactor R via line 5 and directly sent to the anode compartment of the heat pump 15. - The gaseous effluents from the heat pump 15 leave at high temperature and are directed towards the burner B by the conduit 7 to complete their combustion.

This combustion is ensured by an additional supply of very hot air brought in through line 8 and taken off at the outlet of the cathode compartment of the heat pump 15.

During the regeneration phase, as in the case of coupling with a PEM type PAC, the pulverulent carbon accumulated in the reactor R during the pyrolysis sequence must be gasified by oxidation. It should be noted that there is a fundamental difference here with the case of the PEM type heat pump. Indeed, in the present case, the mixture of CO + C0 ₂ gas produced during the regeneration in the reactor R can be sent directly to the anode compartment of the heat pump 15 via the outlet 5. Thus, thanks to the conversion of the CO in PAC 15, an additional contribution to electricity production is obtained. To maximize this contribution, the operating parameters during the regeneration phase must be adjusted so that the ratio α = CO / C0 ₂ resulting from carbon oxidation is maximum. The means of maximizing this ratio are to carry out a gentle combustion of the carbon, during the regeneration phase, in order to stop the reaction to the formation of CO, namely mainly:

- reduce the air supply,

- reduce the reactor temperature,

- inject steam into the reactor. For example, a SOFC type heat pump is considered to operate with an electrical conversion efficiency of 45%, and is supplied with the gases produced during the pyrolysis. The reactor is powered by methane and the pyrolysis reaction produces a complete conversion of this fuel. It turns out with the hydrogen produced that this device makes it possible to obtain a maximum of 217 kJ of electricity per mole of methane, or 27% of the PCI of methane. If the CO produced during the regeneration phase is also converted into electricity, the electric production of the SOFC type heat pump adds an additional contribution of up to 127 kJ of electricity per mole of methane, i.e. 16% of the methane ICP. . Overall electricity production can therefore in principle reach 344 kJ of electricity per mole of methane, or 43% of the methane ICP. The production of thermal energy would be roughly equivalent.

A co-generation system operating with methane according to this principle would therefore make it possible to produce substantially equal electrical power and thermal power with an overall efficiency (heat + electricity) of around 80%.

The same propane-powered system could reach 434 kJ of electricity per mole of propane from the hydrogen formed, or 21% of the propane ICP. Electricity production from the CO formed could reach 381 kJ per mole of propane, or 18.7% of the PCI of propane. Overall electrical production can therefore in principle reach 815 kJ of electricity per mole of propane, or 40% of the propane ICP. In this case too, the production of thermal energy would be roughly equivalent to the production of electricity and the overall efficiency (heat + electricity) would reach around 80%.

Unlike most of the results obtained using the solutions known to date, it should be noted that the electrical and thermal powers released are roughly equal. The performances announced above assume the complete pyrolysis and regeneration reactions, which will not be the case in reality. These are therefore maximum values which will have to be tried to approximate under real conditions.

FIG. 3 represents a system with two reactors RI and R2 in order to obtain continuous and no longer cyclical operation. The two reactors are delimited by an external cylindrical wall and by the cylindrical walls of the burner B '. The reactors like the burner are respectively closed by a top and a bottom in the shape of a spherical cap. The reactor-burner assembly is placed in a cylindrical insulated envelope 16 intended to facilitate the high temperature maintenance of the pyrolysis reactors and to reduce the thermal losses of the system.

The operating principles of the double pyrolysis chamber device are substantially identical to those previously stated with reference to FIGS. 1 and 2. The existence of two reactors makes it possible to operate one in the pyrolysis sequence while the other operates in regeneration sequence and vice versa. As a result, there is always a reactor producing a hydrogen-rich gas produced by pyrolysis and a reactor in regeneration sequence ensuring the oxidation of carbon. On this diagram we distinguish:

- Burner B ': located at the heart of the system. It is cylindrical in shape and has in its middle a ferrule allowing an enlargement of the combustion chamber. This ferrule accommodates the ignition device 10 in the middle of the left side of the burner B 'and the passages of several conduits in the middle of the right side: an exhaust duct 17 collecting the smoke at the top of the burner B', a duct 18 supplying the burner with fuel at its base and a duct 19 supplying the burner with air at its base also.

- An RI reactor located in the upper part of the device and an R2 reactor in the lower part:

The two reactors RI and R2 are identical. Each of the two is connected to a fuel supply duct 20, an air supply duct 21 and a product discharge duct 22.

For the reactor RI, the conduits 20 and 21 are placed in the upper part of the reactor and the conduit 22 in the lower part just above the conduits of the burner B. For the reactor R2, the conduits 20 and 21 are placed in the lower part of the reactor and the conduit 22 in the upper part just below the conduits of the burner B '. The heat transfer between the hot gases (fumes) from the burner B 'and each reactor is ensured by high-efficiency heat exchange structures 23 of the same type as those mentioned in the examples of FIGS. 1 and 2. Trapping of the particles of carbon produced by the pyrolysis reactions is carried out in the reactor RI and in the reactor R2 by filters 24 made of refractory fibers, for example made of alumina fibers, situated at the level of the conduits 22, on either side of the right part of the shell. This double reactor system can be used to continuously supply a PEM type fuel cell connected in a similar manner to the case shown in FIG. 1 or a SOFC type fuel cell connected in a similar manner to the case shown in FIG. 2.

FIG. 4 is a representation of a complete circuit incorporating the device of FIG. 3. Here, only the gas supply circuits comprising solenoid valves controlled by an electrical control circuit will be described.

The pyrolysis chambers (RI, R2) are supplied by two supply circuits:

- one for fuel. It includes a three-way EN2 valve which flows alternately in the two reactors

- the other for air. It includes a three-way valve EN3 which flows alternately in the two reactors and controlled by the control circuit so as to inject air into the reactor which is not supplied with fuel in order to cause the combustion of the pulverulent carbon from the pyrolysis reaction carried out during the previous cycle.

The two reactor gas outlet conduits converge on a set of two three-way solenoid valves, EN4 and EN5, which can send the gases produced during pyrolysis and during the partial combustion of carbon, to the heat pump for solenoid valve EN4 and in the burner for solenoid valve EV5. The burner is supplied with air by the same supply circuit as the pyrolysis chambers but upstream of the solenoid valve EN3 and with fuel via either solenoid valve EV5 as described above or an solenoid valve EN6 regulating the arrival at the choice of the gases from the heat pump or fuel through an outlet upstream of the EN2 solenoid valve.

FIGS. 5, 6 and 7 describe a variant of the device, object of the invention, consisting in incorporating, before the heat pump, a membrane for purifying hydrogen in the circuit for extracting the gases produced by pyrolysis. The system can thus be used as a generator of very pure hydrogen.

There are two categories of hydrogen permeable membranes that can be integrated into the system:

- polymer membranes. They are widely used for the purification of hydrogen in industry. Such membranes only operate at low temperature, less than 120 ° C, and can therefore only be used outside the reactor, after the hydrogen-rich gas has cooled (Fig. 5).

- metal membranes. These are very selective membranes made of a metal very permeable to hydrogen, generally an alloy of palladium. These membranes can be used at high temperature, typically 500 to 550 ° C. They can therefore be integrated either in the high temperature gas circuit (Fig. 6), or in the reactor itself (Fig. 7).

FIG. 5 represents a device using a polymer membrane 25. The device is similar to that of FIG. 1 except on the following points:

- The membrane 25 is interposed between the heat exchanger 6 and the fuel cell;

- The mixture of hydrogen-rich gas, extracted from the reactor through line 5, then cooled to less than 120 ° C by means of the exchanger 6 is sent to the membrane purifier 25. It comes out in two ways. The first channel NI routes the very pure hydrogen thus extracted to the PEM fuel cell to supply it and the second channel N2 evacuates the residual gases which are recompressed using the heating compressor 26 so as to be recycled with the fuel supplying the pyrolysis reactor via line 3.

FIG. 6 represents a device using a metallic membrane 27 made of palladium alloy operating at high temperature. This device is similar to that of FIG. 4 except for the fact that the membrane purifier 27 is located before the heat exchanger 6. The very pure hydrogen thus extracted is sent to the fuel cell of PEM type after having been cooled by means of the heat exchanger 6. The residual gases are recompressed by means of the compressor 26 so as to be recycled with the fuel supplying the pyrolysis reactor via line 3.

FIG. 7 represents a device having a metal membrane 28 placed inside the pyrolysis reactor itself. This palladium alloy membrane operates at high temperature, typically at 500-550 ° C and has the shape of a cylindrical pencil. In order to avoid an accumulation of carbon particles in direct contact with the membrane, the latter is protected by a sleeve 29 of refractory fibers, an alumina fabric for example. The purpose of this sleeve is to keep the carbon particles away from the membrane. It will be noted that the pyrolysis reactor R can contain, if necessary, several identical membranes so as to increase the surface area of active membrane and thereby the flow of extracted hydrogen.

It will also be noted that if a membrane consisting of a cylindrical rod or a bundle of cylindrical rods is one of the possible possibilities, other configurations are also possible. Thus membranes in the form of plates or a stack of plates can also be envisaged. The main advantage of having the membrane inside the pyrolysis reactor is the simplicity of the system since the compressor 26 and the fuel recirculation loop are saved.

The devices shown in Figures 5 to 7 can be adapted to the case of the double pyrolysis reactor of Figure 3. This adaptation does not pose any particular problem.

Among the applications of the process are the production of co-generation boilers (heat and electricity) in the housing sector as well as recreational vehicles (camping cars, caravans ...). For residential applications, for example individual houses, the power level of a co-generation module will be approximately 5 kWe + 5 kWth.

Depending on the case, the fuels to be used will be: natural gas, propane, heating oil, etc. Fuel cells, in particular of the PEM and SOFC type, make it possible to propose schemes adapted to this type of application.

For more powerful installations, such as urban co-generation for buildings, groups of buildings, hospitals, modules of the order of 200 kWe + 200 kWth should be developed. . Given its relatively low cost and its highly developed distribution, natural gas will be the most used fuel for this application.

Opportunities in the field of agricultural applications are also to be considered. Agricultural greenhouses, for example, show the need for heat and electricity. The _. Green fuels such as ethanol should be able to be used for such applications.

An application of the process has its place in the field of petrochemistry. Indeed, the process constitutes a simple and economic means of production of synthesis gas (CO + H) of which there are important uses in the chemical industry (manufacture of acetic acid, formic acid, acrylic acid, phosgene, isocyannates ...).

Claims

claims

1. Method for producing a hydrogen-rich gas usable in a fuel cell, by thermal pyrolysis of hydrocarbons, characterized in that it consists in carrying out, in a reactor (R, RI, R2), thermal cracking , without catalyst, for pyrolyzing a fuel chosen so as to produce either a gas rich in hydrogen and free of carbon monoxide, or a gas rich in hydrogen and containing carbon monoxide, and to use the effluent gases produced during the pyrolysis and inert with respect to the cell as fuel at the burner (B, B ') to ensure the heating of the reactor with a view to bringing it to reaction temperature.

2. Method according to claim 1, characterized in that the pulverulent carbon produced in said reactor during the pyrolysis reaction is collected for the purposes of industrial use.

3. Method according to claim 1, characterized in that the pulverulent carbon produced in said reactor during the pyrolysis reaction is burned either so as to produce carbon monoxide, at least in part, or so as to produce carbon dioxide carbon (heat) to complete the heating of the reactions and possibly to provide a related heating system.

4. Method according to claim 1, characterized in that the reactor comprises at least one pyrolysis chamber (R).

5. Method according to claim 1, characterized in that the reactor comprises at least two pyrolysis chambers (RI, R2) used alternately, the unused chamber being regenerated by an introduction of air which causes the combustion of carbon deposited during the pyrolysis phase.

6. Method according to claim 1, characterized in that the reactions are initiated and maintained using the heat supplied by the carbon combustion reaction and by a burner (B, B ').

7. Method according to claim 6, characterized in that said burner is located in the heart of the pyrolysis chambers (R, R1, R2).

8. Method according to claim 6, characterized in that said burner is supplied by an air / fuel mixture which is ignited by a spark created between the two electrodes of a spark plug (9).

9. Method according to claim 8, characterized in that said burner, once the reactions have started, is supplied by the gases leaving the fuel cell and / or by the carbon monoxide contained in the gas produced during cogeneration.

10. Method according to one of the preceding claims, characterized in that the heat transfer between the burner and the pyrolysis chamber is provided by a specific structure (10, 23).

11. Method according to claim 10, characterized in that said structure is metallic.

12. Method according to claim 10, characterized in that said structure is in the form of fins, honeycomb, foam.

13. Method according to claim 10, characterized in that said structure is arranged on the internal and external walls of the burner.

14. Method according to one of the preceding claims, characterized in that the reactors contain a filter (4, 24) making it possible to prevent the carbon from leaving the reactor and from entering the fuel cell.

15. The method of claim 14, characterized in that the above filter consists of a porous refractory material.

16. The method of claim 14, characterized in that the above filter consists of a layer of alumina wool.

17. Method according to claim 1, characterized in that the extraction and the purification of the hydrogen are carried out by means of a permeable and selective membrane with hydrogen (25, 27, 28).

18. The method of claim 17, characterized in that the membrane is a polymer membrane (25).

19. The method of claim 17, characterized in that the membrane is a metal membrane (27, 28).

20. The method of claim 1, characterized in that the pyrolysis reaction is activated by means of a plasma generator incorporated in the reactor.

21. The method of claim 20, characterized in that the plasma generator is "barrier" discharges.

22. The method of claim 20, characterized in that the plasma generator is microwave discharges.

23. Device for the production of a hydrogen-rich gas usable in a fuel cell, by thermal pyrolysis of hydrocarbons according to a process consisting in carrying out, in a reactor (R, RI, R2), thermal cracking, without catalyst , to pyrolyze a fuel chosen so as to produce either a gas rich in hydrogen and free of carbon monoxide, or a gas rich in hydrogen and containing carbon monoxide, and to use the effluent gases produced during pyrolysis and inert with respect to the cell as fuel at the burner (B, B ') for heating the reactor with a view to bringing it to reaction temperature, characterized in that it comprises at least one reactor (R, RI , R2) surrounding a burner (B, B ') and means allowing alternating operation in the pyrolysis sequence and the regeneration sequence.

24. Device according to claim 23, characterized in that the reactors are of variable number and of various shapes.

25. Device according to claim 23, characterized in that the reactor (R) is delimited by the cylindrical wall of the burner (B) and an external cylindrical wall coaxial with the burner, the reactor (R) being closed by a bottom (FD) in the shape of a spherical cap and by an annular top (DA) situated around the top of the burner; the reactor-burner assembly being placed in a thermally insulated envelope (2).

26. Device according to claim 23, characterized in that the number of reactors (RI, R2) are delimited by an external cylinder and by the cylindrical wall of the burner (B '); reactors such as the burner being closed by a top and a bottom in the form of a spherical cap; the reactor-burner assembly being placed in a thermally insulated envelope (16).

27. Device according to claim 26, characterized in that the burner (B ') has in its middle a ferrule.

28. Device according to claim 23, characterized in that it comprises means for recovering the unused thermal power.

29. Device according to claim 28, characterized in that the above means are heat exchangers (6, 12).

30. Device according to claim 23, characterized in that it comprises solenoid valves which manage the various gas flows.

31. Device according to claim 30, characterized in that the above-mentioned solenoid valves are located on the various air and fuel supply conduits (EN, EN1, EV2, EN3, EN6).

32. Device according to claim 30, characterized in that the above-mentioned solenoid valves are located on the outlet conduits of the pyrolysis chambers (EN4, EN5, EN6).

33. Device according to claim 30, characterized in that the above-mentioned solenoid valves are located at the level of the inputs to the fuel cell (14).