FR3091712A1 - Electrolyser for domestic use - Google Patents

Abstract

The invention relates to a water electrolyser (2), comprising: an electrochemical cell (21) for electrolysis of water comprising a membrane / electrode assembly (31) with a face in communication with a supply volume of water. water (210); -the electrolyser comprises an external liquid water tank (8); -the electrolyser comprises a pipe (80) having a first connection (81) connected to the external tank (8) and a second connection ( 82) connected to the water supply volume (210), the first connection being located below the second connection; said water supply volume (210) comprises an exhaust circuit (211) putting the upper part of the water supply volume with the outside, and dimensioned to verify the following relationship: Pcech ≥ Pccon + (ρ * g * hv), with Pcech. Figure to be published with the abstract: Fig. 8

Description

Electrolyser for home use

The invention relates to water electrolyzers for hydrogen generation, and in particular to electrolyzers fitted with an electrochemical water electrolysis cell with membrane/electrode assemblies based on a proton exchange membrane.

Among the known applications of home electrolysers is the use in a hydrogen lighter. A lighter is a device used to produce a flame. A hydrogen lighter uses dihydrogen as fuel. Alessandro Volta designed a hydrogen lighter. Döbereiner then perfected the design of the lighter to make it suitable for home use. The lighter thus designed produces its hydrogen by using an attack of zinc by sulfuric acid.

The electrolysis of water by diffusion through a proton exchange membrane has also been developed. Such electrolysis of water is done at relatively low temperature and uses an anode and a cathode in the form of catalytic layers fixed on either side of the proton exchange membrane. This assembly forms a membrane/electrode assembly, generally designated by the acronym AME. The membrane acts as a proton-conducting electrolyte, electron separator, and gas separator. The catalytic layers are the site of oxidation-reduction reactions. At the anode, the water is oxidized into oxygen and protons, according to the following half-reaction:

H ₂ O1/2 O ₂ + 2H ⁺ + 2 e ^- ; E ⁰ O ₂ / H ₂ O=1.23V vs. V _ENH

The proton produced crosses the electrolyte under the effect of the electric field carrying with it water molecules. The proton then recombines on contact with the cathode catalyst into hydrogen thanks to the flow of electrons provided by the power supply:

2H ⁺ + 2e ^- H ₂ ; E ⁰ H ⁺ / H ₂ =0V vs V _ENH

On each side of the catalytic layers, there are generally porous transport layers, for example based on carbon felt for the cathode and titanium grid for the anode. These porous layers are in contact with current collector plates.

To increase the hydrogen production rate, the unit electrolysis cells can be assembled electrically either in parallel or in series. In the case of paralleling, the voltage is identical to all the cells and the electrolysis current applied by the power supply is divided between each cell.

For a domestic application of such a burner, the burning head and the anode can be arranged on the same side of a proton exchange membrane. For reasons of compactness, a flow circuit, putting the burning head in communication with a hydrogen collector, can be brought to cross the proton exchange membrane. When the burner stops, a flame can propagate inside this flow circuit, at the risk of damaging the proton exchange membrane.

For a domestic application of such an electrolyser, for example as a table lighter, the evacuation of the dioxygen resulting from electrolysis creates risks of burns, since this is generated at a temperature close to 80°C.

During the electrolysis of water by the proton exchange membrane, a flow of water crosses the membrane. The dihydrogen produced and led to the burning head is therefore heavily loaded with water. To prevent the flame from being extinguished by this excess water, it is necessary to integrate a phase separator to recover the water present in the dihydrogen produced and thus prevent this water from reaching the burning head. In particular for a domestic application of such a burner, it is necessary to maintain great compactness. The integration of a phase separator in such a burner is therefore problematic, since the known solutions are either excessively complex or too bulky to satisfy the compactness requirement.

For a domestic application of such a burner, for example as a table lighter, and depending on the different operating phases, the collection of dihydrogen from electrolysis can induce the presence within the pipes of unburned hydrogen. The evacuation of this dihydrogen creates risks of burns, since it is generated with a temperature close to 80°C. Moreover, the generation of dihydrogen may be insufficient for a certain period of time, making the operation of the burner not visually controllable.

In general, it is desirable to be able to avoid too high a concentration of dihydrogen near the electrolyser. Moreover, for a portable application, it is desirable to have a small electrolyser.

The invention aims to solve one or more of these drawbacks. The invention thus relates to a water electrolyser, comprising:
-an electrochemical water electrolysis cell comprising a membrane/electrode assembly including a proton exchange membrane, a first face of the membrane/electrode assembly comprising an anode and being in communication with a water supply volume, a second face of the membrane/electrode assembly being in communication with a dihydrogen collector;
the electrolyser comprises an external liquid water tank having an opening in the upper part placing the inside of the external tank in communication with the outside;
- the electrolyser comprises a pipe having a first connection connected to the external tank and a second connection connected to the water supply volume at least at the level of an upper end of the anode, the first connection being located below the second connection;
-said water supply volume comprises an exhaust circuit placing the upper part of the water supply volume in communication with the outside, the exhaust circuit being sized to verify the following relationship:
Pcech ≥ Pccon +, with Pcech the pressure drop of the oxygen flow through the exhaust circuit, Pccon the pressure drop of the oxygen flow through the said pipe, ρ the density of the water, g the earth's gravity, and hv the difference in height between the first connection and the maximum water level in said external reservoir.

The invention also relates to the following variants. Those skilled in the art will understand that each of the characteristics of the following variants can be combined independently with the characteristics above, without however constituting an intermediate generalization.

According to a variant, said opening of the external reservoir is configured to allow the water filling of the external reservoir.

According to a variant, said external reservoir comprises a transparent wall extending over a height between the first and second orifices.

According to another variant, the water supply volume is delimited by an upper wall, the exhaust circuit being made up of a bore passing through the upper wall.

According to yet another variant, the external tank comprises a marker of a maximum filling level, the marker being placed at least 10 mm below said through-bore.

According to yet another variant, said bore is formed in a removable element of the upper wall.

According to a variant, said bore is cylindrical.

According to another variant, the first connection is located at the level of the lower part of the outer tank.

According to yet another variant, the electrolyser further comprises an electrical power supply configured to apply an electrical power or a current or a potential difference between said first and second faces of the membrane/electrode assembly so as to hydrolyze water present in said supply volume.

According to another variant, the power supply is configured to generate a volume flow of oxygen Qg, and the pressure drop Pcech is dimensioned so that Qe<Qg/3, Qe is the volume flow of oxygen through the circuit of exhaust.

According to another variant, the exhaust circuit is sized to allow balancing of the water levels of the external tank and of the volume of water supply in a period of less than 5s after the shutdown of said electrical supply.

The invention also relates to a burner, including a water electrolyser as defined above, further comprising a burning head connected to the dihydrogen collector and configured to ignite the dihydrogen coming from the collector.

Other characteristics and advantages of the invention will emerge clearly from the description which is given below, by way of indication and in no way limiting, with reference to the appended drawings, in which:

is an exploded perspective view of an example of a hydrogen burner including two electrochemical water electrolysis cells, according to one aspect of the invention;

is a detail of the exploded perspective view of the hydrogen burner of FIG. 1, with a section at the level of a membrane-electrode assembly;

is an exploded top view of the hydrogen burner of Figure 1;

is a block diagram of the hydrogen burner of Figure 1;

is a perspective view of the hydrogen burner of Figure 1 with a section at a burning head;

is a side view of the hydrogen burner of Figure 1 with a section at a dihydrogen manifold;

is a detail front sectional view of the hydrogen burner of FIG. 1 at the level of flow circuits putting the burning head in communication with dihydrogen collectors;

is a schematic cross-sectional side view of a fluid circuit between a water supply volume and an external reservoir;

,

,

,

and

are side views of the fluid circuit of Figure 8 at different stages of operation of an electrolyser of the burner of Figure 1;

is a perspective view of the hydrogen burner of Figure 1 with a section at an aeration volume of a feed compartment;

is a schematic sectional side view at a receiving tank.

Certain aspects of the invention relate to a water electrolyser, other aspects of the invention relate to a burner including such an electrolyser.

Figure 1 is an exploded perspective view of an example of a hydrogen burner 1 including a water electrolyser 2. Figure 2 is a detail of the exploded perspective view of the burner of Figure 1. Figure 3 is a exploded top view of this same burner 1. In this example, the electrolyzer 2 includes two electrochemical water electrolysis cells 21 and 22, the invention also applying to an electrolyzer having a single electrochemical cell. In this example, the electrochemical cells 21 and 22 are electrically connected in parallel.

The electrochemical cell 21 includes a membrane/electrode assembly 31, gas diffusion layers 314 and 315 and an openwork electrical current distribution plate 418. The gas diffusion layer 315 is for example made of carbon felt and is in contact with the plate. perforated 418. The days of the plate 418 allow passage of the dihydrogen produced. The membrane/electrode assembly 31 includes a proton exchange membrane 311, an anode 312 disposed on a first face of the membrane 311, and a cathode 313 disposed on a second face of the membrane 311. The anode 312 is in communication with a water supply volume 210. The water supply volume 210 is here delimited both by a body 5 and by a metal cage 20 (for example made of titanium).

Electrochemical cell 21 is housed on one side of body 5. Electrolyser 2 further comprises a clamping plate 41 which covers electrochemical cell 21. Gaskets 417 and 419 are positioned on either side of membrane 311. Gasket 419 surrounds cathode 313 and is compressed between clamping plate 41 and membrane 311. Gasket 417 surrounds anode 312 and is compressed between body 5 and membrane 311. Gaskets 417 and 419 provide sealing between the anodic zone and the cathodic zone and vis-à-vis the outside. The face of the membrane/electrode assembly 31 comprising the cathode 313 is in communication with a dihydrogen collector provided in the clamping plate 41 and detailed below. The clamping plate 41 is screwed onto the body 5. The clamping plate 41 has a function of clamping, mechanically holding the components, and collecting the dihydrogen generated.

The electrochemical cell 22 includes a membrane/electrode assembly 32, gas diffusion layers 324 and 325 and an openwork electric current distribution plate 428. The gas diffusion layer 325 is for example made of carbon felt and is in contact with the plate. perforated 428. The days of the plate 428 allow passage of the dihydrogen produced. The membrane/electrode assembly 32 comprises a proton exchange membrane 321, an anode 322 disposed on a first face of the membrane 321, and a cathode 323 disposed on a second face of the membrane 321. The anode 322 is in communication with a water supply volume 220. The water supply volume 220 is here delimited both by the body 5 and by the metal cage 20.

Electrochemical cell 22 is housed on one side of body 5. Electrolyser 2 further comprises a clamping plate 42 which covers electrochemical cell 22. Gaskets 427 and 429 are positioned on either side of membrane 321. Gasket 429 surrounds cathode 323 and is compressed between clamping plate 42 and membrane 321. Gasket 427 surrounds anode 322 and is compressed between body 5 and membrane 321. Gaskets 427 and 429 seal between the anodic zone and the cathodic zone and vis-à-vis the outside. The face of the membrane/electrode assembly 32 comprising the cathode 323 is in communication with a dihydrogen collector formed in the clamping plate 42 and detailed below. The clamping plate 42 is screwed onto the body 5. The clamping plate 42 has a function of clamping, mechanically holding the components, and collecting the dihydrogen generated.

The metal cage 20 is bathed in water. The metal cage 20 is perforated, to allow access of water to the anodes and to allow the evacuation of the dioxygen produced. The water necessary for the reaction is thus directly in contact with the anodes in the water supply volumes 210 and 220. The metal cage 20 is electrically conductive and forms a current collector shared by the electrochemical cells 21 and 22. , the cage 20 is used to accommodate the water supply volumes 210 and 220 of the two electrochemical cells 21 and 22. The volumes 210 and 220 are here one and the same volume. Consequently, the compactness of the electrolyser 2 is optimized, despite the presence of several electrochemical cells. The cage 20 is here positioned in a middle housing of the body 5.

A cover 50 closes the burner assembly 1.

The electrolyser 2 further comprises a power supply 7. The power supply 7 makes it possible to selectively apply (switching on or off) a potential difference or a current or an electric power between the anode and the cathode of each electrochemical cell. A negative potential is applied by the power supply 7 at the cathodic level of each of the cells 21 and 22, via a conductive yoke 71 and conductive screws 73. The conductive screws 73 are screwed into the clamping plates 41 and 42 so as to form a cathodic electrical contact on the perforated electrical current distribution plates 418 and 428. A positive potential is applied by the electrical power supply 7 at the anode level, via a conductive yoke 72 and conductive screws 74. The conductive screws 74 are screwed into the body 5 so as to form an anode electrical contact with the cage 20. The electrical power supply 7 can be a controlled power supply. The electrical power supply 7 can for example be driven by current, voltage or power. Current control makes it possible to control the hydrogen production rate. Power control of electrochemical cells can for example be achieved up to 100W. Limiting the cell voltage to less than 2.3V prevents excessive heating and degradation of the electrochemical cells.

The power supply 7 is controlled by an electronic control unit, not shown here. The power supply 7 can be activated for example by activating a push button. The electronic unit can deactivate the electrical power supply 7 after a predefined period, for example 20 seconds, if the push button is pressed. The power supply 7 may include batteries for storing electrical energy. Burner 1 may have a charging interface for such batteries, for example via a USB port.

An external tank 8 is intended to receive water. The external tank 8 is in communication with the water supply volumes 210 and 220, according to an example detailed below. The external tank 8 is intended to supply additional water to volumes 210 and 220, as water is consumed. Indeed, water is consumed in the electrolysis reaction and partly evacuated with the gases produced. The outer reservoir 8 has a water filling orifice 84 at its upper end. Orifice 84 can for example be provided in a cover of external reservoir 8.

The burner 1 comprises a burning head 19, configured to ignite dihydrogen coming from the dihydrogen collectors. The flame generated can advantageously be colored (notably for safety reasons) by passing it through a reservoir of suitable salts. The burning head 19 may include an ignition device, not shown, in order to ignite the dihydrogen.

FIG. 4 is a block diagram of the hydrogen burner 1. The operation of the electrochemical cells 21 and 22 being equivalent, only the operation of the cell 21 will be described below.

When the power supply 7 applies a potential difference on either side of the membrane 311, water in the volume 210 is oxidized at the level of the anode 312, H ⁺ ions cross the membrane 311. The ions H ⁺ recombine with electrons at cathode 313, to form dihydrogen. The dihydrogen formed is recovered at the level of a collector 410 made in the plate 41. The dihydrogen is then led to the burning head 19 via a flow circuit 11. A part of the dioxygen produced in the volume 210 is evacuated to the outside via an exhaust circuit 211. Another part of the dioxygen produced is evacuated through the external tank 8, after having been introduced at the level of the bottom of this tank 8. This part of the dioxygen produced thus passes through the water in reservoir 8 and thus undergoes cooling. The oxygen having passed through the reservoir 8 is thus evacuated at the level of the filling orifice 84, after having been cooled. Water contained in the flow of dihydrogen formed can be recovered by a phase separator 51, then reintroduced into tank 8, as will be detailed below. Advantageously, reservoir 8 is either transparent or has a transparent wall. It is thus possible to check the level of water in the tank 8, or to check the correct operation of the electrolyser 2, by observing bubbles of oxygen passing through this tank 8.

Figure 5 is a perspective view of the burner 1 with a section at the level of its burning head 19. Figure 6 is a side view of the burner 1 with a section at the level of the dihydrogen collector. As illustrated, the burn head 19 and the water supply volume 210 (in the body 5) are separated from the dihydrogen collector 410 of the plate 41 via the membrane/electrode assembly 31, and in particular by the membrane 311. Similarly, the burn head 19 and the water supply volume 220 (in the body 5) are separated from the hydrogen collector 420 of the plate 42 via the membrane assembly / electrodes 32, and in particular by the membrane 321. By separating is meant to mark a boundary. The membranes 311 and 321 are thus interposed between respective dihydrogen collectors on the one hand, and the burning head 19 and the water supply volumes 210 and 220 on the other hand.

A flow circuit 11 puts the burning head 19 and the dihydrogen collector 410 in communication. The circuit 11 is here partially formed in the plate 41 and in the body 5. The circuit 11 is here substantially horizontal. The proton exchange membrane 311 comprises a bore 310 through which the flow circuit 11 passes. The circuit 11 is placed in communication with the burning head 19 via a vertical pipe 13.

A flow circuit 12 communicates the burning head 19 and a dihydrogen collector 420. The circuit 12 is here partially formed in the plate 42 and in the body 5. The circuit 12 is here substantially horizontal. The proton exchange membrane 321 comprises a bore 320 through which the flow circuit 12 passes. The circuit 12 is placed in communication with the burning head 19 via the vertical pipe 13.

FIG. 7 is a detail front sectional view of the burner 1, at the level of the flow circuits previously described. A seal 61 is disposed at least partially in the bore 310. The seal 61 separates the proton exchange membrane 311 from the flow circuit 11. Thus, the contents of the flow circuit 11 are not in contact with the membrane. proton exchanger 311 at its bore 310. This avoids damaging the membrane 311 by flashback when the production of dihydrogen is stopped. The seal 61 is here composed of two distinct elements 611 and 612. The elements 611 and 612 are superposed with a part of the membrane 311 around the bore 310. The membrane 311 is compressed between the elements 611 and 612. The elements 611 and 612 are themselves compressed between the plate 41 and the body 5. The elements 611 and 612 project radially with respect to the edge of the bore 310. Due to their compression, the elements 611 and 612 are even projecting radially inside the bore 310 and are in mutual contact in this bore 310.

Advantageously, the elements 611 and 612 project radially into the bore 310 over a distance D such that D>Epm/2, with Epm the thickness of the proton exchange membrane 311. In order to promote significant deformation of the seal 61 and guarantee contact between the elements 611 and 612 in the bore 310, the cumulative thickness of the elements 611 and 612 in the absence of compression is advantageously at least twice the thickness Epm of the proton exchange membrane 311.

In order to present the same deformation in the event of compression, the elements 611 and 612 are advantageously formed with the same material. The seal 61 may be formed with a material having a Shore A hardness at most equal to 60, to promote its deformation. The seal 61 may be formed with a material having an electrical resistivity greater than 100 Ω/cm. The seal 61 is advantageously formed with a material resistant to a temperature of at least 180° C., so as to be able to resist flashbacks in the circuit 11.

A seal 62 is disposed at least partially in the bore 320. The seal 62 separates the proton exchange membrane 321 from the flow circuit 12. Thus, the contents of the flow circuit 12 are not in contact with the membrane. proton exchanger 321 at its bore 320. This avoids damaging the membrane 321 by flashback when the production of dihydrogen is stopped. The seal 62 is here composed of two distinct elements 621 and 622. The elements 621 and 622 are superposed with a part of the membrane 321 around the bore 320. The membrane 321 is compressed between the elements 621 and 622. The elements 621 and 622 are themselves compressed between the plate 42 and the body 5. The elements 621 and 622 project radially with respect to the edge of the bore 320. Due to their compression, the elements 621 and 622 are even projecting radially inside the bore 320 and are in mutual contact in this bore 320.

Advantageously, the elements 621 and 622 project radially into the bore 320 over a distance D such that D>Epm/2, with Epm the thickness of the proton exchange membrane 321. In order to promote significant deformation of the seal 62 and guarantee contact between the elements 621 and 622 in the bore 320, the cumulative thickness of the elements 621 and 622 in the absence of compression is advantageously at least twice the thickness Epm of the proton exchange membrane 321.

In order to present the same deformation in the event of compression, the elements 621 and 622 are advantageously formed with the same material. The seal 62 may be formed with a material having a Shore A hardness at most equal to 60, to promote its deformation. The seal 62 may be formed with a material having an electrical resistivity greater than 100 Ω/cm. The seal 62 is advantageously formed with a material resistant to a temperature of at least 180° C., so as to be able to resist flashbacks in the circuit 12.

According to another aspect of the invention, a phase separation is implemented in the vertical conduit 13 placing the dihydrogen collectors 410 and 420 in communication with the burning head 19. The phase separation by gravity in the conduit 13 s proves to be insufficient, an excessive quantity of water being able to be carried towards the burning head 19.

As illustrated in Figure 7, metal fibers 51 are housed in the vertical conduit 13 and oriented vertically. Such metallic fibers 51 in the duct 13 make it possible to achieve phase separation between the collector 410 (and the collector 420) and the burning head 19, without requiring the interposition of a complex or bulky phase separator device.

Fibers 51 are here housed in an upper part 131 of conduit 13. The fibers 51 here are metal fibers. The flow circuits 11 and 12 are connected to the conduit 13 at the level of a connection 133. The connection 133 forms a delimitation between the upper part 131 and a lower part 132 of the conduit 13. The connection 133 is here housed under the head of burning 19 and under the metal fibers 51 housed in the upper part 131.

Advantageously, the flow circuit 11 (and the flow circuit 12) has an angle of between 70° and 110°, and is preferably perpendicular to the pipe 13 at the level of their connection 133. By such a change in direction of the flow flux at the level of the link 133, a separation by inertia between the dihydrogen and the water is promoted. The metal fibers 51 also form successive baffles in the upper part 131 of the conduit 13, also favoring phase separation.

To further promote phase separation, the conduit 13 advantageously includes a throttle section 135 positioned between the burning head 19 and the upper end of the metal fibers 51.

The fibers 51 are for example in the form of a wick of metal straw, for example steel wool or stainless steel wool. Strip stainless steel wool can thus be rolled up to form a reel introduced into the conduit 13. The metal fibers 51 here form a porous mass. The metal fibers 51 are advantageously made of a metal resistant to corrosion by deionized water, and are for example made of stainless steel.

The vertical orientation of the fibers 51 makes it possible both to guide the flow of dihydrogen towards the burning head 19, and to guide the water collected downwards, towards the lower part 132 of the conduit 13. The water collected by the metal fibers 51 flows downward by gravity.

The lower part 132 of the conduit 13 extends downwards as far as communication with a water collection tank 504. Metal fibers 51 are also housed in the lower part 132 of the conduit below the connection 133. The part bottom 132 of conduit 13 can thus participate in the phase separation of the dihydrogen flow. Metal fibers 51 can extend in one piece between the upper part 131 and the lower part 132 of the pipe 13. The lower part 132 of the pipe 13 can also serve as a storage tank for the water resulting from the separation of phase.

To allow good control of the quantity of water present in the reservoir 504, its internal volume is advantageously devoid of any solid element, and in particular devoid of metal fibres.

According to another aspect of the invention, the burning head 19 is placed in communication with the dihydrogen collector via a phase separation device, as described previously. The burner 1 further comprises a tank 504 for storing or recovering the water resulting from the phase separation, as described above. The tank 8 is placed in communication with the water supply volume 210 via a pipe 80, as described later. Burner 1 further comprises a valve 505 controlled to selectively open or close fluid communication between tank 504 and tank 8. Valve 505 may also be associated with a pump to promote flow from tank 504 to tank 8. The valve 505 is here connected to the bottom of the external tank 8 via a hydraulic connector 85.

When the electrolyser 2 is electrically powered and produces dihydrogen, an excess of dihydrogen can be introduced into the tank 504 via the lower part of the conduit 13. At the start of the electrical supply to the electrolyser 2, the valve 505 can be opened to evacuate this excess dihydrogen through tank 8.

This excess dihydrogen will also be cooled by the water present in tank 8. During this phase, the oxygen produced in the electrolyser does not yet reach tank 8. The introduction of hydrogen into tank 8 n is therefore not a problem during this phase. In addition, the passage of dihydrogen in tank 8 makes it possible to visually check the operation of burner 1 by bubbles formed in this tank 8, if the flame is invisible and before the dioxygen reaches tank 8 to form bubbles. The dihydrogen evacuated through tank 8 is humidified, which makes it possible to increase its proportion in the air, which may involve an explosion.

During the continuation of the generation of dihydrogen, for example after 2 seconds after the supply of the electrolyser 2 by the supply 7, the valve 505 is closed. This prevents a flow of dihydrogen from crossing the tank 8, when a flow of dioxygen will cross this tank 8.

As in the example detailed above, the connection 133 is arranged under the burning head 19. This connection 133 is advantageously positioned above the height h2 (height of a connection orifice 82 detailed later), which makes it possible to maintain a water level in the pipe 13 below the level of the connection 133. This prevents the water initially present in the pipe 13 from initially disturbing the flow of dihydrogen towards the burning head 19. This proves to be particularly advantageous if the valve 505 is of the normally open type, and is thus kept open after the end of the power supply to the electrolyser 2 by the electrical power supply 7, since the level of the water in the tank 8 and the level in the conduit 13 then balance each other. Such balancing makes it possible in particular to reintroduce water into the tank 8, this water coming from the phase separation of the dihydrogen flow generated by the electrolyser 2. A valve 505 of the normally open type can then prove to be advantageous for achieve such water level balancing in a simple and compact manner.

The pipe 13 is here advantageously vertical in order to promote the recovery of the water intended for the tank 504.

Advantageously, a dimensioning will be implemented to guarantee a sufficient quantity of water in the electrolyser 2 throughout the electrolysis.

Faraday's law makes it possible to deduce the quantity of protons exchanged by the reaction in electrolyser 2, depending on the current and the reaction time of the electrolysis:

nx =Iel*t/(nel*Far) [eq 1]

with nx the quantity of a species exchanged in mol, Iel the electrolysis current, t the duration of the electrolysis, nel the ratio between the number of electrons exchanged in the reaction with respect to the species considered, and Far Faraday's constant.

By application to the electrolysis of water, taking protons as the species studied, the nel ratio is equal to 1 since the number of protons and electrons exchanged is identical. We then obtain the following equation:

np =Iel*t/Far [eq 2]

with np the quantity of protons exchanged.

Studies have empirically shown that a proton crossing a proton exchange membrane entrains 5 molecules of water through the membrane. We then obtain:

neau =5*Iel*t/Far [eq 3]

with neau the amount of water passing through the membrane.

We then deduce the volume of water crossing the membrane:

Calf=5*Iel*t*M/(Far* ρ) [eq 4]

With Veau the volume of water crossing the membrane, ρ the density of water, M the molar mass of water.

The volume available in line 13 (under connection 133) and in tank 504 must be greater than water, to be sufficient to be able to return to tank 8 the volume of water which will pass through the proton exchange membranes during the reaction of electrolysis. In the presence of metal fibers 51 in conduit 13, the volume of these fibers 51 will have to be subtracted from the volume of conduit 13.

Advantageously, a sizing can be implemented to guarantee a purge of the water present between the connection 133 and the tank 8, when starting the electrolyser 2. Such a purge makes it possible to limit the risk of too large a quantity of water in the flow of dihydrogen at the level of the burning head 19, in particular if it is estimated that the efficiency of the phase separation is insufficient after several successive cycles of use of the electrolyser 2. Such a purge can induce a flow of dihydrogen destined for the tank 8, when the electrical power supply of the electrolyser 2 by the power supply 7 begins. In general, the pressure drop of the flow of dihydrogen between the connection 133 and the flame formation zone of the burning head 19 so that the latter is greater than the height of water in the tank 8 and the dihydrogen flow pressure drops between the connection 133 and this tank 8.

As illustrated in Figure 15, a flame formation zone 192 is formed at the upper end of a needle 191. The lower end of the needle is in communication with a chamber 193 inside the head of the needle. burning 19. The chamber 193 is in communication with the upper part 131 of the duct 13. The needle 191 has a diameter which defines the speed of the dihydrogen in the zone of ignition or flame formation. The diameter of the needle 191 thus defines the height of the desired flame and will therefore generally be adapted according to this parameter. Needle 191 can also be used as a pipe with a section smaller than that of conduit 13, including by extension chamber 193. Such a reduction in section thus makes it possible to increase the pressure drop between connection 133 and zone 192.

More precisely, if the diameter of the needle 191 is fixed to define the flame height, the pressure drop can be calibrated by varying the length of the needle 191 in particular. In the present example, the needle 191 forms a cylindrical pipe of circular section with a calibrated diameter Da, extending over a length La. The pressure drop caused by the needle 191 is the sum of all the singular pressure drops and regular.

The singular head loss ∆Psa is here due to the section changes of the needle 191:

∆Psa=(kea+ksa)*ρh*Vh ² )/2 [Eq5]

with :
- ∆Psa the singular head loss in Pascal generated by needle 191;
- kea a unitless coefficient due to the entry of the needle 191;
- ks a unitless coefficient due to the output of the needle 191;
- ρh the density of hydrogen saturated with humidity;
- Vh The speed of dihydrogen in the diameter of needle 191.

The coefficients kea and ksa are notably defined by the changes in diameter respectively at the entry and exit of the needle 191.

The regular head losses ∆Pra are defined as follows:

∆Pra=(fa*La*ρh*Vh ² )/(2*Da) [Eq6]

with fa a coefficient of friction of the flow through the needle 191. The coefficient of friction fa, assuming a flow in laminar regime is given by the formula
fa=64/Re [Eq7]
with Rea the Reynolds number given by the relation
Rea=ρh*Vh*Da/µh [Eq8]
with µa the dynamic viscosity of hydrogen saturated with humidity.

The head loss Pca of needle 191 can be defined as follows:

Pca=(((kea+ ksa)*ρh*Vh ² )/2)+(32µh*La*Vh)/(Da²) [Eq9]

The speed Vh is defined by the following equation:
Vh = Qea/Sea [Eq10]
with
- Qea the flow through the needle 191;
- Sea the needle passage section 191
with
Sea=π*Da²/4 [Eq11]

We deduce from the equation [Eq10]
Vh=4*Qea/(π*Da²) [Eq12]

We can then express Pca as follows:

Pca=(((kea+ ksa)*ρh*(4*Qea/(π*Da²)) ² )/2)+(128µh*La*(Qea/( π*Da ⁴ ))) [Eq13]

The same reasoning can be applied to all the singularities and linearities between link 133 and zone 192 to calculate a total pressure drop Pcs between link 133 and zone 192. For a known geometry between link 133 and zone 192 192, the only variables or unknowns remain the values Qea and La. We can then simplify the pressure drop Pcs by the following relationship:

Pcs= fm (Qea, La) [Eq14]

with fm a function.

By applying the same reasoning for the flow between connection 133 and reservoir 8, we obtain Pcr the pressure drop of the hydrogen flow to reservoir 8:
Pcr= fn (Qrr) [Eq15]
with fn a function and Qrr the flow rate of dihydrogen to tank 8. This pressure drop Pcr is in particular a function of the pressure drop coefficient through valve 505, generally provided with the technical documentation of its manufacturer.

It will be considered that the pressure variation Pcres between the connector 85 and a filling orifice 84 is due to the height of the water column in the tank 8.

Pcres=(ρ*g*hmax) [Eq16]
with ρ the density of the water, g the earth's gravity, and hmax the height of the water column in the reservoir 8. Hmax is for example defined by the height at rest of the water column in the reservoir 8 , added to the height of water due to the volume of water discharged into the tank 8 via the valve 505.

Hmax=hinit+Calf/ Srext [Eq17]

with hinit the height of the water column in tank 8 at rest, Srext the section of the outer tank 8 along a horizontal plane.

By combining [Eq15] and [Eq16], we deduce:

Pcr+Pcres= fn (Qrr)+ (ρ*g*hmax) [Eq18]

The pressures at the level of the zone 192 and at the level of the opening 84 are here equal to the atmospheric pressure denoted Patm. We can then deduce from equations [Eq14] and [Eq18]:

fm (Qea,La) -fn (Qrr)-(ρ*g*hmax)=0 [Eq19]

We can define the total flow of dihydrogen Qh=(22.4*10 ^-3 *La)/193000 [Eq20]

And with Qh=Qrr+Qea, we then have:

fm (((22.4*10 ^-3 *La)/193000-Qrr), La)- fn (Qrr)-(ρ*g*hmax)=0 [Eq21]

By solving the equation [Eq21], for example by a numerical solver, au can determine the flow rate Qrr returned to tank 8 when the water has been purged between link 133 and tank 8 via the valve 505. The Reynolds Rea number can then be determined from [Eq8] to verify the hypothesis of a laminar flow of dihydrogen towards reservoir 8 and needle 191. If the flow is not laminar, it recalculate the coefficient of friction fa of linear pressure losses with a Moody diagram.

La will advantageously be sized so that, in steady state, so as to favor the draining of water:
Qrr>0.8*Qh

Guaranteeing this relationship in steady state also guarantees this condition in transient state, since the pressure of the water column in reservoir 8 is lower in transient state. The sizing can be carried out so that the water purge (and therefore the steady state is reached) is carried out in less than 0.5 seconds after the start-up of the electrolyser 2.

Figure 8 is a schematic cross-sectional side view of a fluid circuit between the water supply volume 210 and the external reservoir 8. Figures 9 to 13 are schematic cross-sectional side views of the same circuit at different Burner electrolyser operating steps 1.

In the volume 210, it is desirable for the membrane/electrode assembly 31 to remain immersed in the water contained in the volume 210, to prevent, for example, part of the active surface of the anode 312 from coming into contact with a volume of air/dioxygen in the upper part of the supply volume 210. For a minimum height hmin of the water in the supply volume 210, it can therefore be considered that hmin must be greater than a height hmem corresponding to the top of the active surface of the membrane/electrode assembly 31.

Furthermore, the tanks 210 and 8 are connected by a pipe 80, putting an orifice 82 of the tank 210 in communication with an orifice 81 of the tank 8. At rest, to also guarantee a rebalancing between the water level of the external tank 8 and the supply volume 210, the level in the outer reservoir 8 must be maintained at a height h2 of the orifice 82. The minimum height hmin is here defined by the height h2. The orifice 82 is thus located above the height hmem.

Finally, the volume 210 is surmounted by a cover 212 in which the exhaust circuit 211, also designated exhaust vent, makes it possible to evacuate the air/dioxygen mixture contained in the upper part of the supply volume 210 A maximum height hms of the water in the supply volume 210 is defined such that, at rest of the burner 1 and as indicated in FIG. 9, the water level of the supply volume 210 is advantageously at least at 10mm below exhaust vent 211, to prevent water blockage.

It is thus possible to give a particular dimensioning of the exhaust vent 211.

With a tank 8 and a supply volume 210 of constant horizontal sections over their height, the outer tank 8 has a reserve height Δh, to contain the volume of water expelled from the supply volume 210, which can be defined as follows:

∆h=(Svalim*(hms-hmin))/Srext [Re1]

with :
- Svalim the section of the supply volume 210 according to a horizontal plane,
- Srext the section of the outer tank 8 along a horizontal plane.

The head loss Pcech of the calibrated vent 211 can be determined by following the design rule below: part of the total flow of oxygen produced is evacuated by the vent 211, the other part of the flow having to escape by bubbling through tank 8 so that it can be cooled before escaping to the outside.

In the present example, the vent 211 is a cylinder of circular section with a calibrated diameter d, extending over a length l corresponding to the thickness of the cover 212. The pressure drop caused by the vent 211 of calibrated diameter is the sum of all the singular and regular head losses.

The singular head loss ∆Psing is here due to changes in section of vent 211:

∆Psing=(kent+ksor)*ρo*V ² )/2 [Re2]

with :
- ∆Psing the singular head loss in Pascal through the vent 211;
- kent a coefficient without unit due to the entry of the vent 211;
- ksor a unitless coefficient due to the outlet of the vent 211;
- ρo the density of oxygen saturated with humidity;
- V The oxygen velocity in the diameter of vent 211.

The regular head losses ∆Preg are defined as follows:

∆Preg=(f*l*ρo*V ² )/(2*d) [Re3]

with f a coefficient of friction of the flow through the vent 211. The coefficient of friction f, assuming a flow in laminar regime is given by the formula
f=64/Re [Re4]
with Re the Reynolds number given by the relation
Re=ρo*V*d/µ [Re5]
with µ the dynamic viscosity of oxygen saturated with humidity.

The head loss Pcech of the calibrated vent 211 can be defined as follows:

Pcech=(((kent+ ksor)*ρo*V ² )/2)+(32µ*l*V)/(d²) [Re6]

The speed V is defined by the following relationship:
V = Qe/Se [Re7]
with
- Qe the flow through the vent 211;
- Se the passage section of the circular vent 211
with
Se=π*d²/4 [Re8]

We deduce from the relation [Re7]
V=4*Qe/(π*d²) [Re9]

We can then express Pcech as follows:

Pcech=(((kent+ ksor)*ρo*(4*Qe/(π*d²)) ² )/2)+(128µ*l*(Qe/( π*d ⁴ ))) [Re10]

The geometry of the volume 210 being known between the height h2 and a height hmc corresponding to the lowest level of the lower surface of the cover 212, the only unknowns/variables are Qe, Se and l for the pressure difference between the vent 211 and the height h2 inside the volume 210. This pressure difference can thus be expressed by:
P _{21 1 -h2} = m (Qe, Se, l) [Re11]
with m a function.

By applying the same reasoning for the flow through the pipe 80, we obtain Pccon the pressure drop of the flow of oxygen through this pipe 80:
Pccon= n (Qr) [Re12]
with n a function and Qr the oxygen flow through line 80 and reservoir 8.

It will be considered that the pressure variation between a height h1 of the orifice 81 and the height hms is essentially due to the height of the water column in the tank 8. In this case, we can write Pcol the pressure due to the column of water as follows:
Pcol=(ρ*g*hv) [Re13]
with ρ the density of water, g the earth's gravity, and hv=(hmr-h1). Furthermore :
hmr=hms+∆h [Re14]
We can then define the pressure difference between points 82 and 84 as follows:
P _82-84 = n (Qr)+(ρ*g*hv) [Re15]

The pressures at the level of the vent 211 and at the level of the opening 84, at a height hmr corresponding to the top of the reservoir 8, are here equal to the atmospheric pressure denoted Patm. We can therefore deduce from the relation [Re11] that the pressure P82 at the level of the orifice 82 is defined as
P82= m (Qe, Se, l)+Patm [Re16]

We can also deduce from the relation [Re15] that the pressure P82 can also be defined as
P82= n (Qr)+(ρ*g*hv)+Patm [Re17]

We then deduce the following relationship:
m (Qe, Se, l)- n (Qr)-(ρ*g*hv)=0 [Re18]

The total flow of oxygen Qg generated by the electrolyser 2 can be defined by the following relationship:
Qg=(22.4*10 ^-3 *Iel)/386000 [Re19]
with Iel the electrolysis current.
The relationship
Qg=Qe+Qr [Re20]
is also checked.
We can deduct
Qe=(22.4*10 ^-3 *Iel)/386000-Qr [Re21]

By combination with the relation [Re18], we deduce:
m (((22.4*10 ^-3 *Iel)/386000-Qr), Se, l)- n (Qr)-(ρ*g*hv)=0 [Re22]

The resolution of the relation [Re22] (for example by means of a numerical solver) makes it possible to determine the flow rate Qr of dioxygen which crosses the external tank 8 in steady state, for a given length l of the vent 211, as well as the flow rate Qe. The Reynolds number Re can then be determined from [Re5] to verify the hypothesis of laminar flow in pipes. If the flow is not laminar, it is necessary to recalculate the coefficient of friction f of linear head losses with a Moody diagram.

Once the gas flow Qe passing through the calculated vent 211, it is possible to compare the proportion of this flow Qe with respect to the flow Qg. Advantageously, we will make sure to respect the following relationship in steady state:
Qg/2≥Qe [Re23]
and preferably Qg/3≥Qe [Re24]

Advantageously, the vent 211 (and in general the exhaust circuit 211) is sized so that the level in the supply volume 210 and the external tank 8 is balanced after the electrolyser has stopped. For this purpose, the following relation must be solved, in addition to the relation [Re23]:

ρ*g*hv- [Re25]

By derivation of the relation [Re16], taking into account that only the flow through the vent 211 participates in the evolution of the pressure P82, we obtain:

Nevertheless, the resolution of relation 25 leads by default to an infinite level balancing time. We can consider the rebalancing as satisfactory when the difference after a duration x defined between the level of volume 210 and the level of reservoir 8 is less than 10% of ∆h. In this case, the relation [Re25] becomes:

A sizing result can be obtained by iterations of the relation [Re27] by modifying the diameter and/or the length of the vent 211 according to the flow rate of oxygen produced and fixing an equilibration duration x for example at most equal at 5s.

When this aspect of the invention is combined with the purge between the connection 133 and the tank 8, as described above, the flows and the openings/closings of the valve 505 are advantageously dimensioned so that:
a dihydrogen flow to tank 8 is established before a dioxygen flow to this tank 8;
the flow of dihydrogen to tank 8 is already interrupted when the flow of dioxygen to this tank 8 is established. The instant of appearance of the flow of oxygen to the tank 8 can be determined experimentally.

The electrolysis reaction generates a volume of dihydrogen twice that of dioxygen, due to the stoichiometry. If the volume of water to be moved to reach tank 8 were identical for the dioxygen (via pipe 80) and for the dihydrogen (via valve 505), the dioxygen should reach tank 8 in twice the time required for the dihydrogen. In order to guarantee that the flow of dihydrogen can be interrupted when the flow of dioxygen towards the tank 8 is established, the following relationship is advantageously respected:
Vdh<Vdo/4 [Re28]
with Vdh the volume to be traversed for the dihydrogen from the configuration at rest, before reaching the reservoir 8 via the valve 505, Vdo the volume to be traversed for the dioxygen from the configuration at rest, before to reach reservoir 8 via line 80.

In order not to excessively delay the flow of oxygen in the reservoir 8, the following relationship is advantageously respected:
Vdo/8<Vdh<Vdo/4 [Re29]

According to another aspect of the invention, an air cooling circuit 700 is arranged in a box 70. The power supply 7 is housed in this box 70 and therefore benefits from the cooling by the circuit 700.

FIG. 14 is a perspective view of the hydrogen burner 1 with a section at the level of a ventilation volume of the casing 70. The movement of air in the cooling circuit 700 is here generated by a fan 703. The fan 703 is advantageously placed between an inlet 701 and an outlet 702 formed in the walls of the box 70. The fan 703 comprises for example an electric motor powered by the power supply 7.

The inlet 701 is advantageously made in the lower part of an outer wall of the casing 70. The inlet 701 can for example be made in the outer side wall of the casing 70, as shown in FIG. 14, or for example in the wall constituting the caseback 70.

The outlet 702 is advantageously provided in the upper part of the side wall of the housing 70 immediately adjacent to the outer reservoir 8. This side wall is here opposite the outer side wall in which the inlet 701 described above is provided, in order to optimize the power supply cooling 7.

As illustrated in FIG. 14, the input 701 and the output 702 have, for example, rectangular shapes. The lower and upper faces of the wall delimiting the outlet 702 have for example an inclination with respect to a horizontal plane parallel to the outer upper face of the box 70. For these lower and upper faces, their edges internal to the box 70 are advantageously placed in below their outer edges; it can thus be considered that the outlet 702 has an upward inclination from the inside to the outside of the housing 70, in order to direct the outgoing air flow towards the top of the tank.

Due to the bubbling produced in the reservoir 8 according to another aspect of the invention described above, gaseous reactants produced during the electrolysis reaction are caused to escape from the orifices 84 made in the lid of the reservoir 8. The flow of air created by the cooling circuit 700 and passing via the outlet 702 contributes to the dilution of these gaseous reactants by emerging at the level of an upper part of the electrochemical cell, and in particular at the level of the upper part of the tank 8 The outlet 702 could have different possible shapes depending on the shape and direction desired for the flow of dilution air.

Such a dilution makes it possible to reduce the risk of ignition by guaranteeing a very low concentration of dihydrogen above the tank. Such a dilution is also carried out without making the electrolyser more complex or energy-consuming, since an air flow already available is reused to ensure the cooling of the electrical power supply 7.

The direction of flow chosen (suction of fresh air via 701 then blowing of air from the electrical compartment 70 via outlet 702 towards the top of the tank potentially enriched with H ₂ , and not suction of air potentially enriched with hydrogen via the inlet 701) prevents hydrogen-enriched air from passing through housing 70 and coming into contact with electrical components of power supply 7, which could potentially generate sparks.

This aspect of the invention proves to be particularly advantageous in combination with the placing of the external tank 8 in communication with the recovery tank 504 via the valve 505, due to a potential evacuation of dihydrogen via of the external tank 8.

The outlet 702 and the cooling circuit 700 can be designed to adapt the outgoing air flow, so that this flow is laminar, turbulent or vortex depending on a desired dilution mode.

Claims (12)

Water electrolyser (2), comprising:
-an electrochemical cell (21) for water electrolysis comprising a membrane/electrode assembly (31) including a proton exchange membrane (311), a first face of the membrane/electrode assembly comprising an anode (312) and being in communication with a water supply volume (210), a second face of the membrane/electrode assembly (31) being in communication with a dihydrogen collector (410);
characterized in that:
the electrolyser comprises an external liquid water tank (8) having an opening (84) in the upper part placing the inside of the external tank in communication with the outside;
- the electrolyser comprises a pipe (80) having a first connection (81) connected to the external tank (8) and a second connection (82) connected to the water supply volume (210) at least at one end upper part of the anode (312), the first connection being located below the second connection;
-said water supply volume (210) comprises an exhaust circuit (211) putting the upper part of the water supply volume into communication with the outside, the exhaust circuit (211) being sized to check the following relationship:
Pcech ≥ Pccon + (ρ*g*hv), with Pcech the pressure drop of the oxygen flow through the exhaust circuit (211), Pccon the pressure drop of the oxygen flow through said pipe (80), ρ the density of water, g the earth's gravity, and hv the difference in height between the first connection (81) and the maximum water level in said external tank (8). Water electrolyser (2) according to claim 1, in which said opening (84) of the external tank (8) is configured to allow water to be filled in the external tank (8). Water electrolyser (2) according to claim 1 or 2, wherein said outer tank (8) has a transparent wall (83) extending over a height between the first and second orifices (81, 82). Water electrolyzer (2) according to any one of the preceding claims, in which the water supply volume (210) is delimited by an upper wall (212), the exhaust circuit (211) being constituted of a bore passing through the upper wall. Water electrolyser (2) according to Claims 3 and 4, in which the external tank (8) comprises a marker of a maximum filling level, the marker being placed at least 10mm below the said through-bore (211). Water electrolyzer (2) according to Claim 4 or 5, in which the said bore is provided in a removable element of the upper wall (212) Water electrolyser (2) according to any of claims 4 to 6, wherein said bore is cylindrical. Water electrolyzer (2) according to any one of the preceding claims, in which the first connection (81) is located at the level of the lower part of the outer tank (8). A water electrolyser (2) according to any preceding claim, further including a power supply (7) configured to apply electrical power or current or potential difference between said first and second faces of the membrane assembly /electrodes (31) so as to hydrolyze water present in said supply volume (212). Water electrolyser (2) according to claim 9, in which the power supply (7) is configured to generate a volume flow rate of dioxygen Qg, and in which the pressure drop Pcech is dimensioned so that Qe<Qg/3 , where Qe is the volume flow rate of oxygen through the exhaust circuit (211). Water electrolyser (2) according to claim 9 or 10, in which the exhaust circuit (211) is sized to allow balancing of the water levels of the external tank (8) and of the water supply volume ( 210) in a period of less than 5s after the shutdown of said power supply (7). Burner (1), characterized in that it includes a water electrolyzer (2) according to any one of the preceding claims, further comprising a burning head (19) connected to the dihydrogen collector and configured to ignite the dihydrogen from the collector.