Patent Grant
11302942
The present invention relates to a method for detecting leakage of a reducing fluid throughout an electrolyte membrane of an electrochemical cell, the membrane being interposed between an anode channel for the flow of an anode stream, comprising the reducing fluid, along a first side of the membrane, and a cathode channel for the flow of a cathode stream, comprising the oxidizing fluid, along a second side of the membrane, the electrochemical cell comprising means for the flow of electric current between the two sides of the membrane.
This electrochemical cell is for example part of a stack of electrochemical cells of a fuel cell system.
Fuel cell systems are known comprising stacks of electrochemical cells each comprising a cathode channel, an anode channel, and an electrolyte membrane interposed between the two channels. The electrolyte membrane is designed to serve as a barrier between a reducing fluid flowing in the anode channel and an oxidizing fluid flowing in the cathode channel. The electrolyte membrane is nevertheless suitable for being crossed through by protons, so as to allow the occurrence of an oxidation-reduction reaction in the electrochemical cell between the reducing fluid and the oxidizing fluid.
This oxidizing-reducing reaction generates an electric current that is collected at the ends of each stack, and is used to supply a charge.
However, the electrolyte membranes of the cells are generally not perfectly tight with respect to the reducing fluid, and that tightness tends to decrease as the membrane ages. It was thus observed, in the known fuel cell systems, that after a certain operating time of each system, the electrolyte membranes of certain cells of the system were becoming permeable to the reducing fluid, which was causing a significant decrease in the performance of the fuel cell system. For example, for a proton exchange membrane (PEM) of the Nafion 112 type having a thickness of approximately 50 μm, the tightness level is considered abnormal when the so-called “crossover” current, i.e., the permeation current of the hydrogen from the anode channel toward the cathode channel, becomes greater than an equivalent current of 1 mA/cm2.
As a general rule, the loss of sealing of the electrolyte membrane is very disparate and only affects certain cells of the fuel cell system. It is therefore desirable to be able to identify the relevant cells, so as to be able to replace them. To that end, leakage detection methods have been developed in which the voltage across the terminals of each cell of the fuel cell system is monitored, so as to identify the voltage drops across the terminals of the cells.
However, a voltage drop across the terminals of a cell may have causes other than the mere loss of sealing of the membrane, for example fluid swelling of the cell by water accumulation in the anode or cathode channel, or local drying of the membrane of the cell, or a short-circuit phenomenon across the terminals of the cell.
Furthermore, a significant loss of tightness at the cell does not systematically lead to a drop in the voltage of the cell.
To eliminate this ambiguity, a method has been developed for detecting a leak in which the monitoring of the voltage across the terminals of the cells of the fuel cell system is coupled to a measurement of the reducing fluid concentration in the outgoing stream of a cathode collector of a stack of cells, said collector collecting the streams exiting the cathode channels of all of the cells of the stack. Thus, if a voltage drop across the terminals of a cell from the stack is observed at the same time as an increase in the measured concentration of reducing fluid, it is deduced that the voltage drop is originating from a loss of tightness of the electrolyte membrane of the cell.
Such a method is known from U.S. Pat. No. 5,763,113.
However, this method does not make it possible to identify small variations in the permeability of the membranes of the cells in the stack. Furthermore, this method requires keeping voltage sensors to measure the voltage across the terminals of each cell, which makes the fuel cell system expensive and complex.
One aim of the invention is to provide a method suitable for identifying small variations in the permeability of the electrolyte membrane. Another aim is for this method to be able to be implemented on a simple and cost-effective fuel cell system.
To that end, a method of the aforementioned type is provided, comprising the following successive steps:
According to preferred embodiments of the invention, the method also has one or more of the following features, considered alone or according to any technically possible combination(s):
The invention also relates to a fuel cell system, comprising at least one electrochemical cell, the or each electrochemical cell comprising:
According to preferred embodiments of the invention, the fuel cell system also has one or more of the following features, considered alone or according to any technically possible combination(s):
Other features and advantages of the invention will appear upon reading the following description, provided solely as an example and done in reference to the appended drawings, in which:
The fuel cell system 12, shown in
It will be noted that this number of two stacks 14A, 14B is purely arbitrary, and that the fuel cell system 12 may comprise any number of stacks 14A, 14B, without going beyond the scope of the invention.
Likewise, although in the illustrated example each stack 14A, 14B comprises five cells 15, this number is purely arbitrary. Alternatively, the number of cells 15 per stack 14A, 14B is less than or greater than five.
Each cell 15 comprises a membrane-electrode assembly 16 inserted in the longitudinal direction between an anode conductive plate 18 and a cathode conductive plate 20.
The membrane-electrode assembly 16 comprises an electrolyte membrane 22 sandwiched in the longitudinal direction between an anode 24a and a cathode 24b.
The membrane 22 separates the oxidizing and reducing fluids.
The membrane 22 is generally a proton-conducting membrane, suitable for only allowing protons to cross through it. In particular, the membrane 22 forms a barrier to free electrons. Thus, it electrically isolates the anode 24a from the cathode 24b, and the anode plate 18 from the cathode plate 20.
The membrane 22 is typically made from a polymer material.
The anode 24a and the cathode 24b each comprise a catalyzer, typically made from platinum or a platinum alloy, to facilitate the reaction. They are positioned together on either side of the membrane 22, and together define an active zone of the cell 15 where the electrochemical oxidizing-reducing reaction occurs.
The anode plate 18 delimits an anode channel 30 for the flow of an anode stream, comprising the reducing fluid, along and in contact with the anode 24a. To that end, the plate 18 is provided with at least one channel arranged in the face of the plate 18 turned toward the membrane-electrode assembly 16 and closed by said membrane-electrode assembly 16. The anode plate 18 is formed from an electrically conductive material, typically a composite made up of a graphite-filled polymer. The reducing fluid is for example dihydrogen.
The cathode plate 20 delimits a cathode channel 32 for the flow of a cathode stream comprising the oxidizing fluid along and in contact with the cathode 24b. To that end, the plate 20 is provided with at least one channel arranged in the face of the plate 20 facing the membrane-electrode assembly 16 and closed by said membrane-electrode assembly 16. The cathode plate 20 is formed by an electrically conductive material, typically a composite made up of a graphite-filled polymer. The oxidizing fluid is for example dioxygen.
The anode 24 is in electric contact with the anode plate 18. The cathode 24b is in electric contact with the cathode plate 20. The oxidation of the reducing fluid occurs and the electrons and protons are generated at the anode 24a. The electrons next pass through the anode plate 18 toward the cathode 24b of an adjacent cell 15, to participate in the reduction of the oxidizing fluid in the adjacent cell 15.
In each stack 14A, 14B, the anode plate 18 of each cell 15 of the stack 14A, 14B is in contact with the cathode plate 20 of the adjacent cell 15. The conductive plates 18, 20 thus ensure the transfer of electrons from the reducing fluid flowing in one of the cells 15 of the stack 14A, 14B to the oxidizing fluid flowing in another cell 15 of the stack 14A, 14B. Preferably, a channel (not shown) for the flow of a cooling fluid is formed at the interface between the anode 18 and cathode 20 plates.
Alternatively, the anode 18 and cathode 20 plates of two adjacent cells 15 of the stack 14A, 14B are integral and together form a bipolar plate.
The cell 15 further comprises seals 34, 36 to provide tightness between the conductive plates 18, 20 on the one hand and the membrane-electrode assembly 16 on the other hand. A first seal 34 is inserted in the longitudinal direction between the anode conductive plate 18 and the membrane 22, and a second seal 36 is inserted in the longitudinal direction between the cathode conductive plate 20 and the membrane 22, at the first seal 34. Each seal 34, 36, respectively, extends from the anode 24a, the cathode 24b, respectively.
The cells 15 are kept stacked owing to two gripping plates (not shown) positioned at the longitudinal ends of the stack 14A, 14B. Tightening bolts (not shown) exert a tightening force on said plates to keep them compressed against the cells 15.
The cathode channel 32 of each cell 15 of each stack 14A, 14B emerges in a cathode outlet collector 40 of the stack 14A, 14B. The cathode outlet collector 40 is suitable for collecting the cathode stream leaving each cell 15 of the stack 14A, 14B.
The cathode outlet collector 40 extends in the longitudinal direction of the stacks 14A, 14B. It is closed at a longitudinal end 42, and emerges by an opposite longitudinal end 44 in a discharge channel 46 of the cathode streams leaving the stack 14A, 14B.
The discharge channel 46 fluidly connects the cathode outlet collector 40 to a vent 48. In the illustrated example, the vent 48 is shared by the two stacks 14A, 14B.
The fuel cell system 12 also comprises a supply source 49A for supplying each cell 15 with reducing fluid, a valve 49B for regulating the flow rate and the supply pressure of the reducing fluid, a source 50 for supplying each cell 15 with oxidizing fluid, and a valve 51 for regulating the flow rate and the supply pressure of oxidizing fluid.
The supply source 49A is typically a reducing fluid reservoir.
The supply source 50 is typically an air compressor or an oxidizing fluid reservoir.
The fuel cell system 12 also comprises two channels 52 for supplying oxidizing fluid, each specific to a respective stack 14A, 14B. Each supply channel 52 fluidly connects the supply source 50 to a cathode inlet collector 53 (
The cathode inlet collector 53 fluidly communicates with the cathode channel 32 of each cell 15 of the stack 14A, 14B, and is suitable for supplying said cathode channel 32 with oxidizing fluid.
The fuel cell system 12 additionally comprises a recirculation device 54 for the cathode stream. The device 54 is suitable for removing part of the cathode streams leaving each stack 14A, 14B, and reinjecting it at the inlet of each stack 14A, 14B.
To that end, the reinjection device 54 comprises a reinjection channel 56, fluidly connecting the vent 48 to each supply channel 52, and a pump 57, to force the flow of fluid in the channel 56 from the vent 48 toward each supply channel 52.
It will be noted that in
The stacks 14A, 14B are electrically connected to one another in series using a first electrical connection 58. A second electrical connection 59A electrically connects the stacks 14A, 14B to a charge 59B. A switch 59C, on the line 59A, is suitable for selectively connecting or disconnecting the charge 59B with respect to the stacks 14A, 14B.
The fuel cell system 12 is suitable for generating a nominal charge current in the second electrical connection 59A when:
The fuel cell system 12 lastly comprises a first sensor 60 (
The first sensor 60 is suitable for communicating with a leakage detection module 61, suitable for deducing the presence of leakage in a cell 15 of the system 12, when the first concentration C1 exceeds a threshold concentration.
In the first version shown in
To that end, said cell 15 comprises, in reference to
The bypass channel 62 is suitable for the fluid flowing inside to have a lower flow rate than the fluid flowing in the cathode channel 32.
The first sensor 60 comprises a first electrode 64, positioned in the bypass channel 62, and a second electrode 66, positioned in a connecting zone 68 connecting the cathode channel 32 to the outlet collector 40. It also comprises a resistance 70, electrically connected by one terminal to the first electrode 64, and by the other of its terminals to the second electrode 66. It lastly comprises a filter 72, to measure a voltage across the terminals of the resistance 70, and a module 74 for deducing the concentration C1 as a function of the voltage measured by the filter 72.
Still in this first alternative, each other cell 15 of the fuel cell system 12 also comprises a cathode bypass channel similar to the cathode bypass channel 62, and the sensor, identical to the first sensor 60, is associated with each of these other cells 15 to measure a reducing fluid concentration in the cathode stream leaving said cell 15.
The first sensor 60 and the other sensors are suitable for communicating with the leakage detection module 61, which is suitable for deducing the presence of a leakage in a cell 15 of the system 12 when the reducing fluid concentration measured by the sensor associated with said cell 15 exceeds a threshold concentration.
Thus, it is possible to detect, for each cell 15, whether an abnormal elevation of the reducing fluid concentration in the cathode stream leaving said cell 15 is occurring. It is therefore easily possible to identify a loss of tightness of the membrane 22 of a cell 15, and to ensure very precise monitoring of the aging of the membrane 22 of each cell 15.
However, such a solution is expensive, inasmuch as it requires integrating a sensor in each cell 15 of each stack 14A, 14B. Furthermore, such a solution is complex, inasmuch as it requires managing a large number of reducing fluid sensors to monitor the cells 15.
In the second, third and fourth alternatives, shown in
In the second and third alternatives, shown in
The second sensor 80 is also suitable for communicating with the detection module 61.
In the second alternative, shown in
The detection module 61 is suitable for comparing the first concentration C1 to the second concentration C2, and for deducing the presence of a leakage in one of the stacks 14A, 14B when one of the following conditions is met:
The detection module 61 is suitable for locating the leakage in the stack 14A, 14B associated with the sensor 60, 80 having measured the highest concentrations C1, C2, or having the largest deviation relative to time.
The detection module 61 is also suitable for locating the leakage among the cells 15 of said stack 14A, 14B as a function of the delay between a moment t0 (
“Brisk variation” means that, over a length of time greater than or equal to 1 second, the deviation relative to time of the parameter in question is greater than:
“Significant variation” means that the deviation relative to time of the measured concentrations C1, C2 is greater than 40 ppm per second, which represents 0.5% of the flammability limit of the hydrogen in the oxygen.
In the third alternative, shown in
Each second measuring point 82 is upstream from the corresponding first measuring point, in a fluid flow direction in the associated cathode outlet collector 40. Each second measuring point 82 is in particular positioned substantially mid-distance with respect to the longitudinal ends of the stack 14A, 14B, such that it has the same number of cells 15 of the stack 14A, 14B whereof the cathode channel 32 emerges in the cathode outlet collector 40 upstream and downstream from the second measuring point 82.
To that end, the fuel cell system 12 comprises two multi-channel valves 84, 86. A first 84 of the multi-channel valves 84, 86 comprises a first inlet 88A fluidly connected to the first measuring point of the stack 14A, a second inlet 88B fluidly connected to the second measuring point 82 of the stack 14A, and an outlet 88C fluidly connected to the sensor 60. A second 86 of the multi-channel valves 84, 86 comprises an inlet 90A fluidly connected to the first measuring point of the stack 14B, a second inlet 90B fluidly connected to the second measuring point 82 of the stack 14B and an outlet 90C fluidly connected to the sensor 80.
The detection module 61 is suitable for comparing the concentrations C1, C2, C3, C4 to one another, and deducing the presence of leakage of one of the stacks 14A, 14B when one of the following conditions is met:
The detection module 61 is suitable for locating the leakage in:
The detection module 61 is suitable for locating the leakage among the cells 15 of the identified stack as a function of the delay between a moment t0 (
In the fourth alternative, shown in
The multi-channel valve 92 is programmed to successively connect each inlet 94A, 94B, 94C, 94D to the outlet 94E, at constant time intervals.
The detection module 61 is suitable for comparing the concentrations C1, C2, C3, C4 to one another, and deducing the presence of leakage of one of the stacks 14A, 14B when one of the following conditions is met:
The detection module 61 is suitable for locating the leakage in:
The detection module 61 is suitable for locating the cell having the leakage from among the cells 15 of the identified stack half as a function of a delay between a moment t0 (
A method for detecting a leakage using the fuel cell system 12 according to the fourth embodiment will now be described, in light of
According to this method, the stacks 14A, 14B are supplied with a cathode stream and an anode stream. For each cell 15 of each stack 14A, 14B:
In a first step for monitoring the fuel cell system 12, the sensor 60 measures the concentrations C1, C2, C3, C4 at the corresponding measuring points. To that end, the multi-channel valve 92 successively connects each inlet 94A, 94B, 94C, 94D to the outlet 94E, at constant time intervals.
The detection module 61 determines the evolution over time of each concentration C1, C2, C3, C4 as well as the differences between those concentrations in pairs.
The detection module 61 determines the presence of a leakage in one of the stacks 14A, 14B when one of the following conditions is met:
When a leakage is thus detected, a second step for locating the leakage is carried out following the first monitoring step. During the second step, the detection module 61 locates the stack half in which the leakage is located based on the condition set out above that has been met. The detection module locates the leakage in:
In the example shown in
Following the second step for locating the leakage, there is a third step for identifying the leaking cell 15. During this step, the multi-channel valve 92 fluidly connects, to the sensor 60, the measuring point that corresponds to the stack half in which the leakage has been located, to measure the concentration of interest continuously and to individually locate the cell 15 with the leakage.
In particular, the multi-channel valve 92 connects, to the sensor 60:
In the cited example, the multi-channel valve 92 therefore connects the sensor 60 to the first measuring point of the first stack 14A.
At least one parameter is then varied briskly and in a controlled manner, in the stack in which the leakage is detected (i.e., in the cited example, the first stack 14A), at a moment t0, said parameter being chosen from among the following parameters:
Preferably, at least two parameters are varied simultaneously or sequentially from among the preceding parameters.
In the example shown in
Thus, if the moment t1 is close to the moment t0, the detection module 61 identifies the leaking cell 15 as being close to the upstream end of the stack 14A, 14B, and if the moment t1 is far from the moment t0, the detection module 61 identifies the leaking cell 15 as being close to the downstream end of the stack 14A, 14B. One skilled in the art will know how to calibrate the detection module 61 so as to precisely identify the cell 15 affected by the leak as a function of the delay observed between the moments t0 and t1.
It will be noted that this method is easy to adapt to the fuel cell system according to the third embodiment of the invention, illustrated in
The method can also be easily adapted to the fuel cell system according to the second embodiment of the invention, illustrated in
The leakage detection method implemented using the fuel cell system according to the first embodiment of the invention, illustrated in
pressure of the anode stream in the anode channel 30 of the or each cell 15;
pressure of the cathode stream in the cathode channel 32 of the or each cell 15,
flow rate of the anode stream in the anode channel 30 of the or each cell 15;
flow rate of the cathode stream in the cathode channel 32 of the or each cell 15; to that end, the recirculation pump 57 is for example activated; and
intensity of the current exchanged between the two sides of the membrane 22 of the or each cell 15.
Preferably, at least two parameters among the preceding parameters vary simultaneously or sequentially during this step.
This brisk variation increases the permeation rate of the reducing fluid through the membrane 22 of each cell 15. Each sensor measures the reducing fluid concentration in the cathode stream leaving the associated cell 15, and the detection module 61 compares the measured concentrations in pairs, as well as their respective evolutions over time.
The detection module 61 detects the presence of a leakage of one of the cells 15 when one of the following conditions is met:
one of the measured concentrations exceeds the average of the measured concentrations by a predetermined threshold, or
the deviation relative to time of the difference between one of the measured concentrations and the average of the measured concentrations exceeds a predetermined threshold, said measured concentration being higher than the average of the measured concentrations.
The detection module 61 identifies the leaking cell 15 as being the cell 15 in which said concentration was measured.
Owing to embodiments of the invention, the leakage detection method, which uses an brisk variation of a relevant parameter combined with a concentration measurement of a stream of reducing fluid, makes it possible to identify leakages, even small ones, in a membrane of an electrochemical cell.
Furthermore, the detection method makes it possible to detect a leakage in a stack of electrochemical cells, and to next precisely and individually detect the or each electrochemical cell exhibiting the leakage.
The method may be implemented reliably and precisely on a simple fuel cell system. In particular, the method makes it possible to identify an electrochemical cell having a leakage from among the electrochemical cells of one or more stacks with a reduced number of reducing fluid concentration sensors in a fluid, in particular with a single sensor.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1350793 | Jan 2013 | FR | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2013/076126 | 12/10/2013 | WO | 00 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2014/117893 | 8/7/2014 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5763113 | Meltser et al. | Jun 1998 | A |
| 20090226773 | Takekawa | Sep 2009 | A1 |
| Number | Date | Country |
|---|---|---|
| 0 827 226 | Mar 1998 | EP |
| 0827226 | Mar 1998 | EP |
| 1942541 | Jul 2008 | EP |
| H03101061 | Apr 1991 | JP |
| H07169488 | Jul 1995 | JP |
| H11 67255 | Mar 1999 | JP |
| 2005-150007 | Jun 2005 | JP |
| 2005150007 | Jun 2005 | JP |
| 2006-339080 | Dec 2006 | JP |
| A2006339080 | Dec 2006 | JP |
| 2007-048540 | Feb 2007 | JP |
| 2007048540 | Feb 2007 | JP |
| 2007059120 | Mar 2007 | JP |
| A2007157544 | Jun 2007 | JP |
| WO 2004027369 | Apr 2004 | WO |
| Entry |
|---|
| Search Report for corresponding International Application PCT/EP2013/076126. |
| Number | Date | Country | |
|---|---|---|---|
| 20150357661 A1 | Dec 2015 | US |
