Method for detecting a leak in a membrane in a fuel cell stack of a fuel cell system, computer program product, control unit, fuel cell system, and fuel cell vehicle

Abstract

The invention relates to a method for detecting a leak in a membrane in a fuel cell stack of a fuel cell system (FCS), wherein the fuel cell system (FCS) can have at least one or more fuel cell stacks, comprising: - carrying out a passive bleed-down process, - measuring a pressure (PAnode, PKathode, PStack) in the fuel cell stack, - analyzing a change over time and / or a gradient of the pressure (PAnode, PKathode, PStack), - determining a leak in the membrane on the basis of the analysis.

Description

[0001] R. 416940

[0002] - 1 -

[0003] Description

[0004] title

[0005] Method for detecting a leakage of a membrane in a fuel cell stack of a fuel cell system, computer program product, control unit, fuel cell system and fuel cell vehicle

[0006] The invention relates to a method for detecting a leakage in a membrane within a fuel cell stack of a fuel cell system. Furthermore, the invention relates to a corresponding computer program, a corresponding control unit, a corresponding fuel cell system, and a corresponding fuel cell vehicle for carrying out this method.

[0007] State of the art

[0008] In vehicles (Fuel Cell Vehicle or FCV for short) where propulsion energy is supplied (among other things) by one (or more) (PEM) fuel cell system(s) (Fuel Cell System or FCS for short), oxygen from the ambient air is usually used as the oxidizing agent and hydrogen as the reducing agent to react in a fuel cell stack to form water (or water vapor) and thus deliver electrical power through electrochemical conversion.

[0009] If a membrane in the fuel cell stack has a hole (larger than the size of an H2 molecule), hydrogen can reach the cathode and thus be released into the environment via the exhaust gas. If the hole exceeds a certain size, a flammable or explosive mixture can form. This must be avoided. R. 416940

[0010] - 2 -

[0011] Currently, a potential hole in the membrane is detected by monitoring the pressure in the stack (especially the anode) after the fuel cell system has been completely shut down. This allows a leak to be detected after a completed driving cycle.

[0012] Disclosure of the invention

[0013] The present invention provides a method for detecting a leakage in a membrane of a fuel cell stack of a fuel cell system, comprising the features of the independent method claim. Furthermore, the invention provides a corresponding computer program, a corresponding control unit, a corresponding fuel cell system, and a corresponding fuel cell vehicle, comprising the features of the dependent claims. Features and details described in connection with the different embodiments and / or aspects of the invention naturally also apply in connection with the other embodiments and / or aspects, and vice versa, so that the disclosure relating to the individual embodiments and / or aspects always includes, or can include, reciprocal references.

[0014] According to the first aspect, the present invention provides:

[0015] a method for detecting (or in other words, diagnosing) a leak in a membrane of a fuel cell stack of a

[0016] (PEM) fuel cell system.

[0017] The fuel cell system (which can be referred to as the system for short) can be designed with at least one fuel cell stack (which can be referred to as the stack for short) or several fuel cell stacks (which can be referred to as the stacks for short).

[0018] The procedure comprises the following steps:

[0019] (- possibly putting the fuel cell stack into standby mode,) Performing a passive bleed-down procedure, R. 416940

[0020] - 3 -

[0021] Capturing pressure in the fuel cell stack,

[0022] Evaluating a time course and / or gradient of pressure,

[0023] Determining membrane leakage depending on the evaluation.

[0024] The fuel cell system (FCS) can be used to provide electrical energy for a vehicle (Fuel Cell Vehicle or FCV).

[0025] The at least one fuel cell stack or the multiple fuel cell stacks may have associated functional systems, comprising: media systems (air or cathode system, hydrogen or anode system, cooling system) and an electrical system.

[0026] Preferably, the fuel cell system can comprise several modules in the form of individual stacks and the associated functional systems.

[0027] A fuel cell stack can, for example, be put into a standby mode when no power is required from the fuel cell stack (e.g., briefly).

[0028] If, for example, the fuel cell system has several fuel cell stacks and the power requirement for the entire system is reduced, then one of the multiple fuel cell stacks can be put into a standby mode.

[0029] The affected stack can remain in standby mode for a short or longer period of time, but can be put into operation at any time.

[0030] Standby mode does not refer to a complete system shutdown, but rather to maintenance measures such as system drying. R. 416940

[0031] - 4 -

[0032] The invention recognizes that if the pressure in the stack (especially the anode) is monitored after a complete shutdown of the system, it may happen that the system operates with a leakage for an entire driving cycle.

[0033] Therefore, it is suggested here to use standby mode in order to carry out appropriate diagnostics early (already during an operating cycle of the system or a driving cycle of the vehicle).

[0034] The standby mode can

[0035] 1) a passive bleed-down process

[0036] 2) and possibly a preceding (comparatively short, e.g. of a few seconds) active bleed-down process

[0037] exhibit.

[0038] In an active bleed-down process, a cathode region of the stack is closed off, and the hydrogen in the anode region is recirculated to consume the oxygen in the cathode region. During this process, electrical current can be drawn from the stack until a certain system voltage is reached, for example, until the system voltage drops to 0.8 volts.

[0039] In a passive bleed-down process, the stack current is zero. The cathode and anode valves are closed, so the stack is a closed system.

[0040] One advantage of performing diagnostics during standby mode is that the diagnostics can be carried out while the system or vehicle is operating normally. This means the system does not need to take any shutdown measures and the vehicle does not need to go to a workshop.

[0041] Advantageously, a diagnosis can therefore be performed more frequently within a system operating cycle or a vehicle driving cycle. R. 416940

[0042] - 5 -

[0043] The idea is that during a passive bleed-down process, the following pressure profiles in the stack can be evaluated to detect a leak.

[0044] During the passive bleed-down process, the oxygen on the cathode side of the stack reacts with the hydrogen diffusing from the anode side to the cathode side. This causes both the anode and cathode pressures to drop until a certain initial point at which the oxygen in the cathode region is depleted.

[0045] From then on, only a very small fraction of the hydrogen reacts with any subsequently diffusing oxygen. The average stack pressure then decreases only slowly.

[0046] After the first determined point in time, the cathode pressure initially rises further due to the hydrogen diffusing from the anode area.

[0047] Simultaneously, nitrogen diffuses from the cathode side to the anode side of the stack. Since the diffusion of nitrogen (from cathode side to anode side) is slower than the diffusion of hydrogen (from anode side to cathode side), a local maximum of the cathode pressure occurs at a specific second time point, and a minimum of the anode pressure occurs at a specific third time point.

[0048] Afterwards, the pressure profiles in the anode region and the cathode region approach each other and eventually slowly decrease (due to the reaction of diffusing oxygen and hydrogen and because the temperature in the stack falls).

[0049] The invention recognizes that different pressure gradients in the stack (e.g. in the anode region, in the cathode region and / or an average pressure in the stack) can indicate a leakage, e.g. a hole, in the membrane.

[0050] The advantage of a diagnosis during a passive bleed-down procedure can lie particularly in the fact that the pressure gradient in the anode area is greater. R. 416940

[0051] - 6 -

[0052] This is after an active bleed-down process, which is carried out down to a low voltage. This allows tolerances of pressure sensors to be compensated for.

[0053] The reason for this is that oxygen is still present on the cathode, which reacts with the hydrogen passing from the anode. This keeps the driving force for diffusion from anode to cathode high, as the hydrogen partial pressure difference across the membrane remains high. While the hydrogen partial pressure in the anode decreases, the hydrogen partial pressure on the cathode remains at zero until the oxygen is consumed.

[0054] This method could involve measuring and analyzing the pressure in an anode area of ​​the fuel cell stack. Diagnosis by evaluating the pressure's temporal profile and / or gradient in the anode area can be advantageous because the pressure gradient is easily measurable. Furthermore, such a diagnosis can be beneficial because a leaking membrane causes the pressure to drop more rapidly than an intact membrane, and / or because the pressure gradient measurably increases.

[0055] The anode pressure gradient in a passive bleed-down process can be highly dependent on the exchange across the membrane. If a hole is present in the membrane, the pressure gradient increases. Empirical or analytical approximations can be used to evaluate the gradient, allowing the determination of the correct gradient, its time course, and its limits for different boundary conditions. The advantage of the passive bleed-down process lies in the fact that, during the phase in which oxygen is still present on the cathode side of the stack, the corresponding boundary conditions (e.g., the hydrogen partial pressure on the cathode side is zero) are easy to determine.

[0056] Furthermore, it is conceivable that a pressure at an input to an anode region of the fuel cell stack is detected and evaluated, and / or that a pressure at an output from an anode region of the fuel cell stack is detected. R. 416940

[0057] - 7 -

[0058] The data is recorded and evaluated. In this way, existing pressure sensors at the inlet to and / or at the outlet from the anode area can be used flexibly.

[0059] This method could involve measuring and analyzing the pressure in a cathode region of the fuel cell stack. Diagnosis by evaluating the pressure's temporal profile and / or gradient in the cathode region could be advantageous because the pressure gradient there is easily measurable, at least as long as oxygen is present.

[0060] The cathode pressure gradient in a passive bleed-down process can depend on how much hydrogen diffuses onto the cathode and reacts with the oxygen there. This results in a pressure drop in the cathode region. This pressure drop can also be evaluated until the oxygen in the stack is consumed, or subsequently, the corresponding pressure rise. Here, too, empirical or analytical approximations can be used to evaluate the gradient, making it possible to determine the correct gradient or time course, or its limits, for different boundary conditions. The advantage of the passive bleed-down process lies in the fact that, during the phase in which oxygen is still present on the cathode side of the stack, the relevant boundary conditions (e.g., the hydrogen partial pressure on the cathode side is zero) can be easily determined.

[0061] In this method, it is conceivable that an average pressure between a cathode region and an anode region of the fuel cell stack is recorded and evaluated. The average pressure can preferably be determined as a function of the volume of a cathode region and / or as a function of the volume of an anode region of the fuel cell stack. The average pressure depends primarily on the reaction of hydrogen and oxygen at the cathode and thus on the exchange across the membrane. The average pressure can therefore very accurately reflect the condition of the membrane. A further advantage of this parameter is that membrane deformation plays no role when the pressure conditions in the anode region are stable. R. 416940

[0062] - 8 -

[0063] and reverse in the cathode region from a certain initial point in time.

[0064] Furthermore, it is conceivable that a limit value, particularly a maximum value, for the pressure gradient is determined when evaluating the pressure. Additionally, it can be provided that a membrane leak is detected when the pressure exceeds a limit value, particularly a maximum value. In this way, simple and reliable detection of a membrane leak in the stack can be enabled.

[0065] In principle, it is conceivable that a limit value, particularly a maximum value, for the pressure gradient could be determined empirically or analytically, for example, using a model. Empirically, different operating points in the stack (e.g., with regard to temperature, hydrogen partial pressure, humidity, etc.) can be explored during a passive bleed-down process to determine the correct gradient or time course, or the limits for such a gradient or time course. Analytically, models for hydrogen exchange across the membrane can be used, which can, for example, take into account different operating points in the stack (e.g., with regard to temperature, hydrogen partial pressure, humidity, etc.).

[0066] Advantageously, a limit value, in particular a maximum value, for the pressure gradient can be determined as a function of at least one operating parameter of the fuel cell stack:

[0067] - Temperature,

[0068] - Hydrogen partial pressure and / or

[0069] - Moisture.

[0070] In this way, different operating points in the stack can be flexibly considered in order to determine a correct gradient or a correct time course, or the limits for a correct gradient or a correct time course. R. 416940

[0071] - 9 -

[0072] Furthermore, it can be provided that, if the determination of leakage is dependent on a pressure in an anode region and / or a pressure in a cathode region of the fuel cell stack, the evaluation is performed before the complete consumption of oxygen in the cathode region of the fuel cell stack. An evaluation based on the anode or cathode gradient can be improved by this, since at the moment the oxygen in the cathode region is consumed, the force conditions at the membrane reverse (cathode pressure becomes greater than anode pressure) and the membrane shifts, which can lead to pressure changes in the anode and the cathode (which are not visible in the stack pressure). These pressure changes are unrelated to the mass transfer between the anode and cathode being determined.

[0073] Furthermore, it can be provided that, if a recirculation pump is still rotating in an anode path during a passive bleed-down process, an average pressure between an inlet and an outlet pressure from an anode section of the fuel cell stack is recorded and evaluated. The recirculation pump may have a run-on time and not stop immediately (the recirculation pump coasts to a halt). As long as the recirculation pump is rotating, there is flow in the anode recirculation path. An average between the two anode pressures (before and after the stack) can be advantageous to reduce the influence of the changing flow.

[0074] According to a further aspect, the invention provides a computer program product comprising instructions which, when executed by a computer, such as the processing unit of a control unit, cause the computer to carry out the method, which can proceed as described above. The same advantages described above in connection with the method according to the invention can be achieved with the computer program product. These advantages are fully referenced herein.

[0075] A corresponding control unit provides a further aspect of the invention. A computer program can be stored in a memory unit of the control unit. R. 416940

[0076] - 10 -

[0077] A code is stored in the form of a process which, when executed by a processing unit of the control unit, performs a procedure that can proceed as described above. The same advantages described above in connection with the method according to the invention can be achieved using the control unit. These advantages are fully referenced here. The proposed evaluation can be performed online or offline. In the latter case, raw signals can be transmitted to a backend device for evaluation. In other words, the evaluation can be outsourced from the system.

[0078] A corresponding fuel cell system (FCS) also constitutes an aspect of the invention. The fuel cell system can serve to provide electrical energy for a vehicle. The same advantages described above in connection with the method according to the invention can be achieved with the fuel cell system. These advantages are fully referenced herein.

[0079] A corresponding fuel cell vehicle (FCV) also represents an aspect of the invention. The same advantages described above in connection with the method according to the invention can be achieved with the fuel cell vehicle. These advantages are fully referenced herein.

[0080] Preferred embodiments:

[0081] The invention, its further developments, and its advantages are explained in more detail below with reference to the drawings. Each drawing schematically shows:

[0082] Figure 1 shows exemplary pressure profiles in a stack.

[0083] Figure 2 shows exemplary pressure profiles in a stack.

[0084] Figure 3 shows exemplary pressure gradients for printing examples of Figure 2, andR. 416940

[0085] - 11 -

[0086] Figure 4 shows an enlarged representation of exemplary pressure gradients of Figure 2.

[0087] Figures 1 to 4 serve to explain the invention, which, according to the first aspect, provides a method for detecting (or in other words, diagnosing) a leakage of a membrane in a fuel cell stack of a (PEM) fuel cell system FCS.

[0088] The fuel cell system FCS (or simply system) can be configured with at least one or more fuel cell stacks (or simply stack(s)).

[0089] The procedure comprises the following steps:

[0090] (- preferably putting the fuel cell stack into a standby mode,)

[0091] Performing a passive bleed-down procedure,

[0092] Measuring the pressure PAnode, Pcathode, Pstack in the fuel cell stack, evaluating a time course and / or gradient of the pressure PAnode, Pcathode, Pstack,

[0093] Determining membrane leakage depending on the evaluation.

[0094] The fuel cell system FCS (English: “Fuel Cell System” or FCS for short) can be specifically designed for mobile applications, e.g. in a vehicle (English: “Fuel Cell Vehicle” or FCV for short).

[0095] The system can include at least one fuel cell stack or several fuel cell stacks with their associated functional systems, comprising: media systems (air or cathode system, hydrogen or anode system, cooling system) and an electrical system.

[0096] Preferably, the fuel cell system (FCS) can comprise several modules in the form of individual stacks with the associated functional systems.

[0097] A standby mode can be advantageous, for example, when no power is required from the fuel cell stack(s) (or one of the several) (e.g., temporarily). R. 416940

[0098] - 12 -

[0099] If the fuel cell system (FCS) has, for example, several fuel cell stacks and the power requirement for the entire system is reduced, then at least one of the multiple fuel cell stacks can be put into a standby mode.

[0100] The affected stack can remain in standby mode for a short or longer period of time, but can be put into operation at any time, since a complete shutdown is not intended in standby mode, which would entail corrective measures such as drying the system. Furthermore, the procedure for determining the leakage has no influence on the standby mode.

[0101] It is suggested here to use the standby mode to diagnose a membrane leak early (already during an operating cycle of the system or a driving cycle of the vehicle).

[0102] The standby mode may include a passive bleed-down process. This may be preceded by a (relatively short, e.g., a few seconds) active bleed-down process.

[0103] In a passive bleed-down process, no electrical current is drawn from the stack. The cathode and anode valves remain closed, so the stack forms a closed system.

[0104] One advantage of performing diagnostics during standby mode is that diagnostics can be carried out more frequently within an operating cycle of the system or a driving cycle of the vehicle.

[0105] As Figures 1, 2, and 4 indicate, the following pressure profiles in the stack can be evaluated for leak detection during a passive bleed-down process. R. 416940

[0106] - 13 -

[0107] During the passive bleed-down process, the oxygen on the cathode side of the stack reacts with the hydrogen diffusing from the anode side to the cathode side. This causes both the anode and cathode pressures to decrease until a certain initial time point (see time 1 in Fig. 1), at which point the oxygen in the cathode region is depleted.

[0108] From then on, only a very small fraction of the hydrogen reacts with any subsequently diffusing oxygen. The average stack pressure then decreases only slowly (see Fig. 2 and Fig. 3).

[0109] After the first defined time point (see time point 1 in Fig. 1), the cathode pressure Pcathode initially increases due to the diffusion of hydrogen from the anode region. Simultaneously, nitrogen diffuses from the cathode side to the anode side of the stack. Since the diffusion of nitrogen (from cathode side to anode side) is slower than the diffusion of hydrogen (from anode side to cathode side), a local maximum of the cathode pressure Pcathode occurs at a specific second time point (see time point 2 in Fig. 1), and a minimum of the anode pressure Panode occurs at a specific third time point (see time point 1 in Fig. 1).

[0110] Afterwards, the pressure profiles in the anode region and the cathode region approach each other and eventually slowly decrease (due to the reaction of diffusing oxygen and hydrogen and because the temperature in the stack falls).

[0111] The different pressure gradients in the stack (e.g. in the anode area, in the cathode area and / or an average pressure in the stack) can indicate a leak, e.g. a hole, in the membrane.

[0112] In a passive bleed-down process, the pressure gradient in the anode area is greater than after an active bleed-down process performed down to a low voltage. This allows for reliable detection where the tolerances of pressure sensors are irrelevant. R. 416940

[0113] - 14 -

[0114] As long as oxygen is present on the cathode, reacting with the hydrogen passing from the anode, the driving force for diffusion from anode to cathode remains relatively high, since the hydrogen partial pressure difference across the membrane remains high. Although the hydrogen partial pressure in the anode decreases, the hydrogen partial pressure on the cathode remains at zero until the oxygen is consumed.

[0115] This method allows for the measurement and evaluation of the anode pressure gradient in an anode region of the fuel cell stack. During passive bleed-down, the anode pressure gradient can be highly dependent on the exchange across the membrane. If a hole is present in the membrane, the pressure gradient increases. Empirical or analytical approximations can be used to evaluate the gradient, enabling the determination of the correct gradient, its temporal profile, and its limits for different boundary conditions. The advantage of passive bleed-down lies in the fact that, during the phase in which oxygen is still present on the cathode side of the stack, the corresponding boundary conditions (e.g., zero hydrogen partial pressure on the cathode side) are easy to determine.

[0116] As shown in Figures 1 to 4, a pressure PAnodeIn at an input to an anode area and / or a pressure PAnodeOut at an output from an anode area of ​​the fuel cell stack can be detected and evaluated. In this way, existing pressure sensors in the anode area can be used flexibly.

[0117] Furthermore, the method allows for the detection and evaluation of a cathode pressure (Pcathode) in a cathode region of the fuel cell stack, e.g., a cathode pressure (PcathodeE) at an inlet in a cathode region of the fuel cell stack. In a passive bleed-down process, the cathode pressure gradient can depend on how much hydrogen diffuses onto the cathode and reacts with the oxygen there. This results in a pressure drop in the cathode region. This pressure drop can also be evaluated until the oxygen in the stack is consumed, or subsequently, the corresponding Rc. 416940

[0118] - 15 -

[0119] Pressure increase in the cathode region. Here, too, empirical or analytical approximations can be used to evaluate the gradient, making it possible to determine the correct gradient or time course, or its limits, for different boundary conditions. Again, the advantage of the passive bleed-down process lies in the fact that, during the phase in which oxygen is still present on the cathode side of the stack, the corresponding boundary conditions (e.g., the hydrogen partial pressure on the cathode side is zero) are easy to determine.

[0120] The method also allows for the measurement and evaluation of an averaged pressure Pstack between a cathode region and an anode region of the fuel cell stack. The averaged pressure Pstack can preferably be determined as a function of a volume Vcathode of a cathode region and / or as a function of a volume Vanode of an anode region of the fuel cell stack, for example, as follows:

[0121] p > P anode -V anode + P cathode -V cathode

[0122] Stack TT TT

[0123] Anode Cathode

[0124] The average pressure depends primarily on the reaction of hydrogen and oxygen at the cathode and thus on the exchange across the membrane. The average pressure can therefore very accurately reflect the state of the membrane. A further advantage of this parameter is that membrane deformation plays no role when the pressure conditions in the anode and cathode regions reverse at a certain initial point in time (see point 1 in Fog. 1).

[0125] When evaluating the pressure PAnode, Pcathode, Pstack, a limit value, in particular a maximum value, for the pressure gradient of PAnode, Pcathode, Pstack must be determined. A membrane leak can then be detected if the pressure PAnode, Pcathode, Pstack exceeds this limit value, in particular this maximum value. R. 416940

[0126] - 16 -

[0127] The limit value, in particular the maximum value, for the gradient of the pressure P anode, P pathode, Pstack can be determined empirically or analytically, e.g. using a model.

[0128] Empirically, in a passive bleed-down process, different operating points in the stack (e.g., with regard to temperature, hydrogen partial pressure, humidity, etc.) can be approached in order to determine a correct gradient or a correct time profile, or the limits for a correct gradient or a correct time profile.

[0129] Analytically, models for hydrogen exchange across the membrane can be used, which can take into account, for example, different operating points in the stack (e.g., with regard to temperature, hydrogen partial pressure, humidity, etc.).

[0130] The limit value, in particular the maximum value, for the gradient of the pressure P anode, P pathode, Pstack can be determined, for example, depending on at least one operating parameter of the fuel cell stack, such as:

[0131] - Temperature,

[0132] - Hydrogen partial pressure,

[0133] - Moisture, etc.

[0134] As indicated in Fig. 4 with a drawn vertical line, it can be advantageous to evaluate the pressure PAnode in an anode region and / or the pressure Pcathode in a cathode region of the fuel cell stack before the complete consumption of oxygen in the cathode region of the fuel cell stack, i.e., before time 1. An evaluation based on the anode or cathode gradient can be improved because, at the moment the oxygen in the cathode region is consumed, the force relationships at the membrane reverse (cathode pressure Pcathode becomes greater than anode pressure PAnode) and the membrane shifts, which can lead to pressure changes in the anode and the cathode (which are not visible in the stack pressure). These pressure changes are unrelated to the mass transfer between the anode and cathode to be determined. R. 416940

[0135] - 17 -

[0136] As further indicated in Fig. 4, it can be advantageous to calculate an average between the pressure PAnodeBn at an inlet and the pressure PAnodeAus at an outlet of an anode region of the fuel cell stack when a recirculation pump is still rotating in an anode path during a passive bleed-down process. The recirculation pump may have a run-on time and not stop immediately (the recirculation pump coasts to a halt). As long as the recirculation pump is rotating, there is flow in the anode recirculation path. An average between the two anode pressures (before and after the stack) can be advantageous to reduce the influence of the changing flow.

[0137] A corresponding computer program product, a corresponding control unit ECU, a corresponding fuel cell system FCS and a corresponding fuel cell vehicle FCV represent further aspects of the invention.

[0138] The preceding description of the figures describes the present invention solely by way of examples. Naturally, individual features of the embodiments can be freely combined with one another, provided it is technically feasible, without departing from the scope of the invention.

Claims

R. 416940 - 18 - Claims 1. Method for detecting a leakage of a membrane in a fuel cell stack of a fuel cell system (FCS), wherein the fuel cell system (FCS) may be configured with at least one or more fuel cell stacks, exhibiting: - Performing a passive bleed-down procedure, - Measuring pressure (PAnode, Pcathode, Pstack) in the fuel cell stack, - Evaluating a time course and / or gradient of the pressure (PAnode, Pcathode? Pstack), - Determining membrane leakage depending on the evaluation.

2. Method according to claim 1 , where a pressure (PAnode) in an anode area of ​​the fuel cell stack is detected and evaluated.

3. Method according to claim 1 or 2, where a pressure (PAnodeIn) at an input to an anode area of ​​the fuel cell stack is detected and evaluated, and / or wherein a pressure (PAnodeOut) is detected and evaluated at an output from an anode area of ​​the fuel cell stack.

4. Method according to any one of the preceding claims, where a pressure (Pcathode) in a cathode region of the fuel cell stack, e.g. a pressure (PcathodeIn) at an input to a cathode region of the fuel cell stack, is detected and evaluated.

5. Method according to any of the preceding claims, R. 416940 - 19 - where an average pressure (Pstack) between a cathode area and an anode area of ​​the fuel cell stack is recorded and evaluated, in particular the average pressure (Pstack) is determined as a function of a volume (Vcathode) of a cathode area and / or as a function of a volume (VAnode) of an anode area of ​​the fuel cell stack.

6. Method according to any one of the preceding claims, wherein when evaluating the pressure (PAnode, Pcathode, Pstack) a limit value, in particular a maximum value, for the gradient of the pressure (PAnode, Pcathode, Pstack) is determined, and / or wherein a leakage of the membrane is determined when the pressure (PAnode, Pcathode, Pstack) exceeds a limit value, in particular a maximum value 7. Method according to any of the preceding claims, where a limit value, in particular a maximum value, for the gradient of the pressure (PAnode, Pcathode, Pstack) is determined empirically or analytically, e.g. using a model, and / or wherein a limit value, in particular a maximum value, for the pressure gradient (PAnode, Pcathode, Pstack) is determined as a function of at least one operating parameter of the fuel cell stack: - Temperature, - Hydrogen partial pressure and / or - Moisture.

8. Method according to any one of the preceding claims, Where, if the determination of a leakage is dependent on a pressure (PAnode) in an anode region and / or on a pressure (Pcathode) in a cathode region of the fuel cell stack, then the evaluation is carried out before the complete consumption of oxygen in the cathode region of the fuel cell stack. R. 416940 - 20 - 9. Method according to any one of the preceding claims, where, if a recirculation pump is still rotating in an anode path during a passive bleed-down process, an average value between a pressure (PAnodeIn) at an input and a pressure (PAnodeOut) at an output from an anode area of ​​the fuel cell stack is recorded and evaluated.

10. Computer program product comprising instructions which, when the computer program product is executed by a computer, cause the computer to perform the method according to any of the preceding claims.

11. Electronic control unit (ECU) comprising a storage unit in which a code is stored and a computing unit, wherein, when the code is executed by the computing unit, the method according to one of the preceding claims 1 to 9 is carried out.

12. Fuel cell system (FCS) comprising a control unit (ECU) according to the preceding claim.

13. Fuel cell vehicle (FCV) comprising a fuel cell system (FCS) according to the preceding claim.

Images