Method and device for checking the tightness of a fuel cell system

Abstract

The invention relates to a method for checking the tightness of a fuel cell system (2), in particular an anode subsystem (20), which is provided in a fuel cell system (2) for supplying at least one fuel cell (4) of the fuel cell system (2) with gaseous hydrogen and which has an anode shut-off valve (28) and a metering valve (30) in a flow direction of the hydrogen, the method comprising: closing valves (10, 12) of a cathode (6b) of the at least one fuel cell (4); adjusting an overpressure in the anode subsystem (20) by supplying a test gas from a gas reservoir (22); closing the anode shut-off valve (28) and the metering valve (30); measuring the anode pressure (pAn) and / or the cathode pressure (pKat) before and after a predetermined first waiting time (TW1); determining an anode pressure gradient (δpAn) and / or a cathode pressure gradient (δpKat); and outputting an error message if the value of the anode pressure gradient (δpAn) exceeds a predefined anode pressure gradient limit value and / or if the value of the cathode pressure gradient (δpKat) exceeds a predefined cathode pressure gradient limit value.

Description

[0001] R.413843

[0002] - 1 -

[0003] Description

[0004] title

[0005] Method and device for checking the tightness of a fuel cell system

[0006] The invention relates to a method and a device for checking the tightness of a fuel cell system, in particular for checking the tightness of an anode subsystem of a fuel cell system.

[0007] State of the art

[0008] Following a suspected fault and / or a repair to a fuel cell system, especially to an anode subsystem of the fuel cell system, the fuel cell system must be checked for leaks.

[0009] To ensure the safety and proper operation of the fuel cell system, in particular the areas of the fuel cell system, including the anode subsystem, which contain hydrogen during operation, must be reliably checked for leaks to ensure that there are no leaks.

[0010] To check the tightness, the anode subsystem is currently filled with a test gas, such as gaseous nitrogen or a forming gas, and pressurized. Depending on the test gas, the entire anode subsystem is then scanned for leaks using a gas sniffer or a leak detection spray. With this conventional method, it can take a very long time to detect a leak or to fully test the tightness. Suspected areas can be difficult to access, and especially in workshops, there is a risk of overlooking small leaks. R.413843

[0011] - 2 -

[0012] It is therefore an object of the invention to provide a method and a device that make it possible to simplify the detection of leaks in a fuel cell system and to increase the reliability of checking the system for leaks.

[0013] Disclosure of the invention:

[0014] The invention comprises a method for checking the tightness of a fuel cell system, in particular an anode subsystem, which is provided in a fuel cell system for supplying at least one fuel cell of the fuel cell system with gaseous hydrogen, and which has an anode shut-off valve and a metering valve in one flow direction of the hydrogen. A method according to the invention comprises: closing a cathode inlet valve and a cathode outlet valve of a cathode of the at least one fuel cell; setting an overpressure in the anode subsystem by supplying a test gas from a gas reservoir;

[0015] Measuring the anode pressure and / or the cathode pressure of at least one fuel cell; closing the anode shut-off valve and the metering valve; measuring the anode pressure and / or the cathode pressure after an initial waiting period; determining an anode pressure gradient and / or a cathode pressure gradient; and issuing an error message indicating a leak in the fuel cell system if the magnitude of the anode pressure gradient exceeds a predetermined anode pressure gradient limit and / or if the magnitude of the cathode pressure gradient exceeds a predetermined cathode pressure gradient limit.

[0016] The invention also includes a device for checking the tightness of an anode subsystem provided in a fuel cell system for supplying at least one fuel cell with hydrogen, and which has an anode shut-off valve and a metering valve in one direction of hydrogen flow. The device is designed and configured to carry out a method according to the invention for checking the tightness of a fuel cell system, as previously described, in order to check the tightness of the anode subsystem. R.413843

[0017] - 3 -

[0018] A method and a device according to the invention make it possible to detect leaks in a fuel cell system, in particular leaks in an anode submodule of a fuel cell system, conveniently, quickly and with high reliability, without triggering too many false alarms.

[0019] This eliminates the need for lengthy and costly leak detection in the fuel cell system. Workshops can document and prove that they have performed their work on the fuel cell system correctly. Furthermore, automated execution of the procedure can increase reproducibility and reduce the risk of errors in the workshop.

[0020] In one embodiment, the anode pressure gradient limit is in the range between -0.2 mbar / s and -1 mbar / s. The anode pressure gradient limit can, in particular, be -0.5 mbar / s.

[0021] In one embodiment, the cathode pressure gradient limit value lies in the range between 0.07 mbar / s and 0.35 mbar / s. The cathode pressure gradient limit value can, in particular, be 0.17 mbar / s.

[0022] Such limit values ​​for the anode pressure gradient or the cathode pressure gradient make it possible to reliably detect leaks in the fuel cell system without causing a significant number of false alarms.

[0023] In one embodiment, the method comprises weighting, in particular multiplying, the anode pressure difference before and after closing the anode shut-off valve and the metering valve by the ratio between the volumes of the anode subsystem and the cathode subsystem; determining the difference between the cathode pressure difference and the weighted anode pressure difference; and issuing an error message if the amount of the weighted difference exceeds a predetermined limit for the weighted difference.

[0024] In an alternative embodiment, the method comprises the cathode pressure difference with the ratio between the volumes of the R.413843

[0025] - 4 -

[0026] to weight, in particular to multiply, the anode subsystem and the cathode subsystem; to determine the difference between the cathode pressure difference and the weighted anode pressure difference; to determine the difference between the anode pressure difference and the weighted cathode pressure difference; and to output an error message if the magnitude of the weighted difference exceeds a predefined limit for the weighted difference.

[0027] By calculating and evaluating such a weighted difference between the anode pressure difference and the cathode pressure difference, wherein either the anode pressure difference or the cathode pressure difference is weighted by the ratio between the volumes of the anode subsystem and the cathode subsystem, the reliability of a leak detection system according to the invention can be further improved and the number of false alarms can be reduced. Furthermore, by calculating and evaluating the weighted difference between the anode pressure difference and the cathode pressure difference, additional information about the fuel cell system can be obtained, which can facilitate the localization of a leak in the fuel cell system.

[0028] The threshold value for the weighted difference can be in the range between 0.02 mbar / s and 0.08 mbar / s, particularly at 0.05 mbar / s. Such a threshold value for the weighted difference has proven to be well suited for reliably detecting and locating leaks in the fuel cell system.

[0029] In one embodiment, the method comprises measuring the pressure in a region of the fuel cell system between the anode shut-off valve and the metering valve while the anode shut-off valve is open; closing the anode shut-off valve and the metering valve of the fuel cell system; waiting for a predetermined second waiting period; measuring the pressure in the region between the anode shut-off valve and the metering valve again after the second waiting period has elapsed; determining a mean pressure gradient for the region between the anode shut-off valve and the metering valve from the measured pressures; and issuing an error message if the R.413843

[0030] - 5 -

[0031] The magnitude of the mean pressure gradient exceeds a predetermined mean pressure gradient limit.

[0032] These steps can further improve the reliability of a leak detection system according to the invention. In particular, leaks in the area upstream of the metering valve can be detected and located even more reliably.

[0033] The mean pressure gradient limit can be in the range between -0.2 mbar / s and -5 mbar / s, especially at -1 mbar / s.

[0034] In one embodiment, the method comprises closing a tank valve through which a gas reservoir is connected to the fuel cell system, or a gas cylinder serving as a gas reservoir, so that no further gas can flow into the anode subsystem. The method further comprises measuring the pressure downstream of the closed tank valve; opening the anode shut-off valve; waiting for a predetermined third interval; measuring the pressure downstream of the closed tank valve after the predetermined third interval; determining the magnitude of the pressure difference downstream of the tank valve when the anode shut-off valve is open and closed, and comparing it to a predetermined differential limit; and issuing an error message if the magnitude of the pressure difference downstream of the tank valve exceeds the predetermined differential limit.

[0035] In this way, the reliability of a leak detection system according to the invention can be improved even further. In particular, leaks located downstream of the tank valve can be detected and localized even more reliably, and the risk of false alarms can be reduced.

[0036] The difference limit value can be in the range between -5 mbar and -15 mbar, especially at -10 mbar.

[0037] The anode pressure gradient limit, the cathode pressure gradient limit, the weighted difference limit, the mean pressure gradient limit, and / or the difference limit can be determined empirically. R.413843

[0038] - 6 -

[0039] The limit values ​​can be determined, in particular by measurements on existing fuel cell systems or with the help of numerical calculations / computer simulations, in such a way that leaks in the fuel cell system can be reliably detected and located within the fuel cell system without causing too many false alarms.

[0040] An embodiment of the invention is described below with reference to the accompanying figures.

[0041] Brief description of the characters

[0042] Figure 1 shows a schematic view of a fuel cell system whose tightness can be checked using a method according to the invention.

[0043] Figure 2 shows a flowchart of an embodiment of a method according to the invention for checking the tightness of a fuel cell system.

[0044] Character description

[0045] Figure 1 shows a schematic view of a fuel cell system 2, the tightness of which can be checked using a method according to the invention.

[0046] The fuel cell system 2 comprises at least one fuel cell 4, in particular a fuel cell stack 4 containing several fuel cells 4.

[0047] At least one fuel cell 4 has an anode 6a and a cathode 6b, which are separated from each other by a membrane 8.

[0048] Oxygen, for example in the form of oxygen-containing air, is supplied to the cathode 6b of the fuel cell 4 via a cathode inlet valve 10. The cathode 6b also has a cathode outlet valve 12 through which, when open, gas can escape from the cathode 6b of the fuel cell 4. R.413843

[0049] - 7 -

[0050] A cathode pressure sensor 14 is provided at the output of the cathode 6b, which makes it possible to measure the pressure in or at the cathode 6b.

[0051] The fuel cell system 2 also includes an anode subsystem 20, which is designed and configured to supply hydrogen gas to the anode 6a of the at least one fuel cell 4 during operation.

[0052] The anode subsystem 20 comprises a gas reservoir 22, for example a tank or a gas cylinder, which contains hydrogen gas during the regular operation of the fuel cell system 2. For testing the leak tightness of the fuel cell system 2, the gas reservoir 22 can contain a test gas, for example nitrogen or a forming gas.

[0053] Through a tank valve 24, located at the outlet of the gas reservoir 22, and optionally an optional heat exchanger 26, the gas from the gas reservoir 22 is fed to an anode shut-off valve 28. The anode shut-off valve 28 makes it possible to selectively activate / release and deactivate / shut off the supply of gas from the gas reservoir 22.

[0054] Downstream of the anode shut-off valve 28 is a metering valve 30. The metering valve 30 makes it possible to supply the gas from the gas reservoir 22 to the anode 6a of the at least one fuel cell 4 with a predetermined dosage when the anode shut-off valve 28 is open.

[0055] A medium pressure sensor 32 is provided between the anode shut-off valve 28 and the metering valve 30. The medium pressure sensor 32 makes it possible to measure the pressure of the gas between the anode shut-off valve 28 and the metering valve 30.

[0056] Between an output of the metering valve 30 and an input of the anode 6a there is an anode pressure sensor 34, which makes it possible to measure the pressure of the gas that is supplied to the anode 6a of at least one fuel cell 4.

[0057] Anode 6a also has an anode outlet 7 through which gas can escape from anode 6a. R.413843

[0058] - 8 -

[0059] An anode purge valve 16 and an anode drain valve 18 are arranged at the anode outlet 7, which make it possible to purge the anode 6a of the at least one fuel cell 4 by opening the anode purge valve 16 or to drain it by opening the anode drain valve 18.

[0060] A fuel cell system 2 according to an embodiment of the invention thus comprises four areas:

[0061] A first area 41 between the gas reservoir 22 and the anode shut-off valve 28.

[0062] A second area (medium pressure area) 42 between the anode shut-off valve 28 and the metering valve 30.

[0063] A third area (low-pressure area) 43 between the outlet of the metering valve 30 and the anode rinsing and draining valves 16, 18 including the anode 6a of the at least one fuel cell 4; and a fourth area (cathode area) 44 on the side of the cathode 6a between the diaphragm 8 of the at least one fuel cell 4 and the two valves 10, 12 at the cathode 6b.

[0064] The first area 41, the second area 42 and the third area 43 are part of the anode subsystem 20.

[0065] Figure 1 also shows an embodiment of a device 50 according to the invention, which is designed and provided for checking the tightness of the fuel cell system 2.

[0066] Figure 2 shows a flowchart of an embodiment of a method 100 according to the invention for checking the tightness of a fuel cell system 2, in particular for checking the tightness of the anode subsystem 20 of the fuel cell system 2. R.413843

[0067] - 9 -

[0068] To test the tightness of fuel cell system 2, a suitable test gas, such as nitrogen gas, can be used. Alternatively, a readily detectable forming gas can be used as a test gas, making it easy to locate any leaks by detecting the forming gas escaping from fuel cell system 2.

[0069] After starting the procedure 100, in a first step 110 the cathode outlet valve 12 of the cathode 6b of the at least one fuel cell 4 is opened in order to establish a defined pressure level with a cathode reference pressure pKatRef in the cathode region 44 of the fuel cell 4. Since the outlet of the cathode outlet valve 12 is usually connected to the environment, opening the valve 12 at the cathode 6b in the cathode region 44 typically sets the ambient air pressure of approximately 1017 mbar as the cathode reference pressure pKatRef.

[0070] After the cathode reference pressure PKatRef has been set in the cathode 6b, the cathode inlet valve 10 and the cathode outlet valve 12 are closed in step 120. The anode purge valve 16 and the anode drain valve 18 at the anode outlet 7 of the at least one fuel cell 4 are also closed in step 120.

[0071] After closing the valves 10, 12, 16, 18 on the at least one fuel cell 4, in step 130 an overpressure of, for example, approx. 500 mbar relative to the environment is created at and in the anode 6a in the low-pressure area 43 by appropriately controlling the anode shut-off valve 28 and the metering valve 30, i.e. an absolute anode pressure pAn of approx.

[0072] Set to 1500 mbar.

[0073] After a successful pressure build-up in at least one fuel cell 4, the cathode pressure sensor 14 indicates a cathode pressure pKat of approximately 1017 mbar for the cathode area 44, and the anode pressure sensor 34 at the anode 6a indicates an anode pressure PAn of approximately 1500 mbar in the low-pressure area 43 at the anode 6a. R.413843

[0074] - 10 - In the first area 41 and in the second area 42 between the gas reservoir 22 and the metering valve 30, there is a so-called intermediate pressure PM of, for example, approximately 15 bar. The intermediate pressure PM is displayed by the intermediate pressure sensor 32.

[0075] In the next step 140, the anode shut-off valve 28 and the metering valve 30 are also closed, so that no pressure adjustment can take place in the anode 6a of at least one fuel cell 4.

[0076] After the anode shut-off valve 28 and the metering valve 30 have been closed, a predetermined first waiting time Twi was observed in step 150.

[0077] The initial waiting time Twi can be between 60 s and 300 s, for example. Specifically, the initial waiting time Twi can be 180 s.

[0078] After the first waiting period Twi has elapsed, various pressure differences or pressure gradients can be calculated, which make it possible to conclude about any leaks in the fuel cell system 2:

[0079] For the cathode region 44 at cathode 6b, the cathode pressure pKat is measured in step 160, and a cathode pressure difference ApKat or a cathode pressure gradient öpKat = ApKat / Twi is determined. Due to partial pressure equalization between the anode 6a and the cathode 6b of the at least one fuel cell 4 through the membrane 8, a pressure increase, i.e., a positive cathode pressure gradient öpKat, is expected for the cathode region 44.

[0080] The cathode pressure pKat after the first waiting time Twi can, for example, be 1077 mbar, resulting in a cathode pressure difference ApKat of 60 mbar for an initial cathode pressure of 1017 mbar. With an initial waiting time Twi of 600 s, this results in a cathode pressure gradient öpKat of öpKat = (1077 mbar - 1017 mbar) / 600 s = 60 mbar / 600 s öpKat = 0.1 mbar / s. R.413843

[0081] - 11 -

[0082] In a subsequent step 170, the anode pressure pAn in the low-pressure region 43 is measured at the anode 6a, and an anode pressure difference ApAn or an anode pressure gradient öpAn on the side of the anode 6a of the at least one fuel cell 4 is determined. In the low-pressure region 4, a pressure drop, i.e., a negative anode pressure gradient öpAn, is expected due to the partial pressure equalization between the anode 6a and the cathode 6b through the membrane 8 of the at least one fuel cell 4.

[0083] For example, the anode pressure pAn after the first waiting time Twi can be 1320 mbar, resulting in an anode pressure difference ApAn of 180 mbar for an initial anode pressure of 1500 mbar. With an initial waiting time Twi of 600 s, this results in an anode pressure gradient öpAn of öpAn = (1500 mbar - 1320 mbar) / 600 s = 180 mbar / 600 s öpAn = -0.3 mbar / s.

[0084] Optionally, in a next step, a total deviation Ap of 180 can be calculated. geS The pressure p in the at least one fuel cell 4 is calculated. In particular, different volumes VAP and VKat of the low-pressure area 43 at the anode 6a and of the cathode area 44 at the cathode 6b are taken into account.

[0085] For example, if the volume VKat of the cathode region 44 is three times larger than the volume VAP of the low-pressure region 43 at the anode 6a (VKat = 3 * VAH), the cathode pressure increase ApKat is multiplied by a volume factor Kvoi of three. Then the difference between the pressure increase ApKat at the cathode 6b weighted by the volume factor and the pressure drop ApAn in the low-pressure region 43 at the anode 6a is calculated:

[0086] Total = | (Kvol ApKat) + ApKat| -

[0087] The difference is calculated by addition, since the pressure drop ApAn in the low-pressure area 43 at the anode 6a is negative according to the definition chosen here.

[0088] For the example mentioned above, this results in: R.413843

[0089] - 12 -

[0090] Aptotal = (3 * 60 mbar) + (-180 mbar) = 180 mbar - 180 mbar Ap total = 0 mbar.

[0091] This means that the pressure drop ApKat in the low-pressure area 43 at the anode 6a is completely compensated by a corresponding pressure increase ApKat in the cathode area 44. From this, it can be concluded that no gas has escaped from the fuel cell system 2, and therefore the fuel cell system 2 is sealed.

[0092] For the anode pressure drop ApAnt or the anode pressure gradient öpAn on the side of anode 6a, the cathode pressure increase ApKat or the cathode pressure gradient öpKat on the side of cathode 6b and / or for the total deviation Ap geS For each pressure p, separate limit values ​​can be set.

[0093] By comparing the anode pressure drop ApAn or the anode pressure gradient öpAn on the side of anode 6a, the cathode pressure rise ApKat or the cathode pressure gradient öp at on the side of cathode 6b and / or the total deviation Ap geSIn step 190, a leak in the fuel cell system 2 can be inferred from the pressure p with the corresponding limit values ​​if these limit values ​​are exceeded or not reached.

[0094] In a subsequent step 200, the so-called mean pressure PM is measured with the mean pressure sensor 32 in the area between the closed anode shut-off valve 28 and the closed metering valve 30 before and after a second waiting time Tw2.

[0095] The second waiting time Tw2 can be the same as the first waiting time Twi or different from it.

[0096] The second waiting time Tw2 can be, for example, between 60 s and 300 s, in particular 180 s. R.413843

[0097] - 13 - In step 210, a mean pressure-pressure difference APM = pwfc) - Pwi(ti) or a mean pressure-pressure gradient is calculated from the mean pressures pw(ti), pwife) measured before and after the second waiting time Tw2 (t2 = ti + Tw2).

[0098] ÖPM = ApM / TW25 of the mean pressure PM calculated and evaluated, e.g.:

[0099] ÖPM = (14.4 bar - 15 bar) / 600 s = - 1 mbar / s.

[0100] The calculated mean pressure difference Apw and / or the calculated mean pressure gradient ÖPM can also be compared with predefined limit values. A leak in fuel cell system 2, in particular a leak in the second area 42 of fuel cell system 2, can be inferred if the magnitude of the mean pressure difference Apw and / or the magnitude of the mean pressure gradient öp m exceeds a predetermined limit.

[0101] In a subsequent step 220, the supply of test gas to the fuel cell system 2 from the gas reservoir 22 is deactivated, e.g. by closing the tank valve 24, so that no further test gas can flow from the gas reservoir 22 into the fuel cell system 2.

[0102] In the next step 230, the pressure pa downstream of the anode shut-off valve 28 is first measured with the medium pressure sensor 32 with the anode shut-off valve 28 closed and the metering valve 30 closed.

[0103] Then the anode shut-off valve 28 is opened and after a third waiting period Twa, the pressure pa is measured again with the medium pressure sensor 32.

[0104] The third waiting time Twa can be the same as the first waiting time Twi and / or the same as the second waiting time Tw2, or it can differ from them.

[0105] The third waiting time Tws can, for example, be between 5 s and 300 s, in particular 15 s. R.413843

[0106] - 14 -

[0107] Here too, the pressure difference Aps = pfc) - p(ti) (t2 = h + Twa) or the pressure gradient epa = Aps IT can be used. W3 ; calculated and compared with predefined limits.

[0108] For example, before opening the anode shut-off valve 28, a pressure ps(ti) of 14.95 bar can be measured, and after opening the anode shut-off valve 28 and after the third waiting time Twa has elapsed, a pressure psfc) of 14.9 bar can be measured. With a third waiting time Twa of 10 s, this results in a pressure gradient öps of -5 mbar / s.

[0109] Spa = (14.90 bar - 14.95 bar) / 10 s = - 5 mbar / s.

[0110] The specified limit value for the pressure gradient εps can, for example, be -10 mbar / s.

[0111] Using the aforementioned criteria, and in particular using a combination of the aforementioned criteria, leaks in the fuel cell system 2 can be reliably detected and located or ruled out.

[0112] In step 240, one or more messages can be issued regarding the presence of a leak in fuel cell system 2, the suspected area of ​​fuel cell system 2 where the leak is located, and a possible cause of the fault. The messages can be displayed, for example, on a display device 52 of the device 50 for checking the tightness of fuel cell system 2.

[0113] For example, if the cathode pressure pKat on the side of cathode 6b does not increase or only increases slightly, and at the same time the calculated total deviation Ap geS If the value is greater than the corresponding limit value, this indicates that at least one of the cathode valves 10, 12 is leaking.

[0114] A rapid and / or significant pressure drop ApAn in the low-pressure area 43 on the side of the anode 6a, which is associated with an excessive increase ApKat of the R.413843

[0115] - 15 -

[0116] cathode pressure pKa and with a total deviation Ap geS If the device is connected in the expected range, this may indicate a defective membrane 8 in the fuel cell 4.

[0117] A successfully completed test run, in which no indication of a leak was found, can also be displayed on the display device 52.

[0118] A successfully completed test run can also be stored in a storage device 54, for example, for logging purposes. The storage device 54 can be located inside or outside the device 50 for checking the tightness of the fuel cell system 2. The storage device 54 can, for example, be located in a virtual cloud.

Claims

R.413843 - 16 - Patent claims 1. Method for checking the tightness of a fuel cell system (2), in particular an anode subsystem (20) which is provided in a fuel cell system (2) for supplying at least one fuel cell (4) of the fuel cell system (2) with gaseous hydrogen, and which has an anode shut-off valve (28) and a metering valve (30) in a flow direction of the hydrogen, wherein the method comprises: Closing a cathode inlet valve (10) and a cathode outlet valve (12) of a cathode (6b) of the at least one fuel cell (4); Setting an overpressure in the anode subsystem (20) by supplying a test gas from a gas reservoir (22); Measuring an anode pressure (pAn) and / or a cathode pressure (pKat) at the at least one fuel cell (4); Closing the anode shut-off valve (28) and the metering valve (30); Measuring the anode pressure (pAn) and / or the cathode pressure (pKat) after a predetermined initial waiting time (Twi); Determining an anode pressure gradient (epAn) and / or a cathode pressure gradient (epCat); and Output an error message if the magnitude of the anode pressure gradient (öpAn) exceeds a predefined anode pressure gradient limit and / or if the magnitude of the cathode pressure gradient (Sp at) exceeds a predefined cathode pressure gradient limit. R.413843 - 17 - 2. Method according to claim 1, wherein the anode pressure gradient limit is in the range between -0.2 mbar / s and -1 mbar / s, in particular at -0.5 mbar / s.

3. Method according to claim 1 or 2, wherein the cathode pressure gradient limit value is in the range between 0.07 mbar / s and 0.35 mbar / s, in particular at 0.17 mbar / s.

4. A method according to any of the preceding claims, wherein the method comprises weighting an anode pressure difference (ApAn) by the ratio (Kvoi) between the volumes of the anode subsystem (20) and the cathode subsystem (44); determining the difference between the cathode pressure difference (ApKat) and the weighted anode pressure difference (Kvoi * ApAn); and issuing an error message if the magnitude of the difference exceeds a predetermined difference limit; or weighting the cathode pressure difference (ApKat) by the ratio (Kvoi) between the volumes of the anode subsystem (20) and the cathode subsystem; determining the weighted difference between the anode pressure difference (ApAn) and the weighted cathode pressure difference (Kvoi * ApKat); and issuing an error message if the magnitude of the weighted difference exceeds a predetermined limit for the weighted difference.

5. Method according to claim 4, wherein the limit value for the weighted difference is in the range between 0.02 mbar / s and 0.08 mbar / s, in particular at 0.05 mbar / s.

6. Method according to one of the preceding claims, wherein the method comprises measuring the pressure (PM) in a region (42) between the anode shut-off valve (28) and the metering valve (30); R.413843 - 18 - to close the anode shut-off valve (28) and the metering valve (30); to wait for a predetermined second waiting time (Tw2); to measure the pressure (PM) again in the area (42) between the anode shut-off valve (28) and the metering valve (30) after the predetermined second waiting time (Tw2) has elapsed; to determine a mean pressure gradient (PM) for the area (42) between the anode shut-off valve (28) and the metering valve (30) from the measured pressures (PM); and to issue an error message if the magnitude of the mean pressure gradient (PM) exceeds a predetermined mean pressure gradient limit.

7. Method according to claim 6, wherein the mean pressure gradient limit is in the range between -0.2 mbar / s and -5 mbar / s, in particular at -1 mbar / s.

8. A method according to any of the preceding claims, wherein the method comprises closing a tank valve (24) so ​​that no gas can flow into the anode subsystem (20); measuring the pressure (pa) in the anode subsystem (20) downstream of the closed tank valve (24); opening the anode shut-off valve (28); and waiting for a predetermined third waiting period (Twa). R.413843 - 19 - to measure the pressure (ps) in the anode subsystem (20) downstream of the closed tank valve (24) after the specified third waiting time (Twa) has elapsed; to determine the magnitude of the pressure difference (ps) downstream of the tank valve (24) between the open and closed anode shut-off valve (28) and to compare it with a specified differential limit; and to issue an error message if the magnitude of the pressure difference downstream of the tank valve (24) exceeds the specified differential limit.

9. Method according to claim 8, wherein the difference limit is in the range between -5 mbar and -15 mbar, in particular at -10 mbar.

10. Method according to one of the preceding claims, wherein at least one of the anode pressure gradient limit, the cathode pressure gradient limit, the difference limit, the mean pressure gradient limit and the difference limit has been empirically determined.

11. Device (50) for checking the tightness of a fuel cell system (2), in particular an anode subsystem (20) which is provided in a fuel cell system (2) for supplying at least one fuel cell (4) with hydrogen, and which has an anode shut-off valve (28) and a metering valve (30) in a flow direction of the hydrogen, wherein the device (50) is provided and configured to carry out a method according to one of the preceding claims in order to check the tightness of the anode subsystem (20).

Images