Method for operating a fuel cell system, control unit

Abstract

The invention relates to a method for operating a fuel cell system (1), comprising a fuel cell stack (200) having a plurality of fuel cells (2a, 2b) and having cooling channels (201) running through it, which are supplied with a coolant via a cooling circuit (25) using a coolant pump (26), wherein at the beginning of a freeze start (40) according to the principle of air depletion, in an initial step (41), a constant current setpoint (Itarget) is first set in order to achieve a desired target voltage (Utarget) by voltage regulation via the air mass flow (Mfair), wherein when the current actual voltage (Uist) falls below at least a predetermined minimum voltage (Umin) (42), the initial voltage regulation via the air mass flow (Mfair) is interrupted and, instead, a constant air mass flow (Mfair const) is set until the actual voltage (Uist) rises above the minimum voltage (Umin) by current regulation.

Description

The present invention relates to a method for operating a fuel cell system and to a control device which is configured to carry out steps of the method.Prior ArtA fuel cell usually converts a fuel, for example, hydrogen, and oxygen into electrical energy, heat, and water. To increase the power, a multiplicity of fuel cells are generally connected to form a fuel cell stack and supplied with the reaction gases via supply channels passing through the fuel cell stack. The heat generated in the fuel cells during the electrochemical process is dissipated by means of a cooling circuit and is dissipated to the environment via a cooler-generally the vehicle cooler in mobile applications. The coolant of the cooling circuit is pumped through the fuel cell stack by means of a coolant supply channels integrated into the cooling circuit. To bypass the cooler, a directional valve is usually integrated into the cooling circuit. Bypassing the cooler can be advantageous, for example, in the case of start-up, since, in particular at ambient temperatures below 0° C., the fuel cell stack should be heated as quickly as possible in order to avoid condensation water and / or ice accumulations, which could delay or even prevent the start-up. However, the risk of icing is not prevented until the coolant has been reliably warmed above 0° C. before entering the fuel cell stack. This does not create freezing conditions when the coolant is pumped into the fuel cell stack.In the case of freeze starts, the coolant according to the general prior art is either heated externally from the stack or by the electrochemical reaction in the fuel cell stack. In both cases, however, the starting process is extended as a result. Furthermore, the ice tolerance of the fuel cells must be increased due to the constant cooling below 0° C. This is usually done by incorporating ice buffers in the fuel cells, additional heaters in the fuel cell system, and the like.During the start of freezing with heat production directly in the fuel cell stack, the coolant volume flow must be high enough to prevent what are known as "hotspots" and an excessively large temperature difference between cell inlet and outlet, but low enough to not ice the inlet region of the fuel cells as a result of the temperature reduction caused by the cold coolant flowing in. At very low temperatures from about -20° C., the water formed on the cathode side cools approximately instantaneously and thus blocks the active centers. As a result of the continuous decrease of the active area, the voltage decreases ever further down to values close to 0 V. In this case, the reduction and recombination of H+ to gaseous H2 takes place more strongly on the cathode side, so that the H2 concentration in the cathode off-gas increases significantly. As soon as preheated coolant flows into the fuel cell stack, the ice dews and the liquid water formed can be transported away.As a countermeasure, it is also generally known at the beginning of a freezing start according to the so-called principle of air depletion to initially set a constant current setpoint value in order to achieve the desired target voltage by means of a voltage regulation of the air mass flow. However, this countermeasure is often not sufficient.It is the object of the present invention to identify an icing on the cathode side in fuel cells as early as possible in order to initiate effective countermeasures based on an initial current limitation using simple technical means.Disclosure of the InventionThe object is achieved on the basis of a method according to the preamble of claim 1 in conjunction with its characterizing features. The dependent claims which follow represent advantageous further developments of the invention. Claim 10 is directed to a control device implementing the method.The invention includes the technical teaching that at the beginning of a freezing start according to the principle of air depletion, a constant current setpoint is initially set in order to achieve a desired target voltage via a voltage control of the air mass flow, wherein in the event of the current actual voltage dropping below at least a predefined minimum voltage, the initial voltage control of the air mass flow is interrupted and instead a constant air mass flow is set until the actual voltage rises again above the minimum voltage via current control.In other words, according to the invention, after detection of icing of the cathode as countermeasures, an initial limiting of the current takes place first, wherein at the same time or independently thereof an adaptation of the coolant flow is carried out in order not to cool the cell inlet too much. This achieves a stabilization of the voltage above the level from which H2 formation begins on the cathode side, i.e. from which the voltage converter can no longer operate. This allows a rapid and uninterrupted freezing start by means of specific regulatory measures and increases the service life of the fuel cell stack.In other words, the method according to the invention includes the fact that a constant current setpoint value is initially set at the beginning of a freezing start. At the same time, attempts are made to reach the target voltage by regulating the coolant flow. Up to this point, the countermeasures initiated correspond to a start of freezing with depletion of air, the so-called "air stagnation", in order to generate as much heat as possible with the fuel cell stack. This initial countermeasure fails, however, when fuel cells increasingly ice on the cathode side. In this case, voltage regulation via the air mass, i.e. the coolant flow, is no longer possible. Therefore, the method according to the invention provides that compliance with a voltage corridor or at least a predefined minimum voltage is monitored. The minimum voltage may be the total voltage of the fuel cell stack or an average single voltage of fuel cells. As soon as one of these variables or both variables fall below the minimum value or the target corridor, the regulation of the coolant flow is stopped and a constant value is instead set until the actual voltage rises above the minimum voltage by current regulation.Preferably, the air of the coolant stream is supplied here slightly superstoichiometrically with a lambda of between 1.1 and 1.3, preferably 1.2. At the same time, the current is regulated with the aim of stabilizing the voltage back to the original target voltage. This state is maintained until the originally set current setpoint value is reached again. At the connection, the air system again assumes the regulation during the voltage regulation and the current is kept constant. A constant current setpoint value is thus set in order to reach the target voltage again via a voltage regulation via the air mass flow.Alternatively, in the above-described procedure, the H2concentration in the exhaust gas can also be used as a trigger for the current limiting. However, in this case the delay is substantially greater, so that a late response to an icing of the fuel cells can be made. In this case, the icing can only be limited. Instead of limiting the flow, the coolant flow can also be adapted. In this case, the coolant flow would be reduced when icing begins, so that less cold coolant is forced into the fuel cell stack. However, this results in the freezing start being prolonged because it takes longer for the entire amount of coolant to be warmed up to above the freezing point and the fuel cells can be thawed. In addition, depending on the fuel cell stack, the formation of what are known as hotspots in the fuel cells can occur at a low coolant flow and a high current intensity, which can ultimately irreversibly damage the fuel cell stack.According to a measure which further improves the invention, it is proposed that more than one voltage tap is provided for each fuel cell for determining the current actual voltage, which is preferably carried out by means of a high-precision 1-channel voltage module (CVM). The voltage taps should preferably take place in the region between the air inlet of a fuel cell and the air outlet. This is based on the idea that during operation with depletion of air a locally limited reaction zone is formed within which the complete oxygen is consumed. Thus, the voltage distribution in the cell is inhomogeneous. Therefore, it is also not to be assumed that the same voltage is measured at both taps. If the reaction zone freezes during the start of freezing, it is displaced continuously in the direction of the air outlet, that is to say in the direction of the second voltage tap. In this case, the two voltage values change as the reaction zone migrates, i.e. the icing progresses. By comparing the fuel cells, it is therefore possible to react to an incipient icing of the cells already at an early stage.As proposed above, the current is reduced or the air mass flow is throttled as a function of the stack voltage in order to prevent further freezing of the fuel cells on the cathode side. A minimum voltage has been proposed as a criterion here, which can be calculated both from the total voltage of the fuel cell stack and from an average individual voltage of fuel cells. However, a blockage of individual fuel cells on the anode side can no longer be detected in this case. In this case, individual fuel cells assume strongly negative values, while the average voltage hardly changes. This highly damaging condition requires a substantially more aggressive reduction of the current, in contrast to the cathode icing.In order to prevent anode icing as well as cathode icing, according to a first variant, it is proposed as a countermeasure to offload the monitoring and regulation of the minimum voltage to the voltage converter of the fuel cell. This is advantageous because voltage converters have, in principle, a minimum voltage limit below which operation is not possible. Therefore, in order not to jeopardize stable operation, the voltage converter will reduce the current to stabilize the voltage. At the same time, the control unit of the fuel cell system is signalled that the voltage regulation via the air mass flow is stopped. Subsequently, the control device takes over the current setpoint value determined by the voltage converter and then continuously updates the coolant current, so that the cathode stoichiometry is kept constant with respect to the current setpoint value set in each case.This offers the advantage that individual fuel cells are not operated increasingly superstoichiometrically, as a result of which the heat production is greatly reduced. In the further course, the control system on the voltage converter of the fuel cell attempts to successively restore the current to the original current setpoint value as target value without dropping below the predefined minimum voltage.Preferably, the control device for regulating the minimum voltage is activated only when a fuel cell is running into the defined control range. In principle, the total voltage of the fuel cell stack is therefore limited downward, i.e. to a minimum voltage, by limiting the current in the voltage converter of the fuel cell and by matching with the control unit of the fuel cell system in the sense of stopping the lambda controller, as long as the current limitation is active. As a result, strongly negative cell voltages (anode blocking) are intercepted on the control device by means of aggressive current setpoint limiting.According to an alternative approach, implementation and parallel execution of both modes can therefore be carried out on the one hand by current limiting for the total voltage and aggressively attenuating negative cell voltages on the control unit of the fuel cell system. For this purpose, both controllers, on the one hand for controlling the minimum total voltage of the fuel cell stack and, on the other hand, for controlling individual voltages of fuel cells in the voltage module (CVM), are carried out on the control unit of the fuel cell system, whereas the voltage converter of the fuel cells passively assumes the setpoint value calculated there.Exemplary EmbodimentFurther measures improving the invention are described in more detail below together with the description of preferred embodiments of the invention on the basis of the figures. It shows: FIG. 1 shows a schematic illustration of a fuel cell system which is suitable for carrying out a method according to the invention, FIG. 2 ashows a flow chart of the method according to the invention according to a first embodiment, FIG. 2 bshows a flow chart of the method according to the invention according to a second embodiment, FIG. 3 shows a schematic illustration of a fuel cell stack of a fuel cell system, FIG. 4 is a diagram for illustrating a freezing start with a time profile of various parameters of the fuel cell system; and FIG. 5 is a graph comparatively showing cell voltages in the vicinity of end plates of a fuel cell stack.According to FIG. 1, a fuel cell system 1 for a motor vehicle comprises a plurality of fuel cells 2 (by way of example) each having an anode 3 and a cathode 4.Hydrogen is supplied to the anode 3 via an anode circuit 5. The hydrogen is stored in a tank 15 which is connected to the anode circuit 5 via a valve 13, a heat exchanger 15, a pressure regulator 16 and an ejector 17. With the aid of the suction jet pump 17, discharged hydrogen emerging from the fuel cell 2 can be passively recirculated. The suction jet pump 17 is actively supported by a blower 18. Since the anode gas accumulates with nitrogen over time, purging is carried out from time to time. For this purpose, a purge valve 8 is opened, via which the anode gas is removed from the anode circuit 5. The water obtained in the process is separated by means of a water separator 19 and collected in a container 20. By opening a drain valve 9, the container 20 can be emptied.The cathode 4 will be supplied with air via a cathode supply air path 6, which is taken from the environment. The air passes via an air filter 21 to an air compressor 7, with which the air is compressed. The air heats up and is cooled downstream of the air compressor 7 via a heat exchanger 22 in the cathode supply air path 6. Depleted air exiting the fuel cell 2 is discharged to the outside via a cathode exhaust air path 10. The air supply can be blocked by shut-off valves 11 and 12. Furthermore, a bypass path 23 with an integrated bypass valve 24 is provided to the environment of the at least one fuel cell 2.The fuel cell 2 of the fuel cell stack-not shown in detail-is furthermore connected to a cooling circuit 25 in order to remove the heat generated in the process. The cooling circuit 25 is operated via a coolant pump 26 and includes a fuel cell side heat exchanger 27 and a radiator 28, and a cooling circuit can be switched between a small cooling circuit for starting operation and a large cooling circuit including the radiator 28 via a drain valve 29.The fuel cell system 1 further includes, as electrical components, an inverter 30 for converting the direct current generated by the fuel cells 2 into a 3-phase alternating current for driving an alternating current motor 31 of the motor vehicle illustrated here. A direct voltage converter 32 connected upstream converts the direct voltage generated by the fuel cells 2 into a higher direct voltage suitable for the inverter 30. In this exemplary embodiment, at least both components are connected to a superordinate electronic control unit 33 of the fuel cell system 1 or are a component thereof.According to FIG. 2 a, according to a first embodiment of the method according to the invention for operating the fuel cell system described above, at the beginning of a freezing start 40, a constant current setpoint value I target is initially set in step 41 according to the principle of air depletion in order to achieve a desired target voltage U target by voltage regulation via the air mass flow Mf air.If, however, in a comparison step 42 the current actual voltage U ist falls below a predefined minimum voltage U min, then in a subsequent step 43 the regulation of the air mass flow Mf air is interrupted and instead a constant air mass flow Mf air const is set until the actual voltage U ist rises above the set minimum voltage U min by current regulation, which is determined via a comparison step 44. If the condition is fulfilled, the initial voltage control via the air mass flow Mf air is entered again.The variant shown in FIG. 2 b differs in modification from the above variant in that-as depicted on the left-hand side-the monitoring and regulation of the minimum voltage U min on the voltage converter 32 which has at least one fuel cell 2 removed from it and is signaled to the superordinate control unit 33 that the voltage regulation via the air mass flow Mf air has been stopped, after which the voltage converter 32 takes over the current setpoint value I target determined by the control unit 33 and then continuously updates the air mass flow Mf air so that the cathode stoichiometry is kept constant with respect to the respectively set current setpoint value I target. For this purpose, a stepwise increase is achieved in step 44' to I target.The control device 33 becomes active for regulating the minimum voltage U min only when the at least one fuel cell 2 is running into the defined control range. In this case, a current setpoint value is determined on the control unit 33 and transmitted to the voltage converter 32. Both values are continuously compared on the voltage converter 32. In the event that the cell voltage regulator does not respond to the control device, the voltage converter 32 receives its own current setpoint value back. As soon as the cell voltage regulator becomes active, the control device sends a lower setpoint value back to the voltage converter 32. Therefore, the minimum of the own set value and the set value fed back from the control device 33 is always set at the voltage converter 32. When leaving the cell voltage control range, it is advantageous if the current is gently ramp-adjusted again to the original setpoint value according to step 44. Otherwise, there is the risk that critical cells are immediately drawn back into the anode-side depletion. The transition out of the cell voltage control range is thereby formed on the control device 33. As soon as the originally predefined current setpoint value I target is reached again, the voltage converter 32 again passively accepts the current setpoint value from the control unit 33, which in turn attempts to regulate to the desired target voltage I target using the air mass flow Mf air again.In addition to the comparison step 42 for checking whether a predefined minimum voltage I min of the overall system is undershot, a comparison is made in a parallel comparison step 52 as to whether this also applies with respect to a predefined minimum cell voltage. If affirmative, current control is performed in step 53 according to the minimum cell voltage U cell_min. According to the boundary condition 54 applicable to both parallel branches, the air mass flow Mf air is kept constant.According to FIG. 3, an exemplary fuel cell stack 200 consists of a plurality of fuel cells 2 a, 2 b, etc. This is traversed by cooling channels 201 of the heat exchanger 27 of the cooling circuit 25, which heat exchanger is not shown in greater detail. The cathode 4 is supplied with air 202. As illustrated, more than one voltage tap 203 for a voltage module (CVM) is provided per fuel cell 2 aand 2 b, which voltage tap is located close to the air inlet into the fuel cells 2 a, 2 b. The other voltage tap 204 is arranged on the air outlet side. An ice V which is imminent due to cold frozen start starts out on the cathode 4 side.The diagram shown in FIG. 4 illustrates the profile of the parameters air mass flow Mf air, total voltage U ist_stack, the hydrogen concentration xH2, the coolant inlet temperature T in and the coolant outlet temperature T out and a set of cell voltages U CVM of the individual fuel cells of the fuel cell stack during a freezing start measurement starting from -20° C.At the beginning, a constant current setpoint value of 90 amperes is set according to the principle of air depletion. As a result, the air mass flow Mf air is regulated in such a way that a target voltage U target of 60 V is reached. The characteristic of the fuel cell stack is intentionally deteriorated to increase the heat production. It is apparent from the graph that although the air mass flow Mf air remains approximately the same, the total voltage of the fuel cell stack U ist-stack falls continuously up to a minimum of approximately 0 V. In this case, it is thus assumed that the ice growth proceeds on the cathode side, whereby the active area of the fuel cells is reduced ever further. This means that, with the current remaining the same, the current density per active area continues to increase, which is why the voltage falls. At the same time, it can be seen that the H2 concentration xH2 directly at the cathode outlet rises sharply to 1.8%, which indicates the effect described above. As soon as preheated coolant reaches the fuel cells, which takes place at the point X, it can be seen that the cell voltages gradually recover again.According to FIG. 5, it can be seen from the comparison of the cell voltages of fuel cells at the left edge "L" near the media distributor plate with those at the right edge "R" of the fuel cell stack that the fuel cells at the left wheel initially get oxygen at the start, thus water production starts there first. However, these fuel cells also reach the preheated coolant first, so that these fuel cells thaw again first. The right edge is offset in time with respect to this. These fuel cells start later, since the air front must first move through the stack at the start. The preheated coolant also reaches these fuel cells later, so that the cells remain frozen for a longer time.

Claims

Method for operating a fuel cell system (1), comprising a fuel cell stack (200) which has a multiplicity of fuel cells (2a, 2b) and through which cooling ducts (201) are passed, which ducts are supplied with a coolant via a cooling circuit (25) using a coolant pump (26), wherein, at the beginning of a freezing start (40), a constant current setpoint value (I target) is initially set in an initial step (41) according to the principle of air depletion, in order to achieve a desired target voltage (U target) by voltage control via the air mass flow (Mf air) characterized in that, that, when the current actual voltage (U ist) falls below at least a predetermined minimum voltage (U min) the initial voltage control over the air mass flow (Mf air) is interrupted and instead a constant air mass flow (Mf air const) is set until the actual voltage (U ist) rises above the minimum voltage (U min) by current control.Method according to Claim 1, characterized in that the air of the air mass flow (Mf air) is supplied in superstoichiometric form with a lambda of between 1.1 and 1.3, preferably 1.2, the current being regulated in such a way that the voltage stabilizes again at target voltage (U target).Method according to Claim 1, characterized in that the actual voltage (U ist) corresponds to the total voltage (U iststack) of the fuel cell stack (2) and / or to an average single-cell voltage (U ist mittel) of fuel cells ().Method according to Claim 1, characterized in that a constant current setpoint value (I target) is subsequently set again in order to reach the target voltage (U target) again by means of voltage regulation via the air mass flow (Mf air).Method according to Claim 1, characterized in that more than one voltage tap is provided for each fuel cell (2a, 2b) in order to determine the current actual voltage (U ist).Method according to Claim 5, characterized in that the voltage tap is carried out between the air inlet () and the air outlet () of the fuel cell (2a; 2b).Method according to Claim 1, characterized in that the monitoring and regulation of the minimum voltage (U min) is transferred to a voltage converter (32) of the at least one fuel cell (2), and the superordinate control unit (33) is signalled that the voltage regulation via the air mass flow (Mf air) has been stopped, after which the voltage converter (32) takes over the current setpoint value (I target) determined by the control unit (33) and then continuously updates the air mass flow (Mf air) so that the cathode stoichiometry is kept constant with respect to the respectively set current setpoint value (I target).Method according to Claim 7, characterized in that the control device (33) for regulating the minimum voltage (U min) only becomes active if the at least one fuel cell (2) runs into the defined control range.Method according to Claim 1, characterized in that both regulations of the minimum total voltage (U ist stack min) of the fuel cell stack () and of the minimum voltage (U min) are carried out on the control unit (33), and the setpoint value calculated by the control unit (33) is passively adopted by the voltage converter (32).Control device (33) for a fuel cell system (1) which is configured to carry out steps of a method according to one of the preceding claims.

Images