Method for operating a fuel cell system, control unit

Abstract

The invention relates to a method for operating a fuel cell system (1) comprising an anode subsystem (3) for supplying at least one stack (2) with hydrogen to generate a stack current, wherein the hydrogen is supplied in an anode gas which becomes enriched with nitrogen during operation, and wherein a nitrogen concentration in the anode gas is controlled by temporarily discharging anode gas from the anode subsystem (3) via a controllable purge valve (14) during a purge process and replacing it by supplying fresh hydrogen, characterized by the following steps: (a) Determination of a hydrogen concentration in the anode gas, (b) Inference from the hydrogen concentration and the stack current to an actual anode stoichiometry, (c) comparing the actual anode stoichiometry with a target anode stoichiometry, (d) Indirect adjustment of the hydrogen concentration by varying the duration and / or frequency of the purge process if the actual anode stoichiometry deviates from the target anode stoichiometry. The invention further relates to a control device for executing method steps.

Description

The present invention relates to a method for operating a fuel cell system according to the preamble of claim 1.Prior ArtA fuel cell is used for the electrochemical conversion of hydrogen and oxygen into electrical energy, heat and water. The core of a fuel cell is a membrane electrode assembly. This comprises a membrane which separates an anode side and a cathode side from one another. The entire fuel cell system is thus divided into an anode subsystem and a cathode subsystem. In operation, oxygen is supplied to the fuel cell in the form of air via the cathode subsystem and hydrogen is supplied via the anode subsystem. During operation, the hydrogen and the oxygen react with one another to form water, with energy being released. To scale the energy obtained, individual fuel cells are generally combined into so-called stacks by stacking.Since anode gas emerging from a stack generally still contains unused hydrogen, it is recirculated within the anode subsystem and fed back to the anode side of the stack. Over time, however, the recirculated anode gas accumulates with nitrogen and water, wherein the water can be present in liquid or gaseous form. Liquid water is generally removed with the aid of a water separator. The water separator usually comprises a container in which the separated water is collected in liquid form. By opening a valve, the so-called drain valve, the container can be emptied. The water flowing out via the drain valve is combined with the exhaust air from the cathode subsystem and discharged from the fuel cell system.The nitrogen contained in the anode gas is present in gaseous form and cannot be separated meaningfully from the hydrogen. In order to remove nitrogen from the anode subsystem, anode gas is therefore discharged within or downstream of the water separator as part of a purge process via a so-called purge valve. The anode gas thus separated, which contains both nitrogen and hydrogen, is combined with the exhaust air from the cathode subsystem and any separated water and separated out of the fuel cell system. The purge process is variable in its duration and frequency and is directly controllable. The disadvantage of the purge process is that hydrogen is lost in this process, which could actually still be used as fuel.The anode gas deposited during the purge operation is replaced within the anode subsystem by supplying fresh hydrogen. However, this fresh hydrogen stored in tanks is generally contaminated with nitrogen. The hydrogen quality for vehicles with fuel cells is specified in SAE J2719 as well as in ISO 14687:2019. A maximum nitrogen concentration of 300 ppm for the fresh hydrogen is established. Supplying fresh hydrogen into the anode subsystem is thus always also associated with supplying nitrogen.Nitrogen is an inert gas for a fuel cell that reduces cell voltage and causes efficiency losses. Furthermore, an increased nitrogen concentration is associated with a reduced hydrogen concentration, which can lead within the fuel cell or the stack to local undersupply and thus to damage, in particular to the membrane and consequently to a reduced service life. This is always taken into account during the operation of a fuel cell system. Therefore, during operation, typically more hydrogen is supplied to the anode side than would be required for the reaction in the event of an applied current draw. The ratio of the amount of hydrogen supplied on the anode side and the amount of hydrogen required for the reaction is referred to as the anode stoichiometry, or, if the anode side is viewed in isolation, is referred to briefly as stoichiometry. When defining the anode stoichiometry, it should be noted in particular that both too high and too low anode stoichiometry cause damage or accelerated aging phenomena on the stack.The present invention is therefore concerned with fine adjustment of the hydrogen concentration in the anode gas and, associated therewith, adjustment of the anode stoichiometry.To achieve the object, the method having the features of claim 1 is proposed. Advantageous further developments of the invention can be found in the dependent claims. Furthermore, a control device for executing steps of the method is specified.Disclosure of the InventionA method for operating a fuel cell system is proposed, comprising an anode subsystem for supplying at least one stack with hydrogen in order to generate a stack stream, wherein the hydrogen is supplied in an anode gas which accumulates with nitrogen during operation. A nitrogen concentration in the anode gas is controlled by temporary discharge of anode gas from the anode subsystem via a controllable purge valve as part of a purge operation and replacement by the supply of fresh hydrogen. According to the invention, the following steps are carried out: (a) determination of a hydrogen concentration in the anode gas, (b) inference of the hydrogen concentration and the stack current to an actual anode stoichiometry, (c) comparison of the actual anode stoichiometry with a desired anode stoichiometry, (d) indirect adaptation of the hydrogen concentration by varying the duration and / or frequency of the purge process if the actual anode stoichiometry deviates from the desired anode stoichiometry.With the proposed method, the hydrogen concentration is adjusted by specific control of the purge processes and thus enables compliance with a predefined desired anode stoichiometry. The determination of the hydrogen concentration in the anode gas described in step (a) can be carried out by calculation or sensor. Step (b) refers to a relationship that exists between hydrogen concentration, stack current and actual anode stoichiometry. This relationship is to be determined empirically. The anode stoichiometry λ is defined here as where m H2,tats represents the amount of hydrogen actually supplied to the anode side and m H2,st represents the amount of hydrogen required for complete combustion given a predefined stack current drop. The former is typically set by setting an anode gas mass flow. Due to the permanent enrichment of the anode gas with nitrogen and the associated successive lowering of the hydrogen concentration, the amount of hydrogen actually supplied to the anode side varies continuously. Consequently, the actual anode stoichiometry also diverges continuously from the target anode stoichiometry to the extent that the nitrogen enrichment of the anode gas progresses. The amount of this divergence at a specific time is determined in step (c).The adjustment of the hydrogen concentration described in step (d) takes place by means of the active use of the purge process, more precisely the change in its frequency and / or duration, in order to create capacity for fresh hydrogen in the anode gas by removing nitrogen from the fuel cell system. Full control over the nitrogen concentration in the anode gas is thus always exerted, thus preventing the negative consequences of an excessively high nitrogen concentration. Furthermore, it is made possible for the fuel cell system to operate permanently at the desired operating point, with the result that states used for the control correspond better to those actually prevailing. The advantageous consequence is an increase in the precision with which the fuel cell can be adjusted to a specific operating point.It is furthermore proposed that, for carrying out method step (a), the nitrogen concentration in the anode gas is balanced as the sum of a maximum possible nitrogen fraction in the fresh hydrogen and a nitrogen fraction transmitted via diffusion processes within the at least one stack, and the at least hydrogen concentration present is estimated by calculation from the nitrogen concentration in the anode gas determined in this way. This preferred embodiment is advantageous because it carries out an estimation of the hydrogen concentration in the form of a worst-case scenario towards the safe side without using additional components such as measuring devices. Uncertainties in the determination of state variables such as tolerance fields do not have any negative effects due to the estimation on the safe side. This preferred embodiment is based on purely simulative considerations and can also be used well as an upper limit of the nitrogen fraction or a lower limit of the hydrogen fraction in the anode gas in combination with other methods.It is furthermore proposed that, in order to carry out method step (a), the hydrogen concentration is determined by means of a sensor system at the inlet and / or at the outlet of the at least one stack. Typically, such a sensor system is already installed for monitoring the operation, whereby already existing data is used. At the same time, measurements usually provide the most realistic value and thus better map the reality than computational approaches.In a development of the invention, it is proposed that, in order to carry out method step (a), the composition of the anode gas is modeled and the hydrogen concentration is determined by calculation with the aid of the data model. With this preferred embodiment, which can be used both alone and in combination with other methods for determining the hydrogen concentration, a usually very realistic image of the system states is generated without the need for new sensors. In particular in conjunction with a sensory determination of the hydrogen concentration, operating errors can be detected via a divergence that occurs between model predictions and measured values.It is furthermore proposed that, in order to carry out method step (a), a pressure loss within a component of the anode subsystem, for example of the at least one stack or of a water separator, is determined as a function of the stack flow and the hydrogen concentration is deduced by means of this, a characteristic curve field which is produced in advance individually for the system preferably being used. This preferred embodiment uses a relationship between pressure loss in a component and stack flow. This relationship is to be determined in advance on a system-specific basis. Advantageously, once it has been determined and written down in a characteristic curve, it can be used at any time during operation to determine the hydrogen concentration. For this purpose, no computationally intensive modeling is necessary. Preferably, to increase the accuracy of this preferred embodiment, a temperature signal from a suitable sensor system is used at at least one location in the anode circuit.In addition, it is proposed that, in order to carry out method step (a), the hydrogen concentration is determined on the basis of a predefined stack current and a predefined switching frequency of the purge valve via a characteristic curve field which is created in advance on a system-specific basis and represents the respective profile of the hydrogen concentration in the anode gas over the stack current for different switching frequencies of the purge valve. This embodiment is advantageous in that the hydrogen concentration is derived directly from two parameters that can be freely adjusted during operation. In this case, complex switching strategies of the purge valve can also be stored in the characteristic curve field. The characteristic curves are available at any time during operation and supply the required hydrogen concentration quickly and without modelling or computing effort.It is furthermore proposed that, in order to carry out method step (b), a conclusion is drawn from the stack stream to the actual anode stoichiometry by means of a characteristic curve field which is created in advance on a system-specific basis and represents the relationship between the stack stream and the actual anode stoichiometry for different hydrogen concentrations in the anode gas. Carrying out step (b) with the aid of a characteristic curve field is advantageous because, after initial determination, these are available at any time during operation and have dependencies neither on other components nor on a computation-intensive modeling. Preferably, a setpoint anode stoichiometry λ SOLL is also depicted on the characteristic curve field that relates the stack current to an actual anode stoichiometry λ IST by means of which the comparison required in step (c) is determined directly in the form of a difference formation λ IST- λ SOLL.It is furthermore proposed that in method step (d) the frequency and / or duration of the purge process is reduced if the actual anode stoichiometry has been determined to exceed the desired anode stoichiometry. This is based on the following causal relationship, wherein a system state defined by the target anode stoichiometry λ SOLL is referred to below as abbreviated as the target state and a system state defined by the actual anode stoichiometry λ IST is referred to below as abbreviated as the actual state:If λ IST > λ SOLL the definition of anode stoichiometry presented above follows: where m H2_st represents the amount of hydrogen required for a given current drop, which is the same in both the actual state and the desired state, and m H2_tats_IST and m H2_tats_SOLL each represent the amount of hydrogen actually supplied in the actual and the desired state. Consequently, it follows that in the actual state more hydrogen is available for combustion than is required by the desired state. This circumstance is corrected by reducing the duration and / or frequency of the purge process, thus depositing less nitrogen from the fuel cell system. As a result, the nitrogen concentration in the anode gas increases at the sacrifice of the hydrogen concentration and the amount of hydrogen supplied to the anode side decreases, with the result that less hydrogen is supplied in the actual state than is provided in the target state. In this case, an increased nitrogen content is deduced, which is separated from the fuel cell system by increasing the duration and / or frequency of the purge process.Likewise, it is proposed that in method step (d) the frequency and / or duration of the purge process is increased if the actual anode stoichiometry has fallen below the desired anode stoichiometry. In this case, λ IST< λ SOLL applies and the aforementioned causal chain with the opposite operator applies: as a result, less hydrogen is supplied in the actual state than is provided in the desired state. In this case, an increased nitrogen content is deduced, which is separated from the fuel cell system by increasing the duration and / or frequency of the purge process. With these two preferred embodiments shown, the adjustment of the hydrogen concentration in the anode gas is based on a causal relationship which can be followed in a straight line.Furthermore, a control device is proposed which is configured to carry out steps of a method according to the invention. This preferably stores the characteristic fields determined individually for the system, evaluates measurement data and / or carries out the comparison between the actual air ratio and the desired air ratio. Since a fuel cell system is typically already equipped with a control unit which carries out a wide-ranging control over controllable components of the fuel cell system, the control of a method according to the invention is preferably to be placed there.A preferred embodiment of the invention is illustrated in more detail below with reference to figures. FIG. 1 shows a schematic illustration of a fuel cell system FIG. 2 shows a schematic illustration of a method sequence according to the inventionDESCRIPTION OF THE FIGURESFIG. 1 shows, by way of example, a schematic illustration of a fuel cell system 1, the central constituent of which constitutes stack 2. This comprises an anode side 5 belonging to an anode subsystem 3 and a cathode side 6 belonging to a cathode subsystem 4. hydrogen is supplied to the stack 2 via the anode subsystem 3. This reacts within the stack 2 with air supplied on the cathode side 6 to form water. During this process, water and nitrogen diffuse over to the anode side 5 and accordingly enrich the anode gas. The anode gas, which is enriched with water and nitrogen and still contains unused hydrogen, is discharged from the stack 2 for recirculation. Water is separated from the anode gas in a water separator 8, wherein the water separator 8 is emptied by temporary opening of a drain valve 9. The separated water is introduced into an exhaust air path 10 located in the cathode subsystem 4 and leaves the fuel cell system 1 via an exhaust system 11 together with air discharged from the cathode side 6.The anode gas largely freed of water is in each case led out of the water separator 8 and recirculated via a jet pump 12. Instead of a jet pump 12, an active pump module or a combination of the two can also be used. In the jet pump 12, fresh hydrogen is supplied from a tank 7 to the anode gas. The hydrogen stored in the tank 7 is contaminated with nitrogen and typically contains a maximum nitrogen concentration of 300 ppm. The supply takes place via an actuatable hydrogen metering valve 13.Both due to the diffusion processes within the stack 2 and due to the contamination of the fresh hydrogen, the anode gas accumulates permanently with nitrogen. In order to remove this nitrogen from the fuel cell system 1, a so-called purge process is provided. In this purge operation, anode gas is discharged inside or downstream of the water separator by temporarily opening a purge valve 14. The anode gas discharged via the purge valve 14 is introduced into the exhaust air path 10 of the cathode subsystem 4 and is separated out via an exhaust system 11. The hydrogen concentration in the exhaust gas is monitored by a measuring device 15 in order to avoid explosive hydrogen-air mixtures. Within the anode subsystem 3, the anode gas discharged in the purge process is compensated for by supplying fresh hydrogen via the hydrogen metering valve 13. The duration and frequency of the purge operation are variable and controllable.In operation, the correct adjustment of the hydrogen concentration in the anode gas and thus of the anode stoichiometry is of particular importance. It has in particular effects on the service life of the fuel cells within a stack 2. in a method according to the invention, the hydrogen concentration in the anode gas is controlled by adjusting the frequency and / or duration of the purge processes. Such a method according to the invention is illustrated by way of example in FIG. 2. In a first method step S 1, the hydrogen concentration in the anode gas is determined. This is preferably effected via a corresponding sensor system 16 at the inlet and / or at the outlet of the anode side 5 of the stack 2. in other preferred embodiments the method step S 1 is effected by balancing the nitrogen mass flows entering the anode gas and estimating the resulting at least present hydrogen concentration, by modelling the anode gas composition or in some other way.In a subsequent method step S 2, the hydrogen concentration in the anode gas concludes that the anode stoichiometry is actual on the anode side 5 within the stack 2. In a preferred embodiment, for this purpose, the stack current concludes the actual anode stoichiometry via a characteristic curve field which is created in advance for individual systems and represents the relationship between the stack current and the actual anode stoichiometry for different hydrogen concentrations in the anode gas.In a subsequent method step S 3, the actual anode stoichiometry ascertained in method step S 2 is compared with a desired anode stoichiometry. The desired anode stoichiometry results from a target operating point of the fuel cell system defining the control or regulation target.If the actual anode stoichiometry is greater than the desired anode stoichiometry, the hydrogen concentration in the anode gas is above its target value. In this case, in a method step S4.1, the frequency and / or the duration of the purge processes is reduced. As a result, the nitrogen concentration in the anode gas increases and the hydrogen concentration decreases.Conversely, if the actual anode stoichiometry is less than the target anode stoichiometry, the hydrogen concentration in the anode gas is below its target value. In this case, in a method step S4.2, the frequency and / or the duration of the purge processes is increased. As a result, the nitrogen concentration in the anode gas decreases and the hydrogen concentration increases.

Claims

Method for operating a fuel cell system (1) comprising an anode subsystem (3) for supplying at least one stack (2) with hydrogen for generating a stack stream, wherein the hydrogen is supplied in an anode gas which accumulates with nitrogen during operation and wherein a nitrogen concentration in the anode gas is controlled by temporary discharge of anode gas from the anode subsystem (3) via a controllable purge valve (14) within the scope of a purge operation and replacement by supply of fresh hydrogen, characterized bythe following steps: (a) determination of a hydrogen concentration in the anode gas, (b) conclusion of the hydrogen concentration and the stack stream on an actual anode stoichiometry, (c) comparison of the actual anode stoichiometry with a desired anode stoichiometry, (d) Indirect adaptation of the hydrogen concentration by varying the duration and / or frequency of the purge process if the actual anode stoichiometry deviates from the desired anode stoichiometry.Method according to Claim 1, characterized in that, in order to carry out method step (a), the nitrogen concentration in the anode gas is balanced as the sum of a maximum possible nitrogen fraction in the fresh hydrogen and a nitrogen fraction transmitted via diffusion processes within the at least one stack, and the at least hydrogen concentration present is estimated by calculation from the nitrogen concentration in the anode gas determined in this way.Method according to one of the preceding claims, characterized in that, in order to carry out method step (a), the hydrogen concentration is determined by means of a sensor system (16) at the inlet and / or at the outlet of the at least one stack (2).Method according to one of the preceding claims, characterized in that, in order to carry out method step (a), the composition of the anode gas is modeled and the hydrogen concentration is determined by calculation with the aid of the model.Method according to one of the preceding claims, characterized in that, in order to carry out method step (a), a pressure loss within a component of the anode subsystem, for example of the at least one stack (2) or of a water separator (8), is determined as a function of the stack flow and the hydrogen concentration is deduced by means of this, a characteristic curve field which is produced in advance individually for the system preferably being used as a basis.Method according to one of the preceding claims, characterized in that, in order to carry out method step (a), the hydrogen concentration is determined on the basis of a predefined stack current and a predefined switching frequency of the purge valve (14) via a characteristic field which is created in advance on a system-specific basis and represents the respective profile of the hydrogen concentration in the anode gas over the stack current for different switching frequencies of the purge valve (14).Method according to one of the preceding claims, characterized in that, in order to carry out method step (b), the stack current concludes the actual anode stoichiometry via a characteristic curve field which is produced in advance on a system-specific basis and represents the relationship between the stack current and the actual anode stoichiometry for different hydrogen concentrations in the anode gas.Method according to one of the preceding claims, characterized in that in method step (d) the frequency and / or the duration of the purge process is reduced if the actual anode stoichiometry has detected the target anode stoichiometry to be exceeded.Method according to one of the preceding claims, characterized in that in method step (d) the frequency and / or duration of the purge process is increased if the actual anode stoichiometry has fallen below the desired anode stoichiometry as determined.Control device which is configured to carry out steps of a method according to the invention.

Images