Load balancing method, SOFC cluster

Abstract

The invention is based on a load distribution method for an SOFC cluster (10) which has a plurality of SOFC units (12), with which an operating load of the SOFC cluster (10) is distributed among the individual SOFC units (12) of the SOFC cluster (10). It is proposed that an individual operating load of the individual SOFC units of the SOFC cluster is determined by optimizing a result of an objective function of the SOFC cluster.

Description

State of the art

[0001] A load distribution method for a SOFC cluster comprising multiple SOFC units has already been proposed, with which an operating load of the SOFC cluster is distributed among the individual SOFC units of the SOFC cluster. Disclosure of the invention

[0002] The invention is based on a load distribution method for an SOFC cluster having a plurality of SOFC units, with which an operating load of the SOFC cluster is distributed among the individual SOFC units of the SOFC cluster.

[0003] It is proposed that an individual operating load of the individual SOFC units of the SOFC cluster is determined by optimizing a result of an objective function of the SOFC cluster.

[0004] The inventive design of the load distribution method advantageously makes it possible to provide high efficiency, since in particular a high degree of effectiveness can be achieved by optimizing the target function. Low costs can advantageously be achieved, since gas consumption can be reduced, particularly with the same electrical output power, by optimizing the target function. A particularly broad power spectrum can advantageously be provided, since in particular very small operating loads of the SOFC cluster can be provided by switching off individual SOFC units to reduce a minimum power of the SOFC cluster. High flexibility can advantageously be provided, since in particular the individual SOFC units can be operated independently of one another.

[0005] The SOFC (Solid Oxide Fuel Cell) unit is preferably designed as a solid fuel cell. An SOFC cluster (Solid Oxide Fuel Cell) is preferably formed from at least one, preferably at least 5 or at least 10, SOFC units. It is also preferably conceivable for an SOFC cluster to have far more than 10 SOFC units. Alternatively, the SOFC unit could also be designed as a comparable fuel cell. The SOFC units are preferably constructed identically. The SOFC units are preferably electrically connected to one another. Alternatively, however, different SOFC units could also be combined in an SOFC cluster. Energy, in particular electrical energy, is preferably generated in the SOFC units by a reaction between a fuel gas, for example hydrogen, and a reaction gas, for example oxygen.Preferably, in the load distribution method, the operating load, which is requested, for example, by a consumer in a load request, is individually distributed among the individual SOFC units. Preferably, the operating load is in the form of a power, in particular electrical power. Preferably, each SOFC unit is operated with a different operating load. Preferably, the operating load of the SOFC cluster is composed of the sum of the, in particular individual, operating loads of the individual SOFC units. In particular, the individual operation of each individual SOFC unit does not preclude the individual SOFC units from being operated with the same operating load in at least one operating state. Preferably, the load is only distributed individually among the individual SOFC units of the SOFC cluster in a partial load range of the SOFC cluster.Preferably, the SOFC units have at least substantially the same operating load in full load operation.

[0006] Preferably, the operating load of the SOFC cluster is distributed among the individual SOFC units by optimizing the objective function. The objective function is preferably formed as a sum. The objective function preferably represents a mixed-integer optimization problem that can be solved using solution algorithms. The objective function preferably includes at least the operating load, in particular of the SOFC units forming the SOFC cluster, as a variable. The result of the objective function is preferably minimized by the individual operating loads of each of the individual SOFC units. Alternatively, the result of the objective function could also be maximized.An “objective function” is to be understood in particular as a mathematical function which has a functional value, such as costs, efficiency, power or the like, as a result value, which depends on at least one input parameter, such as an operating load, an aging parameter, an efficiency, gas costs and / or a feed-in tariff or the like. In particular, the objective function describes the load distribution. An “individual operating load” is to be understood in particular as a separate load for each SOFC unit. In this context, “at least substantially” is to be understood in particular as meaning that a deviation from a predetermined value deviates in particular by less than 25%, preferably less than 10% and particularly preferably less than 5% of the predetermined value.

[0007] It is further proposed that the objective function be formed as a cost function. Advantageously, low costs can be provided since, in particular, the cost function is minimized. The cost function preferably has costs, preferably operating costs, as a result. For example, the cost function could be minimized by operating the individual SOFC units of the SOFC cluster at maximum efficiency. When optimizing the cost function, it is preferably determined whether, given the current feed-in tariff, the SOFC cluster should be operated with a higher operating power, which is fed into a, preferably public, power grid. Alternatively, it is conceivable that the additionally generated energy is stored. The cost function of the SOFC cluster is preferably formed as a sum of the cost functions of the individual SOFC units forming the SOFC cluster.The cost function preferably includes at least operating costs, purchase costs and / or a feed-in tariff. In particular, the operating costs are designed as operating load-dependent operating costs and / or operating costs for standby operation. The operating costs are preferably determined by determining the individual operating load for each SOFC unit and offsetting this against the efficiency of the SOFC unit. The purchase costs are preferably formed at least by the gas costs. The feed-in tariff is preferably formed as the cost, in particular the negative cost, for the energy fed into a preferably public power grid. It is conceivable that the cost function includes further costs, such as maintenance costs, switch-on and / or switch-off costs, wear and tear costs or the like. Equation (1) shows one of the possible cost functions by way of example. Preferably, “Z” in Eq.1 represents the total costs of the SOFC cluster, in particular in € / h. Preferably, “P. ac , i ” represents the individual operating load of one of the SOFC units, where “i” denotes the number of SOFC units in the SOFC cluster. Preferably, “η αc,i ” represents the efficiency of one of the SOFC units. Preferably, “n SB ” represents the number of SOFC units in standby mode. Preferably, “P SB “ represents power consumption in standby mode. For example, “cost Gas ” represents the gas costs, especially in € / kilowatt hour. Preferably, “P Bezug “ represents the power drawn from the grid for the operation of the SOFC unit. Preferably, “cost Bezug ” represents the cost of the power drawn from the grid for the operation of the SOFC unit. Preferably, “P Speis “ represents the power fed in. Preferably, “cost Speis “ represents the remuneration for the power fed into the grid. Z=(ΣinUnitsPac,iηac,i+nSB⋅PSB)⋅costGas+PBezug⋅costBezug−PSpeis⋅costSpeis

[0008] Furthermore, it is proposed that the objective function takes into account at least one efficiency curve of the individual SOFC units forming the SOFC cluster, which efficiency curve is determined based on the aging state of the individual SOFC units (cf. Eq. 1). Advantageously, a particularly high efficiency can be provided since, in particular, the aging state of the SOFC units is used as a basis for determining the efficiency curve. The efficiency curve is preferably described as a mathematical function. It is conceivable that the efficiency curve is determined as a measurement, preferably carried out regularly, and / or is estimated by a simulation and / or by an efficiency prediction model. In particular, the efficiency curve is formed by support points. Preferably, the efficiency curve is determined / estimated for each SOFC unit forming the SOFC cluster.It is conceivable that the efficiency curve is determined / estimated for several SOFC units, such as SOFC blocks with several SOFC units. An “efficiency curve” should be understood in particular as a profile of the efficiency of at least one of the SOFC units or of the SOFC cluster, which depends on the operating load. In particular, the efficiency curve is determined at least for / in an operating range defined by a minimum load and a maximum load, for example full load. It is conceivable that the efficiency curve has a linear and / or constant profile. An “aging state” should be understood in particular as a chemically and / or physically induced change, preferably a reduction, in the performance of a unit at a specific point in time.

[0009] Furthermore, it is proposed that at least one gas price and / or at least one electricity price be taken into account in the objective function (cf. Eq. 1). Advantageously, low costs can be provided, since in particular the current feed-in tariff and / or a gas price are taken into account. Preferably, operating costs, in particular those dependent on the gas price, are determined. Preferably, the operating costs are formed at least from the gas price. Preferably, the operating costs are compared with the feed-in tariff. It is conceivable that the operating load of the SOFC cluster is increased until the electricity costs are equal to the feed-in tariff. A "gas price" should be understood in particular to mean monetary costs for the gas, in particular fuel gas, for electrical energy generation per SOFC unit. An "electricity price" should be understood in particular to mean the monetary costs for an amount of energy, for example, a kilowatt hour.In particular, the feed-in tariff should be understood as negative monetary costs for a quantity of energy.

[0010] In addition, it is proposed that in at least one method step for providing a minimum load of the SOFC cluster, at least one SOFC unit of the SOFC cluster is switched off or placed in standby mode and at least one SOFC unit of the SOFC cluster continues to operate. Advantageously, a particularly broad power range can be provided since, in particular, very small operating loads of the SOFC cluster can also be provided by switching off individual SOFC units to reduce the minimum power of the SOFC cluster. Preferably, the minimum load of the SOFC cluster is formed by the minimum loads of the individual SOFC units forming the SOFC cluster. Preferably, the minimum load of the SOFC cluster is formed at least substantially by the minimum load of one of the SOFC units. Preferably, a minimum number of SOFC units are switched off during operation.Preferably, the load is distributed among the individual SOFC units after determining which SOFC units are to be shut down. A "standby mode" is understood to mean, in particular, a standby state of the SOFC unit that allows immediate startup of the SOFC unit. In this standby state, the SOFC unit consumes a very small amount of standby energy, particularly compared to operation.

[0011] Furthermore, it is proposed that the SOFC unit is switched off depending on an aging parameter which describes at least one aging state, if an SOFC unit has to be switched off to optimize the objective function. Advantageously, particularly low costs can be provided since, in particular, by taking the aging state into account, the SOFC unit can be operated at a high level of efficiency. Advantageously, a long maintenance interval can be provided since, in particular, by taking the aging state of the SOFC unit into account, a service life can be extended. Preferably, the aging state is described by the aging parameter. For example, the aging parameter could be specified as a percentage. For example, the aging parameter could be described as a value between zero and one.For example, an SOFC unit in its new state could be described by the aging parameter 100% and / or 1. For example, the lower the value for the aging parameter, the worse the aging state. Preferably, the SOFC unit with the lowest aging parameter is shut down. Preferably, the SOFC units with increasing aging parameter values could be shut down.

[0012] In addition, it is proposed that a load distribution among the individual SOFC units be recalculated for each change in the load requirement and / or the aging state, for example at regular intervals and / or based on new input parameters and / or based on an implemented load profile. Particularly low costs can advantageously be provided because, in particular, the operation of the SOFC cluster can be adapted to a change in the load requirement and / or the aging state. Preferably, the change in the load requirement is specified by a consumer, such as the energy demand of a factory hall. Preferably, the load requirement is recalculated in real time. Alternatively, the load requirement could also be recalculated every minute, hourly, daily, or at comparable intervals. It is conceivable that an implemented load profile, in particular one created in advance, is run.The implemented load profile could have different load cycles, such as day / night or full / partial load operation, or similar. It is also conceivable to combine the load demand and the implemented load profile. Preferably, the load distribution is recalculated when a new input parameter is available.

[0013] Furthermore, an SOFC cluster with at least one control and / or regulating unit, in particular a microcontroller, for implementing the load distribution method is proposed, which is configured to determine an individual operating load for the individual SOFC units of the SOFC cluster in order to optimize a target function of the SOFC cluster. Advantageously, high efficiency can be provided, since, in particular, a high degree of effectiveness can be achieved by optimizing the target function. Advantageously, low costs can be provided, since, in particular, with the same electrical output power, gas consumption can be reduced by optimizing the target function.Advantageously, a particularly low power can be provided, since in particular even very small operating loads of the SOFC cluster can be provided by switching off individual SOFC units to reduce the minimum power of the SOFC cluster. Advantageously, a high level of flexibility can be provided, since in particular the individual SOFC units can be operated independently of one another. It is conceivable for the SOFC cluster to be designed so that it can be expanded by any desired number of SOFC units. A “control and / or regulating unit” should be understood in particular to mean a unit with at least one control electronics unit. “Control electronics” should be understood in particular to mean a unit with a processor unit and with a memory unit as well as with an operating program stored in the memory unit."Configured" should be understood in particular as specifically programmed, designed, and / or equipped. The fact that an object is configured for a specific function should be understood in particular as meaning that the object fulfills and / or executes this specific function in at least one application and / or operating state.

[0014] The load distribution method according to the invention and the SOFC cluster according to the invention are not intended to be limited to the application and embodiment described above. In particular, the load distribution method according to the invention and the SOFC cluster according to the invention can have a number of individual elements, components, units, and method steps that differs from the number stated herein to fulfill a functionality described herein. Furthermore, in the value ranges specified in this disclosure, values within the stated limits are also to be considered disclosed and can be used arbitrarily. drawing

[0015] Further advantages will become apparent from the following description of the drawings. The drawing illustrates an exemplary embodiment of the invention. The drawings, the description, and the claims contain numerous features in combination. Those skilled in the art will also expediently consider the features individually and combine them into further meaningful combinations.

[0016] They show: Fig. 1 a SOFC cluster with several SOFC units and a control and / or regulation unit and Fig. 2 a schematic flow diagram of a load distribution procedure. Description of the embodiment

[0017] The Fig. 1 shows an SOFC cluster 10 (Solid Oxide Fuel Cell cluster). The SOFC cluster 10 has 10 SOFC units 12. The SOFC cluster 10 could also have 5 SOFC units 12 or 2x5. It is conceivable that the SOFC cluster is formed from any number of SOFC units 12. The SOFC cluster 10 has a control and / or regulating unit 16. The control and / or regulating unit 16 is designed as a microcontroller. Alternatively, the control and / or regulating unit 16 could also be designed as a cloud or a smartphone or the like. The control and / or regulating unit 16 is designed to determine an individual operating load for the individual SOFC units 12 of the SOFC cluster 10. The individual operating loads of the SOFC units 12 are designed to determine an objective function Eq. 1 of the SOFC cluster 10. The objective function Eq. 1 includes a feed-in tariff. The objective function Eq. 1 includes gas procurement costs. The objective function Eq.1 shows operating loads of the SOFC units 12. The operating loads of the SOFC units 12 are expressed as electrical power. The objective function Eq. 1 shows an efficiency curve per SOFC unit 12. The efficiency curve depends on the aging state of the SOFC units 12. The objective function Eq. 1 shows standby costs of the SOFC units 12. It is conceivable that the objective function Eq. 1 has further parameters. The SOFC cluster 12 has a minimum load. The minimum load of the SOFC cluster 10 is formed by a minimum load of one of the SOFC units 12. Alternatively, the minimum load of the SOFC cluster 10 could be formed by the minimum load of the SOFC unit 12 with the lowest minimum load. The optimal distribution of the operating load among the various SOFC units 12 occurs when the objective function Eq. 1 is at a minimum. The objective function Eq. 1 is formed as a cost function. Alternatively, the objective function could be Eq.1 can also be designed as an efficiency function and / or power function and / or lifetime function or another target function that appears reasonable to the person skilled in the art.

[0018] The Fig. Figure 2 shows a schematic flow diagram of a load distribution procedure for the SOFC cluster 10.

[0019] In a method step 18, it is determined whether a load requirement on the SOFC cluster 12 is lower than the sum of the minimum loads of the individual SOFC units 12. It is conceivable that the load requirement is selected to be greater than required. It is conceivable that if the load requirement is greater than required, excess energy is fed into another power grid / public power grid. If the load requirement on the SOFC cluster 12 is lower than the minimum loads of the SOFC units 12, it is determined how many SOFC units 12 can be switched off. For this purpose, a switch-off heuristic is run. During the switch-off heuristic, SOFC units 12 are iteratively switched off until the load requirement can be met. In method step 18, one or more of the SOFC unit(s) 12 of the SOFC cluster 10 are switched off in order to provide the load requirement to the SOFC cluster 10.Alternatively, to provide the minimum load of the SOFC cluster 10, the SOFC unit 12 of the SOFC cluster 10 is placed in standby mode. As few SOFC units 12 of the SOFC cluster 10 are shut down or placed in standby mode as necessary to achieve the determined load requirement of the SOFC cluster 10. In method step 18, at least one SOFC unit 12 of the SOFC cluster 10 continues to operate.

[0020] In a method step 20, it is determined which SOFC unit 12 of the SOFC cluster 10 will be shut down. The SOFC unit 12 is shut down depending on an aging parameter if an SOFC unit 12 of the SOFC cluster 10 must / can be shut down to optimize the objective function Eq. 1. The aging parameter describes an aging state of the SOFC unit 12. The aging parameter is specified between zero and one. A new state of the SOFC unit has a value for the aging parameter of 1. As the SOFC unit ages, the value for the aging parameter decreases. The SOFC unit 12 with the lowest / worst aging state is shut down first. The SOFC units 12 with ascending aging state are shut down.

[0021] In a method step 22, an individual operating load is determined for the individual SOFC units 12 of the SOFC cluster 10. The individual operating loads are determined for the SOFC units 12 that are not shut down / continued to operate in method steps 18 / 20. The objective function Eq. 1 of the SOFC cluster 10 is optimized using the individual operating loads of the SOFC units 12. A mixed-integer optimization problem is described using the objective function Eq. 1. The optimal individual operating loads of the SOFC units 12 are determined by solving the mixed-integer optimization problem with a solution algorithm. The objective function Eq. 1 takes into account an efficiency curve for each of the individual SOFC units 12 forming the SOFC cluster 10. The efficiency curves of each of the SOFC units 12 are determined based on the aging state of the individual SOFC units 12.The efficiency curves of each of the SOFC units 12 are divided into discrete sections using support points. The efficiency curves of each of the SOFC units 12 are formed over an operating range of the assigned SOFC units 12. The operating range is defined as the range between the minimum load of the SOFC unit 12 and the maximum load of the SOFC unit at full load. A gas price is taken into account in the objective function Eq. 1. An electricity price for the feed-in tariff is taken into account in the objective function Eq. 1. It is conceivable that additional parameters, such as wear costs, maintenance costs, or startup and / or shutdown costs, are taken into account.

[0022] In a method step 24, a load distribution is recalculated at regular intervals. Alternatively or additionally, the load distribution is recalculated if new input parameters are available. Alternatively or additionally, it is conceivable that a load distribution among the individual SOFC units 12 is calculated based on an implemented load profile. For example, the load profile could be configured as a day / night distribution or include breaks. A new load requirement is determined for each change in the load requirement and / or the aging state and / or other parameters.

Claims

[1] Load distribution method for a SOFC cluster (10) having a plurality of SOFC units (12), with which an operating load of the SOFC cluster (10) is distributed among the individual SOFC units (12) of the SOFC cluster (10), characterized by that an individual operating load of the individual SOFC units of the SOFC cluster is determined by optimizing a result of an objective function of the SOFC cluster. [2] Load distribution method according to claim 1, characterized by that the objective function is formed as a cost function. [3] Load distribution method according to claim 1 or 2, characterized by that the objective function takes into account at least one efficiency curve of the individual SOFC units (12) forming the SOFC cluster (10), which is determined on the basis of the aging state of the individual SOFC units (12). [4] Load distribution method according to claim 1 or 2, characterized bythat at least one gas price and / or at least one electricity price is taken into account in the objective function. [5] Load distribution method according to one of the preceding claims, characterized by that in at least one method step (18) for providing a minimum load of the SOFC cluster (10), at least one SOFC unit (12) of the SOFC cluster (10) is switched off or put into a standby mode and at least one SOFC unit (12) of the SOFC cluster (10) continues to be operated [6] Load distribution method according to one of the preceding claims, characterized by that the SOFC unit (12) is switched off as a function of an aging parameter which describes at least one aging state if an SOFC unit (12) of the SOFC cluster (10) has to be switched off in order to optimize the target function. [7] Load distribution method according to one of the preceding claims, characterized bythat a load distribution to the individual SOFC units (12) is recalculated for each change in the load requirement and / or the aging state, for example at regular intervals and / or based on new input parameters and / or based on an implemented load profile. [8] SOFC cluster (10) with at least one control and / or regulating unit (16), in particular a microcontroller, for carrying out the load distribution method according to one of the preceding claims, which is designed to determine an individual operating load for the individual SOFC units (12) of the SOFC cluster (10) in order to optimize a target function of the SOFC cluster (10).

Images