EP2777122A1

EP2777122A1 - Method for providing control power with an energy store using tolerances in determining the frequency deviation

Info

Publication number: EP2777122A1
Application number: EP12781074.5A
Authority: EP
Inventors: Georg Markowz; Wolfgang SCHWEISSTHAL; Carsten Kolligs; Holger BREZSKI; Wolfgang Deis; Michael Igel; Anna FLEMMING; Dennis GAMRAD; Sébastien COCHET; Stefan Winternheimer
Original assignee: Evonik Degussa GmbH; Evonik Industries AG; Steag Power Saar GmbH
Current assignee: Steag Power Saar GmbH; Evonik Operations GmbH
Priority date: 2011-11-10
Filing date: 2012-10-29
Publication date: 2014-09-17
Also published as: US20140327404A1; US9667071B2; DE102011055229A1; WO2013068256A1

Abstract

The invention relates to a method for providing control power for a power network, in which at least one energy store connected to the power network supplies energy to the power network as necessary and/or absorbs energy from the power network as necessary. According to the invention, the control power is generated as a function of a frequency deviation from a target value of a network frequency, wherein the tolerance with respect to the frequency deviation is used in order to adjust the charging state of the energy store simultaneously with the provision of the control power by the energy store. The invention also relates to a device for carrying out such a method, in which the device comprises a control and an energy store, wherein the device is or can be connected to a power network, the control is connected to the energy store and regulates the control power released by and/or absorbed by the energy store.

Description

Method of providing control power with an energy storage using tolerances in the determination of

Frequency deviation "

description

The invention relates to methods for providing control power for a power grid in which at least one connected to the power grid energy storage supplies the power grid as needed energy and / or from the power grid as needed absorbs energy, the control power in response to a frequency deviation from a nominal value of a grid frequency is provided.

The invention also relates to a device for carrying out such a method.

Electricity grids are used to distribute electricity from many energy generators in large areas to many users and to supply households and industry with energy. Energy producers mostly in the form of power plants provide the required energy. As a rule, power generation is planned and provided based on the forecasted consumption.

Both the generation and the consumption of energy can lead to unplanned fluctuations. These can arise on the energy producer side, for example, in that a power plant or part of the power grid fails or, for example, in the case of renewable energies such as wind, that the energy production is higher than predicted. Consumers may also experience unexpectedly high or low consumption. For example, the failure of a portion of the grid, which cuts off some consumers from the power supply, can lead to a sudden reduction in power consumption.

This generally results in power network fluctuations due to unplanned and / or short-term variations in power generation and / or consumption. The desired AC frequency is, for example, in Europe 50,000 Hz. This frequency is often referred to as the desired frequency. A reduction of consumption compared to the plan leads to an increase in the frequency of planned power generated by the energy producers, the same applies to an increase in electricity production compared to the plan for scheduled consumption. On the other hand, a reduction in the output of the energy producers leads to a reduction in the network frequency at scheduled consumption, as well as an increase in consumption compared with the plan when scheduled production.

For reasons of network stability, it is necessary to keep these deviations within a defined range. For this purpose, depending on the amount and direction of the deviation, it is necessary to provide specifically positive control power by connecting additional generators or switching off consumers, or negative balancing power by shutting down generators or adding consumers. There is a general need for economical and efficient provision of these control powers, where the requirements for the capacities to be kept and the dynamics of the control power sources or sinks can vary depending on the characteristic of the power grid.

In Europe, for example, there is a set of rules (UCTE Handbook) that describes three different categories of control power. This also defines the respective requirements for the control power types. The types of balancing power differ, among other things, in the demands on the dynamics and the duration of the service provision. In addition, they are used differently with respect to the boundary conditions. Primary control power (PRL), regardless of the location of the fault, is to be delivered across Europe from all embedded sources, essentially in proportion to the actual frequency deviation. The absolute maximum power is to be provided at frequency deviations of minus 200 mHz and (absolute) below, the absolute minimum power is to be provided at frequency deviations of plus 200 mHz and above. With regard to dynamics, it is necessary to provide the (maximum) performance within 30 seconds from hibernation. In contrast, secondary control power (SRL) and minute reserve power (MR) are to be provided in the balancing rooms in which the fault has occurred. Their task is to compensate for the disturbance as quickly as possible and thus to ensure that the grid frequency is back within the desired range as quickly as possible, preferably at the latest after 15 minutes. In terms of dynamics, lower requirements are placed on SRLs and MRLs (5 and 15 minutes, respectively) full service provision after activation), at the same time these services must be provided for longer periods than primary control power.

In the power grids operated so far, a large part of the control power is provided by conventional power plants, in particular coal and nuclear power plants. Two fundamental problems result from this. On the one hand, the conventional power plants providing control power are not operated at full load and thus maximum efficiencies, but slightly below them in order to be able to provide positive control power if required, possibly over a theoretically unlimited period of time. On the other hand, with increasing expansion and preferential use of renewable energies, fewer and fewer conventional power plants are in operation, which is often the prerequisite for the provision of balancing services.

For this reason, approaches have been developed to increase the use of energy storage in order to store negative control power and to provide it as positive control power when needed.

The use of hydro pumped storage plants to provide control power is state of the art. In Europe, the various types of control are provided by pumped storage. However, hydropumps are also often referred to as the most economical renewable energy storage and retrieval technology today to better match energy supply and demand over time. The potential for expanding storage capacity - especially in Norway - is a matter of controversy, as significant capacity in power lines needs to be approved and installed for use. Consequently, the use for the energy management of load management is in competition with the provision of control power.

Against this background, in the area of primary control power in the recent past, approaches have repeatedly been explored and described for the use of other storage technologies, such as flywheel mass and battery storage, for the provision of control power. From US 2006/122738 A1 an energy management system is known, which comprises a power generator and an energy store, wherein the energy store can be charged by the power generator. This should be an energy producer, which does not guarantee a uniform power generation in normal operation, such as For example, the increasingly favored renewable energy sources, such as wind power or photovoltaic power plants, will be able to distribute their energy more evenly into the power grid. The disadvantage of this is that in this way a single power plant can be stabilized, but all other disturbances and fluctuations in the power network can not be intercepted or only to a very limited extent.

It is known from WO 2010 042 190 A2 and JP 2008 178 215 A to use energy storage to provide positive and negative control power. When the grid frequency leaves a range around the desired grid frequency, either energy is provided from the energy store or added to the energy store to regulate the grid frequency. DE 10 2008 046 747 A1 also proposes operating an energy store in an island power grid in such a way that the energy store is used to compensate for consumption peaks and consumption minima. The disadvantage hereof is that the energy stores do not have the necessary capacity to compensate for a longer disturbance or a plurality of disturbances rectified with respect to the frequency deviation one after the other.

In the article "Optimizing a Battery Energy Storage System for Primary Frequency Control" by Oudalov et al., In IEEE Transactions on Power Systems, Vol. 22, No. 3, August 2007, the capacity of a rechargeable battery is dependent on technical and operational It can be seen that, due to the injection and withdrawal losses, it is inevitable that the energy storage device will be recharged at longer intervals in the long term, but it can still happen temporarily or temporarily The article proposes a (limited use) of resistors, but this leads to the destruction of energy and, as a rule, the devaluation of the energy, despite this measure the energy storage is at a full load runtime of 1, 6 Hours still comparatively large, definitely s significantly larger than the minimum requirements of the UCTE Handbooks. The comparatively high capacities are associated with corresponding investment costs and make the use of storage often uneconomical. The price for the provision of control power depends largely on how quickly the control power can be provided after a request, that is, after a frequency deviation outside the tolerance. At PRL and SRL, the provision of energy is compensated. SRL and MR also pay for work.

In the area of secondary control power provision, the various requirements for the dynamics of service provision (commissioning and decommissioning) of the sources (or pools of sources) are explained below for the example of the European interconnected network of the UCTE. 1 . The prequalifiable secondary control power (short SR power, SRL or nominal power) results from the power change activated and measured within 5 minutes (each control direction)

2. A short-term overshoot of a maximum of 10% above the secondary control power setpoint is permitted. In any case, a short-term overshoot of up to 5 MW is permissible.

3. In the case of secondary control power pools, after 30 seconds at the latest, a reaction of the pool to the transmission system operator shall be measurable.

The compensation for the secondary control power provision consists of a performance fee for the provision of the secondary control power and a remuneration for the actual energy provided in the context of the secondary control power provision.

Such requirements, in particular for groups of energy producers, can be found in the Forum Netztechnik / Netzbetrieb in the VDE (FNN) "TransmissionCode 2007" of November 2009. For this, Annex D2 in particular is relevant for the requirements of SRL pools, in which it is also described which method a control system can be operated at a provider of SRL.

The problem with using power plants or consumers, such as electrolysis plants to provide control power, is that they can not be ramped up fast enough to provide the required speed for MR or SRL in an emergency.

Accumulators and other energy stores can absorb or release energy very quickly, making them basically suitable for providing PRL. The disadvantage, however, is that very large capacities of the batteries must be provided in order to deliver the power over a longer period or repeatedly. However, very large capacity batteries are also very expensive. Due to the losses during storage and withdrawal of energy takes place at a statistically symmetric deviation of the mains frequencies from the setpoint by the operation sooner or later emptying of the energy storage, such as a rechargeable battery. It is therefore necessary to recharge the energy storage more or less regularly targeted. It may be necessary to pay for this charging current separately.

From US Pat. No. 7,839,027 B2, a generic method for providing control power with an energy store is known. In this case, in the times when no control power is required, the energy storage is charged or discharged to achieve a desired output state of charge. The disadvantage of this is that the power must be obtained from the network, so it must be paid. Another disadvantage is that when many control cycles follow each other, the energy storage is still heavily charged or discharged. So there must still be kept a large capacity (energy storage capacity) of the energy storage.

A consistent adherence to the guidelines for the prequalification of primary control technologies requires the provision of adequate power reserves at any operating time and thus the state of charge of the energy storage. This requirement (currently in Germany: the marketed primary control power over a period of 15 minutes) means that a corresponding increase in investment costs must be provided. In fact, such a reserve would be used very rarely (statistically).

It has been found within the scope of the invention that sometimes considerable quantities of energy are monotonously fed in or out, as shown by an analysis of real frequency characteristics of the inventors. For a given storage capacity, this leads to a correspondingly high charge state change. In turn, large state of charge changes tend to cause faster aging than low state of charge changes. Either the memory thus reaches the end of its life rather than having to be replaced, or the capacity is a priori too increase to reduce the relative state of charge change. Both lead to an increase in investment costs.

The object of the invention is therefore to overcome the disadvantages of the prior art. In particular, a possibility should be found to improve the economic efficiency of the operation of energy storage devices, such as accumulators, in particular for the primary control power supply by avoiding inefficient charging states. In this case, the provision of primary control power with a lower energy storage capacity should be made possible. In addition, it would also be advantageous if a lower aging load could be achieved. Furthermore, the provision of primary control power while avoiding intermediate charging would also be desirable. Also, control power is to be provided with efficient energy yield of the control power supplier.

A further object of the invention is to be seen in that, in particular when using galvanic elements, such as accumulators, the capacity of the energy store should be as low as possible in order to provide the required control power.

The object of the invention is achieved in that the tolerance with respect to the frequency deviation is used to adjust the state of charge of the energy storage simultaneously with the provision of the control power through the energy storage.

It can be provided that the tolerance of the amount of the control power to be provided is used as a function of the frequency deviation to adjust the state of charge of the energy storage, preferably simultaneously with the provision of the control power by the energy storage.

The tolerance with regard to the amount of the control power provided and the tolerance in the determination of the frequency deviation is to be understood by the network operator, due to technical conditions such as the measurement accuracy in determining the control power supplied or the grid frequency, certain differences between an ideal target power and the actually delivered control power will be accepted. The tolerance may be granted by the network operator, but could also comply with a legal requirement. The advantage of these procedures is that during the provision of the control power and thus also continuously it is possible to regulate the state of charge (charge state) of the energy store. The terms state of charge and state of charge are to be regarded as equivalent according to the invention. As energy storage in particular LESRs (Limited Energy Storage Resources) come into question.

Furthermore, it can be provided that the frequency deviation is determined by the nominal value of the network frequency and the demand for positive and / or negative control power is determined from the frequency deviation. In the direct determination of the mains frequency or the frequency deviation own measuring equipment can be used. This provides the opportunity to determine the frequency deviation with a higher accuracy than prescribed by the network operator. This allows a greater frequency tolerance can be used to control the state of charge of the energy storage.

A particularly preferred embodiment of the invention provides that the frequency deviation is measured with greater accuracy than necessary for the provision of the control power, preferably with an accuracy of at least ± 8 mHz, more preferably of at least ± 4 mHz, most preferably of at least ± 2 mHz, especially preferred of at least ± 1 mHz.

The higher the accuracy in determining the frequency deviation of the grid frequency from the target value, the greater the margin that can be used to adjust the state of charge of the energy store. A further particularly advantageous embodiment of the method according to the invention results if it is provided that the amount of the requested control power is exceeded in order to utilize the tolerance with respect to the amount of the control power provided, preferably by a maximum of 30% and / or by 10 MW, particularly preferably by a maximum 20% and / or by 5 MW, wherein in particular the percentage exceeding of the amount of the requested control power proportional to the deviation of the state of charge of the energy storage is selected from a desired average state of charge. Exceeding the limit is more appropriate in the provision of control power as a shortfall that could lead to problems in the power grid. Therefore, the grid operator can grant a greater tolerance when exceeding the control power, as in the case of a shortfall. According to a further embodiment of the invention it can be provided that the amount of the control power provided is measured with a greater accuracy than necessary for the provision of the control power, preferably with an accuracy of at least ± 10 kW, more preferably of at least ± 1 kW, especially preferably of at least ± 100 W, more preferably 10 W, in particular with an accuracy of 1%, preferably 0.1% of the prequalified power.

Increasing the measuring accuracy also leads to a greater tolerance in the control power, which can be used to set the state of charge of the energy storage. According to a particularly preferred embodiment of the invention can be provided that the state of charge of the energy storage is adjusted by the fact that in the case in which the state of charge of the energy storage is above a first limit, the energy storage within the tolerance or tolerances larger control power in feeding in the power grid or receiving a smaller control power within the tolerance or tolerances from the power grid and / or in the event that the state of charge of the energy store is below a second threshold, the energy store has a smaller control power within the tolerance or tolerances feeds the power grid or receives a greater control power from within the grid within the tolerance or tolerances.

As a result, details are given as to how an inventive control of the state of charge can be achieved. These principles reflect the general principles of a method according to the invention, which solves the problems underlying the invention. It can be provided that the first limit of the state of charge between 40% and 80% of the maximum charge of the energy storage is, preferably between 45% and 70%, more preferably between 50% and 60%, and / or the second limit of the state of charge between 20% and 60% of the maximum load, Preferably, between 30% and 55%, more preferably between 40% and 50%, or the first and the second limit of the state of charge at 50% of the maximum charge of the energy storage are.

The ranges mentioned for the limit values of the state of charge of the energy store are particularly suitable according to the invention for the realization of methods according to the invention. Particularly with a large capacity of the energy store, the limit values can also be 80% or 20% of the maximum charge of the energy store. The percentage figures refer to the maximum energy content of the energy storage. Furthermore, it can be provided that the first and / or the second limit of the state of charge are selected depending on whether positive or negative control power is needed, so that when a request for positive control power, the first limit and / or the second limit of the state of charge is selected at a larger charge of the energy storage and is selected in a request for negative control power of the first limit and / or the second limit of the charge state at a smaller charge of the energy storage, the limit values are preferably selected as a function of the frequency deviation of the power frequency.

By these measures, the state of charge of the energy storage can be adjusted even better.

According to a further preferred embodiment, it can be provided that a flywheel, a heat accumulator, a hydrogen generator and storage with fuel cell, a natural gas generator with gas power plant, a pumped storage power plant, a compressed air storage power plant, a superconducting magnetic energy storage, a redox flow element and / or or a galvanic element is used, preferably an accumulator and / or a battery storage power plant, particularly preferably a lithium-ion accumulator.

These energy stores are suitable for implementing methods according to the invention. But the best way to do this is to use fast energy storage devices, ie energy storage devices that can provide the full control power within a short time. Therefore, accumulators are particularly preferred. The heat storage must be operated together with a device for producing electricity from the stored thermal energy. It can also be provided that an energy of at least 1 kWh can be stored in the energy store, preferably of at least 10 kWh, particularly preferably at least 50 kWh, very particularly preferably at least 250 kWh. The capacity of electrochemical energy storage can be at least 40 Ah, preferably about 100 Ah. The individual cells of electrochemical energy stores can operate at at least 1 V, preferably at at least 10 V, particularly preferably at least 100 V.

These capacities or these amounts of energy are particularly suitable for the provision of control power for a power grid. Smaller capacities or energy quantities are insufficient to influence the network frequency or to provide control power, at least in the long term. The large capacity compared to commercially available small energy storage devices, such as cell phone batteries and the high energy storage capacity are suitable for being able to provide the considerable energy for the method according to the invention, even in large power grids.

Furthermore, it can be provided that, in the case of a control power requirement requiring a maximum control power, a greater control power is provided than a contracted control power in order to set the state of charge of the energy storage.

Again, exceeding the provision of control power is more than a shortfall that could lead to problems in the power grid. Therefore, the grid operator can grant a greater tolerance when exceeding the control power, as in the case of a shortfall. This applies in particular to the maximum control power. A contracted service is the service that is marketed to the network operator in the current operation. The contracted capacity is at most as high as the prequalified service.

It can also be provided that the state of charge of the energy store is adjusted such that a medium state of charge is sought, preferably a state of charge between 20% and 80% of the maximum charge of the energy store, more preferably between 40% and 60%, most preferably a state of charge of 50% of the maximum charge of the energy store. The maximal Charge refers here again to the maximum energy storable in the energy storage.

Since it is usually not known whether a positive or a negative control power is needed at the next control power request, it makes sense to keep the state of charge in a middle range.

A further embodiment of the invention provides that the energy store is operated together with at least one energy generator and / or at least one energy consumer, which is used to adjust the state of charge of the energy store. Energy producers, such as power plants and energy consumers, such as electrolysis plants can take permanent control power loads. In addition, they can also be used to adjust the state of charge of the energy storage device without taking power from the grid or putting it into the power grid. According to the invention, it can also be provided that a power plant is used as the energy generator, preferably a coal-fired power station, gas-fired power plant or a hydroelectric power station and / or a plant for producing a substance is used as the energy consumer, in particular an electrolysis plant or a metal plant, preferably an aluminum plant or a steel plant. Such energy producers and consumers are well-suited to providing longer-term balancing services. Their inertia can be compensated according to the invention well with energy storage.

A particularly preferred embodiment of a method according to the invention is realized in that the time tolerance is used to adjust the state of charge of the energy storage, with a change in the required control power, the time after the change of the required control power, provided from the energy storage a modified control power is selected depending on the current state of charge of the energy storage. The time tolerance in the provision of control power stems from the fact that most sources of control power are clearly carriers as energy storage, especially as accumulators. Therefore, the suppliers of PRL, SRL and MR becomes time given to respond to a change in the required control power. This time can be used according to the invention to adjust the state of charge of the energy storage.

It can be provided that when a change in the required control power, a positive control power at an early stage, preferably immediately fed into the grid and / or a negative control power is taken at a late date, preferably at the latest possible time from the mains, if the state of charge of the energy store is above a first limit value and / or a change in the required control power a negative control power early, preferably immediately removed from the mains and / or fed a positive control power at a late date, preferably at the latest possible time in the power grid when the state of charge of the energy store is below a second limit. It can also be provided that the time is in a time interval between the change of the required control power and a maximum time after the change of the required control power, the time interval depending on the type of demanded control power, in particular the time interval in case of provision of primary control power 30 seconds, in case of provision of secondary control power 5 minutes and in case of provision of minute reserve 15 minutes.

A further embodiment of the invention, taking advantage of the time tolerance can provide that is adjusted by the choice of the timing of the state of charge of the energy storage. Since energy is power times, the time tolerance can be used to adjust the state of charge of the energy storage.

Furthermore, it can be provided that a control power gradient is selected as a function of the state of charge of the energy store, wherein in particular the time profile of the amount of the control power is set and the tolerance of the amount of the control power to be provided is used over time.

It can be provided that when changing the control power request, especially in a measured change in the frequency deviation, the dynamic behavior with which the energy storage on the change of Setpoint reacts, is selected as a function of the current state of charge of the energy storage, preferably the selected dynamic behavior depends on the type of demanded control power, in particular the time in which the full control power must be provided in the case of provision of primary control power 30 seconds, in the case a provision of secondary control power is 5 minutes and in the case of providing minute reserve 15 minutes.

A further embodiment of the invention, taking advantage of the dynamic tolerance, can provide for responding to the change in a control power request by suitably selecting the dynamic behavior of the energy store, thereby adapting the state of charge of the energy store.

A particularly preferred embodiment of a method according to the invention results if it is provided that a control power gradient is selected as a function of the state of charge of the energy store, wherein in particular the time profile of the amount of the control power is set and the tolerance of the amount of the control power to be provided is used over time.

It is used that, for example, by the network operator or by legal requirements, there are tolerances with respect to the increase or decrease of the standard power to be provided. These tolerances can be used according to the invention to adjust the state of charge of the energy storage. This results in a further improvement in the state of charge optimization, or this embodiment represents a particularly preferred implementation of the method according to the invention. As a result, the size or the capacity of the energy store can be further reduced. Furthermore, it may be particularly preferred that when the frequency deviation changes by less than a dead band, in particular by less than 10 mHz, a modified control power is only provided to adjust the state of charge of the energy store, in particular the changed control power is only provided if As a result, the state of charge of the energy store as much as possible to the middle state of charge or as little as possible from the average state of charge invites or discharges.

The change of the frequency deviation here refers to the frequency deviation at which the last control power adjustment took place. in this connection another tolerance is used. This tolerance may be given by the network operator or the legislator or system immanent. The tolerance allows that if only small changes in the frequency deviation of the mains frequency of the desired frequency occur, then no adjustment of the control power must be made. In order to develop the state of charge of the energy storage in the desired direction, however, control power can be provided. This control power delivery is optionally carried out in the way it is desired, so for example directly antiproportional to the frequency deviation. Thus, if control power is provided to adjust the state of charge, this would contribute to regulating the grid frequency. The adaptation of the state of charge is thus always constructive in the sense of regulating the grid frequency. The prerequisite for this is that the frequency deviation can be measured with a higher accuracy than the dead band. The dead band is a measure of the tolerance of the change in frequency deviation above which a change in the control power is required or necessary.

The objects of the invention are also achieved by a device for carrying out such a method, wherein the device comprises a controller and an energy store, wherein the device is connected or connectable to a power supply, the controller is connected to the energy store and the output from the energy storage and / or recorded control power controls.

Under a control according to the invention is understood in the present case a simple control. It should be noted that each control comprises a control, as in a control, a control in dependence on a difference of an actual value to a desired value takes place. Preferably, the controller is thus designed as a control, in particular with respect to the state of charge. Particularly preferably, the controller is a control system.

It can be provided that the device comprises a device for measuring a frequency deviation from a nominal value of the mains frequency of the power network, a device for measuring the injected and recorded power and a memory, wherein in the memory to be provided, dependent on the frequency deviation control power together is stored with the tolerances and preferably at least one state of charge of the Energy storage is stored as a target value or as a limit or are stored as two states of charge as limits, the controller has access to the memory and is designed, depending on the frequency deviation and the state of charge of the energy storage, the output from the energy storage and / or absorbed energy regulate.

Finally, it can also be provided with regard to methods according to the invention that the energy store is an accumulator, preferably a lithium-ion accumulator, a lead-sulfuric acid accumulator, a nickel-cadmium accumulator, a sodium sulfide accumulator and / or a Li-ion accumulator. Accumulator and / or a composite of at least two of these accumulators.

These accumulators are particularly suitable for the implementation of inventive methods due to their rapid reaction.

Accumulators include in particular lead-acid batteries, sodium-nickel-chloride accumulators, sodium-sulfur accumulators, nickel-iron accumulators, nickel-cadmium accumulators, nickel-metal hydride accumulators.

Accumulators, nickel-hydrogen batteries, nickel-zinc batteries, tin-sulfur lithium-ion batteries, sodium ion batteries and potassium ion batteries.

Here, accumulators are preferred, which have a high efficiency and a high operational and calendar life. As a particularly preferred embodiment of the invention, lithium polymer accumulators, lithium titanate accumulators, lithium manganese accumulators, lithium iron phosphate accumulators, lithium iron manganese phosphate accumulators, lithium iron yttrium Phosphate accumulators, lithium-air accumulators, lithium-sulfur accumulators and / or tin-sulfur lithium-ion accumulators used as lithium-ion accumulators.

The control energy or control power is delivered to the power grid (positive control energy or positive control power) or taken from the mains (negative control energy or negative control power). All mentioned control cycles can be carried out according to the invention and particularly preferably also any number of times in succession, by the energy storage is repeatedly recharged or discharged during a control cycle to again to have a suitable average state of charge for a second cycle.

The inventive method also ensures that the wishes of the customer, that is, the grid operator can be met for a predictable and defined control power and no regulatory oscillations are generated in the power grid.

The rated power of the device for providing control power is the power achievable within a certain time. It is also spoken of the praqualifiable performance, since this meets the criteria of the customer, that is the network operator.

The invention is based on the surprising finding that energy stores can be operated in such a way that the tolerances given by the grid operators when providing control power are used to maintain the state of charge of the energy store in a central area and thus the capacity of the energy store in a cost-effective framework to keep. This is especially true for energy storage, the energy quickly, that is record high power gradients and can deliver, such as flywheels and especially accumulators and battery storage. The tolerances granted by the network operators stem from the fact that the control power is currently mostly provided by larger power plants, such as coal power plants and larger consumers, such as electrolysis plants. These control power suppliers have a comparatively large inertia in the reception and output of control power, that is, their power gradients are small and effects such as overshoot are inevitable to some extent. Therefore, certain tolerances are given in the provision of control power. For example, control power suppliers do not have to deliver control power even with very small deviations, but a tolerance is granted in the frequency deviation of the power frequency, from which control power must be made available. Thus, it can be ensured that the provision of control power only serves to stabilize the network frequency. Above all, a tolerance in the measurement of the frequency deviation of the mains frequency is granted, so that the control power to be provided within a certain scope (in a tolerance) can be chosen freely. The latter tolerance is used in the invention to adjust the state of charge of the energy storage in which more or less energy is stored in the energy storage or more or less energy is released from the energy storage to the power grid. Another tolerance is also the accuracy of what power can be fed into the grid or absorbed from the mains. It is usually not so important for the network operator whether the required control power is provided or whether it is exceeded by a certain amount or slightly below.

These tolerances can be used to provide more or less control power and thus to charge or discharge the energy storage targeted stronger or less strong. As a result, the state of charge can be specifically influenced. Thus, according to the invention, it is possible to maintain the state of charge in a desired, for example, in a central region, so that the energy store is in a good or even optimum initial state for the next or a changed control power request.

The state of charge then no longer has to or can be regulated with less energy from the mains. In addition, the capacity of the energy storage can be chosen smaller. Both save costs. The state of charge in the case of accumulators as energy storage corresponds to the state of charge (English: "State of charge", SOC) or the energy content (English: State of Energy, SoE).

Battery storage devices (accumulators) distinguish themselves from conventional technologies for providing primary and / or secondary control functions, inter alia, in that they can change the power provided much faster. In most cases, however, is disadvantageous in battery storage that they have a relatively small storage capacity, so can provide the required services only over a limited period. In a statistical evaluation of the frequency deviations over time, it has surprisingly been found in the context of the present invention that the requested powers in more than 75% of the active time (that is to say a power deviating from zero) are less than 20% of the maximum power. Performance or marketed performance.

From this realization of the invention also follows that the capacity of the energy storage and thus the stored stored Regulating power amount can be chosen lower and it can be particularly well with a method according to the invention to keep the capacity of the energy storage small.

It is used in the invention that it is possible with modern measuring devices and measuring methods to determine the frequency deviation more accurate than is currently required for the provision of control power. This makes it possible to freely select the control power to be provided within the specified tolerances (ie within the specified limits) and thereby realize the method. Specific and particularly preferred embodiments of the approaches are that the energy store is an accumulator or battery store that is used to provide primary control power.

In a further specific embodiment, the energy taken into the memory in the event of negative PR or SR power can be sold on the spot market, in particular if the conditions there are advantageous.

According to a further embodiment, it may be provided within the scope of the specifications for the provision of control power that, in particular, more energy is absorbed from the network than is fed in by the energy store. This can be done by providing very much negative control power in accordance with the regulations including the procedure outlined above, whereas according to the regulations including the procedure set out above, preferably only the at least guaranteed power is provided at positive control power. Preferably, an average of at least 0.1% more energy is withdrawn from the network than is supplied, in particular at least 0.2%, preferably at least 0.5%, more preferably at least 1.0%, especially preferably 5%, these values being based on a Average measured over a period of at least 15 minutes, preferably at least 4 hours, more preferably at least 24 hours and especially preferably at least 7 days, and refer to the energy fed. In this case, the control power provision set out above can be used to extract a maximum of energy from the network, whereby the maximum possible negative control power is provided, whereas only a minimum of positive control power is provided. In the embodiments for the preferred and especially for the maximum energy consumption, the energy thus extracted from the network can be sold via the energy trade described above, preferably at times when the highest possible price is to be achieved. For this purpose, forecasts of the price development based on historical data can be used.

Furthermore, the state of charge of the energy store at the time of a planned sale of energy may preferably be at least 70%, particularly preferably at least 80% and particularly preferably at least 90% of the storage capacity, the state of charge after sale preferably not exceeding 80%, in particular not more than 70% and especially preferably at most 60% of the storage capacity amounts.

In preferred embodiments of the invention, a plurality of energy stores are pooled and operated in accordance with the method of the invention. The size of the energy storage within the pool can vary. In a particularly preferred embodiment, in the case of the use of tolerances, in the case of the use of tolerances, the change from one parameter setting to another is not synchronized, but deliberately offset in time, in order to minimize or at least tolerate any disturbances in the network.

In a further preferred embodiment, the tolerances used in the various procedures vary depending on the time of day, the day of the week or the season. For example, within a period of 5 minutes before to 5 minutes after the hour change, the tolerances can be more narrowly defined. This is due to the fact that often very rapid frequency changes take place here. It may be in the interest of transmission system operators that there are lower tolerances and thus the control energy supply is more secure in the sense of sharper.

Embodiments of the invention will be explained below with reference to three schematically illustrated figures, without, however, limiting the invention. Showing:

Figure 1 is a schematic diagram of the normalized power P / Psoii on the frequency deviation Af to provide control power; FIG. 2 shows a flow chart for a method according to the invention; and

Figure 3: a schematic representation of an inventive device for providing control power.

Figure 1 shows an exemplary schematic diagram of the normalized power P / Psoii over the frequency deviation Af for providing control power as required by network operators. In the European power grid, a grid frequency of 50,000 Hz is set, the tolerance in the measurement of the frequency deviation is currently ± 10 mHz. The mean solid curve shows the control power P desired by the grid operator as dimensionless power P / Psoii normalized to the prequalified or contracted power P _SO II. For a value of 1, therefore, the power P _SO II is achieved as a positive control power and at a value of - 1 the power -P _SO II as a negative control power. In the present case, the power P _SO II is to be provided from frequency deviations of 200 mHz from the desired power frequency 50,000 Hz. In order to ensure that the provision of control power only serves to stabilize the grid frequency and is not counterproductive, no control power is requested in a dead band of ± 10 mHz by the desired grid frequency of 50,000 Hz. The deadband is herein interpreted as a tolerance in the provision of control power. At the same time, a tolerance and / or a dead band of ± 10 mHz is granted in the measurement of the frequency deviation Af.

It may also be (unlike in Figure 1 shown) that from frequency deviations of +10 mHz or -10 mHz, the control power is to be increased leaps and bounds. Then in the deadband no control power would be provided and when working outside the deadband is a control power that is proportional to the frequency deviation.

In addition, the negative and positive control power provided, including the control power to be provided from a frequency deviation of 200 mHz, may be exceeded by up to 20%, but not undershot. This tolerance is shown in the diagram of Figure 1 by the dotted line. Thereby, the control power P can be provided in a tolerance around the solid curve, which is bounded by the two dashed curves, or even bounded by the dashed and dotted curves. The dashed and Dotted lines (curves) represent the maximum and minimum control power that can be fed into the grid or taken out of the grid using these two tolerances for a given frequency deviation. It should be noted that under real conditions a distance to the dashed and / or dotted curves is to be observed, which is limited by the own measuring accuracy. This prevents the actually provided control power from being above or below the tolerance. The higher the accuracy of the own measurement of the frequency deviation Δf, the more dense the control power can be provided at the dashed or dotted lines and the better the tolerance between these curves can be exploited.

In extreme cases (with arbitrarily accurate frequency measurement), it is therefore possible to run an energy store along the dashed or dotted lines in order to keep the charge state of the energy store in a desired middle range. In particular, when the state of charge is already in the middle range or just in case of a request for positive control power or just below a request for negative control power, it may also be useful to drive the energy storage along the solid line to control power such desired to provide. With a slight deviation, it may also be useful to drive in an intermediate region between the solid and the dashed or dotted lines.

It may also be given a time tolerance, that is, for example, only a few seconds after a frequency deviation Af has occurred, which requires feeding or receiving control power, an actual provision or adjustment of the control power must be made. This time tolerance can be used according to the invention and according to a particularly preferred embodiment of a method according to the invention to adjust the state of charge of the energy storage, which is used to provide control power.

When the energy storage is relatively heavily charged, that is, the state of charge of the energy storage in the upper range, so for example between 75% and Is 90% of the maximum charge of the energy storage, the energy storage is driven near the upper right dashed or dotted curve. If a request for positive control power is detected by a negative frequency deviation Δf, then a positive control power is provided very early and also a higher control power than is actually required by the network operator and as indicated by the solid middle line. This results in the energy storage being discharged more than if it were operated according to the solid middle line. When a request for negative control power, which is determined by a positive frequency deviation Af, on the other hand responded to the request as late as possible and then only with the least energy intake at lower power. As a result, the energy storage device absorbs less energy than if it were operated in accordance with the solid middle line. In both cases, it is thus achieved by the inventive method that the state of charge of the energy storage is closer to the desired average state of charge after a control cycle, as if it would be operated as conventional along the solid line.

If the energy store is charged relatively low, that is, the state of charge of the energy storage is in the lower range, so for example between 10% and 25% of the maximum charge of the energy storage, the energy storage is driven in the vicinity of the lower left dashed or dotted curve , If a request for positive control power is detected by a negative frequency deviation Δf, then a positive control power is provided very late and also a control power that tends to be lower than actually required by the network operator and as indicated by the solid middle line. This results in the energy storage being discharged less than if it were operated according to the solid middle line. When a request for negative control power, which is determined by a positive frequency deviation Af, however, is responded to as early as possible to the request and then with a higher energy intake at higher benefits. As a result, the energy storage device absorbs more energy than if operated according to the solid middle line. Even in these two cases is thus achieved by the inventive method that the state of charge of the energy storage is closer to the desired average state of charge after a control cycle, as if it would be operated as conventional along the solid line.

On average, when using a method according to the invention, a smaller capacity of the energy store must be kept available than with a conventional method. As energy storage devices with larger capacities are more expensive than those with lower capacity, this can save costs and consume fewer resources during construction. Alternatively, a larger control power can be provided with a same sized energy storage, if the power electronics is adapted or sufficient.

Particularly suitable for inventive method galvanic elements, in particular accumulators, most preferably lithium-ion batteries, since they can react very quickly and therefore can take advantage of the given tolerances particularly well. Slower energy storage, such as hydrogen storage and fuel cells or gas power plants with natural gas generators and natural gas storage takes a certain amount of time to provide the power can. Due to this inertia, the tolerances, in particular the time tolerances, can not be exploited so well. In addition, the capacity of these energy storage is not as expensive compared to other components as with accumulators or flywheels, so that the benefit of using a method according to the invention is not so great. FIG. 2 shows a flow chart for a method according to the invention. The method uses an energy store and a controller. Decision step 1 tests whether the state of charge of the energy store is within a desired range or above or below it.

If the state of charge is within the desired range, go to step 4. If the state of charge is below the desired range or the tolerance, as much energy as possible should be supplied to the energy store, or at least only little further energy should be withdrawn, and step 2 is continued. Is the charge state above the desired Range, energy should be removed from the energy storage as possible or at least only a little more energy supplied and it is continued with step 6.

In steps 2, 4 and 6, it is checked how large the frequency deviation Af of the mains frequency of the power grid is and then proceeded with steps 3, 5 and 7 respectively. If the energy store is in the desired state of charge, in step 5, which follows step 4, the control power P provided from the energy store is set as a function of and as a function of the frequency deviation Δf measured in step 4 and thus made available to the power grid. If energy should be supplied to the energy store as far as possible or at least only little further energy should be withdrawn, this is achieved in step 3 by adjusting the control power P provided from the energy store not only as a function of and as a function of the frequency deviation Δf measured in step 2, and thus the Power grid is provided. Instead, the state of charge of the energy store in the amount of the provided control power P is taken into account. Since the state of charge is below the desired range, more energy is supplied to the energy storage device or less further energy is withdrawn.

For this purpose, the tolerances are used in the provision of balancing power, which allows one of the grid operators. The provided positive or negative control power thus always moves within the limits shown in the embodiment of Figure 1. How far this tolerance is exploited can depend on various factors, including the time of day, the season, the day of the week, the current price on the electricity market or other events to be expected. In particular, according to the invention, there may also be a functional relationship between the state of charge of the energy store and the deviation from the desired control power provided or received.

Conversely, in step 7, when the state of charge is above the desired range, the energy store should be deprived of energy as far as possible or at least only a little more energy should be supplied. This is realized in step 7 in that the control power P provided from the energy store is set not only as a function of and as a function of the frequency deviation Δf measured in step 2, and thus made available to the power grid is considered, but also the state of charge of the energy storage in the amount of the provided control power P. Since the state of charge is above the desired range, less energy is supplied to the energy store or additional energy is withdrawn. In turn, the tolerances in the provision of balancing power, which one of the grid operators admits, are used. The provided positive or negative control power thus always moves within the limits shown in Figure 1 as an example in this case. How far this desired area is exploited can in turn depend on the factors mentioned, such as the time of day, the season, the day of the week, the current price on the electricity market or other expected events.

In particular, according to the invention, there may also be a functional relationship between the state of charge of the energy store and the deviation from the desired control power provided or received. As a function, for example, a linear relationship between the frequency deviation Af and the control power P to be provided can be selected. An additional summand S, in which the other factors such as time of day, season, the day of the week, the current price on the electricity market or other expected events but also especially the state of charge L of the energy storage can be considered, can be additionally introduced. An example of such a functional relationship with reference to FIG. 1 is for frequency deviations Af in a range of ± 200 mHz:

P / Psoii = (Af + S) / 200 mHz or P / P _SO II = Af / 200 mHz + S

For the first function S has the dimension of a frequency and ranges between +0.2 ^* Af + 10mHz and - 10mHz for positive control power and between - 0.2 ^* Af - 10mHz and + 10mHz for negative control power. For the second function, the size S is not a dimension and reflects the tolerance for the amount of power. The function stored for S must in both cases be within the given tolerance limits. Outside this range ± 200 mHz, P / Psoii = 1 is selected or, as in FIG. 1, with the tolerance for the maximum control power: P / Psoii = 1 + L1 0.2 for a positive control power requirement or P / Psoii = -1 - L2 at 0.2 a negative control power requirement, wherein the factors L1 and L2 can assume functions of the state of charge of the energy store and values of 0 to 1. In this case, for example, a mean state of charge L = 0.5 (50%) can be assigned the value L1 = L2 = 0, or L1 = 1, if the state of charge is 75% (L = 0.75) or above and L1 = 0 when the charge level is 55% (L = 0.55) or below. Similarly, for L2: L2 = 1 when the state of charge is 25% (L = 0.25) or less and L2 = 0 when the state of charge is 45% (L = 0.45) or above. In between, a linear relationship between L1 or L2 and the state of charge L of the energy store can be assumed. For example, for the range of the state of charge L between 0.55 and 0.75, ie from 55% to 75%, the function L1 = (L - 0.55) ^* 5 be selected, or at negative control power requirement and highly discharged energy storage for the area of the state of charge L between 0.25 and 0.45, ie from 25% to 45%, the function L2 = (L - 0.25) ^* 5. This is based on the assumption that the nominal power P _SO II is up to 20% may be exceeded.

Similarly, the summand S depending on the state of charge of the energy storage can be selected or depend exclusively on the state of charge L of the energy storage. The dependence on the state of charge L can represent a curve of any kind that moves within the limits specified by the tolerances. Also a linear relationship can be used within the limits of tolerance and in the limits ± 200 mHz. It may also be advantageous if the limits of the tolerance are not fully exploited if the measurement accuracy of the network frequency is not accurate enough. The distance from the limits of the tolerance must at least correspond to the distance that corresponds to the uncertainty of the own measurements. Preferably, S is proportional to the charge state deviation of 50% charge (L = 0.5).

The present functions are exemplary embodiments of how, in steps 3, 5 or 7, the state of charge of the energy store can be adapted according to the invention.

After steps 3, 5 and 7, the decision step 1 is continued again and, in principle, can also be run endlessly. 3 shows a schematic view of a device 21 according to the invention comprising an energy store 22. A controller 23 is connected to the energy store 22 so that the controller 23 can adjust the power consumption and output of the energy store 22. The energy storage 22 is connected to a power grid 24 and can absorb power from the power grid 24 and / or deliver. The controller 23 is connected to a device 25 for measuring the frequency deviation of the mains frequency of the power network 24. When there is a need for control power - positive or negative control power - the controller 23 generates a signal. Subsequently, the control power of the energy storage 22 is increased. The device 25 is capable of more accurately determining the frequency deviation of the line frequency than required. Thus, the tolerance in the determination of the network frequency can be used to adjust the state of charge of the energy storage device 22. The state of charge of the energy store 22 can be determined by a suitable measuring device by the controller 23 in order to implement a method according to the invention.

The controller 23 can thus intelligently charge or discharge the energy store 22, so that a certain desired state of charge is sought. Thus tolerances in the frequency deviation, from which a control power is to be provided, and tolerances in the amount to be provided at a certain frequency deviation control power, or tolerances in the determination of the frequency deviation, as well as tolerances in the maximum to be provided prequalified or contracted control power used be developed to develop the state of charge of the energy storage device 22 in the desired direction. For example, the power of the energy storage 22 may be provided even at a lower frequency deviation than a frequency deviation of 10 MHz to charge or discharge the energy storage 22 when deemed necessary. Likewise, an over-fulfillment, for example by an overshoot, of up to 20% beyond the maximum control power can be generated by the energy store 22 in order to regulate the charge state of the energy store 22. In particular, in such cases, a particularly fast-reacting and easy to charge and discharge energy storage 22 is particularly advantageous. Batteries are best suited for this purpose. In particular, Li-ion batteries are quickly and frequently charged and discharged without damaging influences on the battery, so that they are particularly suitable and preferred for all embodiments according to the invention. This requires the provision of Li-ion accumulators of considerable capacity. For example, these are easily accommodated in one or more 40-foot ISO containers.

A device 21 according to the invention is therefore particularly well suited as a primary control power source or as a secondary control power source.

Please refer to the Forum Netztechnik / Netzbetrieb in the VDE (FNN) "TransmissionCode 2007" of November 2009 for details on the regulation of control power and the exchange of information with the network operators.

The features of the invention disclosed in the foregoing description, as well as the claims, figures and embodiments may be essential both individually and in any combination for the realization of the invention in its various embodiments.

LIST OF REFERENCE NUMBERS

P performance

Psoii Prequalified power (rated power) or contracted power f Frequency

Δί frequency deviation

1 decision step

2; 3; 4; 5; 6; 7 process step

21 Apparatus for providing control power

22 energy storage

23 control

24 power grid

25 Device for determining the frequency deviation

Claims

claims

1 . Method for providing control power for a power grid (24) in which at least one energy store (22) connected to the power grid (24) supplies power to the power grid (24) as needed and / or receives power from the power grid (24) as needed; the control power is provided as a function of a frequency deviation from a nominal value of a network frequency, characterized in that

the tolerance with respect to the frequency deviation is used to adjust the state of charge of the energy store (22) simultaneously with the provision of the control power by the energy store (22).

2. The method according to claim 1, characterized in that

the tolerance of the amount of the control power to be provided is used as a function of the frequency deviation in order to adjust the charge state of the energy store (22), preferably simultaneously with the provision of the control power by the energy store (22).

3. The method according to claim 1 or 2, characterized in that

the frequency deviation is determined by the nominal value of the mains frequency and the demand for positive and / or negative control power is determined from the frequency deviation.

4. The method according to any one of the preceding claims, characterized in that

the frequency deviation is measured with greater accuracy than necessary for the provision of the control power, preferably with an accuracy of at least ± 8 mHz, more preferably at least ± 4 mHz, most preferably at least ± 2 mHz, even more preferably at least ± 1 mHz ,

5. The method according to any one of the preceding claims, characterized in that the charge state of the energy store (22) is set by the fact that in the case in which the charge state of the energy store (22) is above a first limit, the energy store (22) has a larger control power within the tolerance or tolerances in the power grid ( 24) or receives within the tolerance or tolerances lying smaller control power from the power grid (24) and / or for the case in which the state of charge of the energy storage (22) is below a second threshold, the energy storage (22) within a the smaller tolerances or tolerances in the power grid (24) feeds or receives within the tolerance or tolerances larger control power from the power grid (24).

Method according to claim 5, characterized in that

the first limit of the state of charge is between 40% and 80% of the maximum charge of the energy store (22), preferably between 45% and 70%, more preferably between 50% and 60%, and / or the second limit state of charge between 20% and 60% of the maximum charge, preferably between 30% and 55%, more preferably between 40% and 50%, or the first and the second limit of the state of charge at 50% of the maximum charge of the energy store (22).

A method according to claim 5 or 6, characterized in that

the first and / or the second limit of the state of charge depending on whether positive or negative control power is required, so that when a request for positive control power, the first limit and / or the second limit of the state of charge at a larger charge of the energy store ( 22) is selected and in a request for negative control power of the first limit value and / or the second limit of the state of charge is selected at a smaller charge of the energy store (22), wherein the limit values are preferably selected as a function of the frequency deviation Af of the mains frequency.

8. The method according to any one of the preceding claims, characterized in that

a flywheel, a heat accumulator, a hydrogen generator and storage with fuel cell, a natural gas generator with gas power plant, a pumped storage power plant, a compressed air storage power plant, a superconducting magnetic energy storage, a redox flow element and / or a galvanic element is used as energy storage (22) preferably an accumulator and / or a battery storage power plant, particularly preferably a lithium-ion accumulator.

Method according to one of the preceding claims, characterized in that

in the energy store (22) an energy of at least 1 kW h can be stored, preferably of at least 10 kW h, more preferably at least 50 kW h, most preferably at least 250 kW h.

Method according to one of the preceding claims, characterized in that

for a control power request requiring a maximum control power, a larger control power is provided than a contracted control power to set the state of charge of the energy storage (22).

1 1. Method according to one of the preceding claims, characterized in that

the state of charge of the energy store (22) is set such that an average state of charge is sought, preferably a state of charge between 20% and 80% of the maximum charge of the energy store (22), more preferably between 40% and 60%, most preferably a state of charge of 50% of the maximum charge of the energy store (22).

12. The method according to any one of the preceding claims, characterized in that the energy store (22) is operated in conjunction with a power generator and / or an energy consumer that is used to adjust the charge state of the energy store (22).

Method according to one of the preceding claims, characterized in that

The time tolerance is used to adjust the state of charge of the energy storage device (22), with a change in the required control power, the time after the change of the required control power, from which the energy storage (22) a modified control power is provided, depending on the current Charging state of the energy store (22) is selected.

A method according to claim 13, characterized in that

when a change in the required control power a positive control power at an early stage, preferably immediately in the power grid (24) is fed and / or a negative control power at a late time, preferably at the latest possible time from the power grid (24) is removed, if the charge state of the energy store is above a first limit value and / or a change in the required control power a negative control power early, preferably immediately removed from the power grid (24) and / or a positive control power at a late time, preferably at the latest possible time in the power grid (24) is fed when the state of charge of the energy storage is below a second threshold.

Method according to one of the preceding claims, characterized in that

a control power gradient is selected as a function of the state of charge of the energy store (22), wherein in particular the time profile of the amount of the control power is set and the tolerance of the amount of the control power to be provided over time is used.

16. The method according to any one of the preceding claims, characterized in that

when the frequency deviation changes by less than a dead band, in particular by less than 10 mHz, a modified control power is only provided to determine the state of charge of the battery

Energy storage (22) adjust, in particular, the changed control power is provided only when the state of charge of the energy storage (22) as much as possible to the average state of charge or as little as possible from the average state of charge invites or discharges.

17. An apparatus for carrying out a method according to one of the preceding claims, characterized in that

the device (21) comprises a controller (23) and an energy store (22), wherein the device is connected or connectable to a power network (24), the controller (23) is connected to the energy store (22) and the energy store ( 22) delivered and / or recorded control energy controls.

18. The apparatus according to claim 17, characterized in that

the device (21) comprises a device for measuring a frequency deviation from a nominal value of the mains frequency of the power network (24) and a memory, wherein in the memory to be provided, dependent on the frequency deviation control power is stored together with the tolerances and preferably at least one Charge state of the energy store (22) is stored as a target value or as a limit or are stored as two states of charge as limits, the controller (23) has access to the memory and is designed in dependence on the frequency deviation and the state of charge of the energy store (22 ) to control the output from the energy storage (22) and / or recorded control power.

19. Device according to one of claims 17 or 18, characterized in that the energy store (22) is an accumulator, preferably a lithium-ion accumulator, a lead-sulfuric acid accumulator, a nickel-cadmium accumulator, a sodium sulfide accumulator and / or a Li-ion accumulator and / or a composite at least two of these accumulators is.