Method and device for the peak shaving of a power draw of a load via a grid connection point of an electrical grid

Abstract

The invention relates to a method and a device for the peak shaving of a power draw of a load (1) via a grid connection point (2) of an electrical grid (3), the method comprising: capturing a time curve of a power draw of the load (1) occurring via the grid connection point (2); ascertaining a time curve of an average value of the power draw from the captured time curve of the power draw; determining a time curve of electrical compensation power to be drawn from an energy store (5) associated with the load (1), on the basis of a determined threshold value (T) for the power draw and both the time curve of the power draw and the time curve of the average value; and causing at least partial compensation of the power draw by drawing the electrical compensation power from the energy store (5), while bypassing the grid connection point (2), according to the determined time curve of said electrical compensation power.

Description

[0001] 120639P1488PC

[0002] Method and device for peak load shaving of a consumer's power consumption via a grid connection point of an electricity network

[0003] The present invention relates to a method and a device for peak load shaving of a consumer's power consumption via a grid connection point of a power grid.

[0004] The term “electricity grid” or “grid” as used herein refers to a network that serves to supply electrical energy to consumers connected to the grid and connects power plants and other energy producers or energy converters, for example wind power or photovoltaic plants, with consumers to enable their electrical energy supply from the grid.

[0005] A grid connection point (GSP) is the physical connection point between the electricity grid and the grid user. The grid user can be an energy producer feeding electricity into the grid, an energy consumer drawing electricity from the grid, or a combination of both. The costs for the grid infrastructure up to the GSP are borne by the grid operators. In the context of photovoltaics (PV), a GSP is understood to be the specific location where a PV system is connected to the electricity grid. Via the GSP, the grid user, such as a building, can draw energy from the grid as an energy consumer and / or feed energy into the grid as an energy supplier.A typical example of a network participant is a building which, in order to meet its energy needs, obtains at least part of the required energy from the grid via its NAP, but also has its own energy generation plant, in particular a photovoltaic system (“PV system”), and feeds at least part of the energy generated there into the grid via the NAP.

[0006] In this context, a "consumer" is defined as a network participant in an electricity grid who is connected to the grid via a grid connection point and can draw electrical energy from the grid via this connection point and potentially also feed it back into the grid. A consumer can therefore be either a pure electricity consumer or an electricity consumer and supplier, meaning someone who both draws electrical energy from the grid and feeds their own electrical energy into the grid. Examples of consumers include households that draw electricity from the grid, businesses that draw electricity from the grid and may also operate their own power generation facilities, and operators of photovoltaic or wind power plants that feed electricity into the grid and also draw electricity from the grid.

[0007] For consumers (electricity customers) with higher electricity demand (with load profile meters), their electricity bill typically includes a first component, which relates to the energy consumed during the billing period (usually specified in kWh), and a second component, which prices the maximum power consumed (usually specified in kW). The latter is often determined by multiplying the highest electricity consumption averaged over a defined time interval (e.g., 15 or 30 minutes) during the billing period by a capacity charge. The higher the peak load, the higher the energy costs.

[0008] The synonymous terms "peak shaving" and "peak cutting", as used herein, refer to the use of at least one battery storage system to reduce the consumption of electrical energy from the grid or to reduce the maximum power consumption via the grid connection point.

[0009] It is known that the aforementioned peak loads can be reduced through peak shaving by using a battery storage system. To optimize the grid connection fee, a threshold is defined as a target value for the upper limit of the average load drawn from the grid connection point per time interval. With typical algorithms ("classic peak shaving"), the battery storage system begins to discharge when the power drawn from the grid connection point exceeds the predefined threshold. The change in the battery capacity of the 120639P1488PC

[0010] The battery storage capacity then corresponds to the difference between the (higher) current power consumption (grid draw) and the (lower) threshold value. If the grid draw is below the threshold value, the battery storage is charged. The change in battery capacity again corresponds to the difference between grid draw and threshold value and therefore has the opposite sign. The battery capacity is calculated at the shortest possible intervals, e.g., every second.

[0011] If, during battery discharge, the difference between grid consumption and the threshold is greater than the maximum battery capacity, such a load peak cannot be adequately compensated, and the average grid consumption may exceed the threshold. Furthermore, with conventional peak shaving, it can happen that individual short (e.g., 30-second) power peaks are compensated by the battery storage, even if the average over the corresponding time interval is below the threshold.

[0012] One object of the present invention is therefore to provide an improved peak shaving system that enables optimized, at least partial, compensation of peak loads.

[0013] To solve this problem, the respective devices or methods are proposed according to the teachings of the independent claims. Various embodiments and further developments of the solution are the subject of the dependent claims.

[0014] terms

[0015] Some of the other terms used herein to describe the present solution are explained in more detail below:

[0016] The term "mean," as used herein, refers to a value, in particular a number or quantified quantity, that is calculated from given values ​​according to a specific formula. Formulas are commonly used for the arithmetic, geometric, and quadratic means. The mean can, in particular, be a parameter that characterizes a typical value of a distribution or expresses the central tendency of a distribution.

[0017] Any terms used herein, such as "comprises," "includes," "includes," "features," "has," "with," or any other variant thereof, are intended to cover non-exclusive inclusion. For example, a method or apparatus that includes or features a list of elements is not necessarily limited to those elements but may include other elements not expressly listed or inherent in such method or apparatus.

[0018] Furthermore, unless explicitly stated otherwise, "or" refers to an inclusive "or" and not an exclusive "or". For example, a condition A or B is satisfied by one of the following: A is true (or present) and B is false (or not present), A is false (or not present) and B is true (or present), and both A and B are true (or present).

[0019] The terms "ein" or "eine," as used here, are defined as "one or more." The terms "ein anderer" and "ein Weitere," as well as any other variant thereof, are to be understood as "at least one more."

[0020] The term "plural," as used herein where applicable, is to be understood as "two or more." The terms "first," "second," "third," and similar terms in the description and in the claims are used to distinguish between similar or otherwise identically named elements and not necessarily to describe a sequential, spatial, or chronological order. It is understood that the terms used in this way are interchangeable under suitable circumstances and that the embodiments of the solution described herein may also function in orders other than those described or illustrated here. 120639P1488PC

[0021] The terms "configured" or "set up" to perform a specific function (and any variations thereof), as used herein, mean that a device or component thereof already exists in a configuration or setting capable of performing the function and / or that it is adjustable—i.e., capable of assuming a selectable configuration within a defined configuration space—so that it can perform the function after appropriate adjustment. Configuration can be achieved, for example, by adjusting parameters, such as a process sequence, or by using switches or similar devices to activate or deactivate functionalities or settings. In particular, the device may have several predetermined configurations or operating modes, allowing configuration by selecting one of these configurations.Operating modes can occur.

[0022] A first aspect of the proposed solution concerns a method for peak load shaving of a consumer's power consumption via a grid connection point of an electricity network, wherein the method has:

[0023] (i) Recording the temporal progression of a consumer's power consumption via the grid connection point;

[0024] (ii) Determining a time course of an average benefit receipt from the recorded time course of benefit receipt;

[0025] (iii) Determining a time profile of electrical compensation power to be drawn from an energy storage device assigned to the consumer as a function of a specific threshold for power consumption and both the time profile of the power consumption and the time profile of the mean value; and

[0026] (iv) Arranging for at least partial compensation of the power consumption by drawing electrical compensation power from the energy storage system, bypassing the grid connection point, according to its specified time profile. 120639P1488PC

[0027] According to the procedure, peak shaving using compensation power can therefore occur not only when the instantaneous value exceeds the threshold, but also when the average value exceeds the threshold, but the instantaneous value does not, or no longer does. This increases the scope for providing compensation power, which promotes a reduction in the average power consumption over the entire time interval and thus compliance with the target upper limit for power consumption via the grid connection point, as defined by the threshold, or at least a minimization of any exceedance of this target upper limit by the average power consumption. In this way, peak shaving can be optimized compared to the aforementioned classic peak shaving. This, in turn, can be used economically to reduce the average peak-related portion of the consumer's electricity bill.

[0028] The compensation power is drawn from the energy storage system, bypassing the grid connection point (commonly referred to as "behind the meter"). This means that the compensation power is not part of the grid consumption or recorded as such, as would be the case if the energy storage system were connected to the grid. Typically, the energy storage system is located at the consumer's premises or is integrated into them, so the compensation power is made available to the consumer directly. The specified compensation power initially represents a target power output from the energy storage system. If and as long as the energy storage system is capable of delivering the target power, this corresponds to the actual compensation power used; otherwise, the latter is lower than the target power output.

[0029] The following describes various exemplary embodiments of the method, which, unless expressly excluded or technically impossible, can be combined with each other and with the other aspects of the present solution described below.

[0030] According to one embodiment, the average power consumption is determined as a moving average. If the temporal profile of the power consumption is given by a sequence of instantaneous values ​​at successive points in time, then, depending on the point in time for which the moving average is determined, the instantaneous values ​​used for this purpose can be derived entirely or partially from the preceding time interval or from the current time interval (in which the point in time lies) from a defined sequence of such time intervals of equal length. These time intervals expediently correspond to the electricity supplier's billing intervals, for which peak shaving is aimed at limiting the average power consumption occurring at the grid connection point in the respective time interval.The length of the recording period relevant for averaging the instantaneous values ​​over the course of the network reference can be chosen in particular so that it does not exceed the length of the time interval, so that either (i) only instantaneous values ​​from the current time interval or (ii) only instantaneous values ​​from the current time interval and the immediately preceding time interval are used for averaging.

[0031] The average value thus evolves over time during the time interval and is based on the actual course of the power consumption (grid consumption) characterized by the sequence of recorded instantaneous values. Unlike the grid consumption, however, the time course of the average value is subject to less abrupt fluctuations and thus smooths the latter. Accordingly, the time course of the compensation power to be drawn from the energy storage system can also be smoothed, which in particular allows the use of slower-reacting energy storage systems that would not be able to closely follow the time fluctuations of the grid consumption.

[0032] According to one embodiment, a sequence of time intervals is defined along the temporal course of the power consumption, for which, within the framework of the method, a peak load shaving is carried out in each case with the aim of keeping the average grid consumption for the time interval at or below the threshold value on the basis of the compensation power taken from the energy storage in the respective time interval; and the respective average value of the power consumption is determined as an average value over a sequence of instantaneous values ​​of the power consumption, all of which are assigned to the same time interval.

[0033] Accordingly, the averaging process restarts in each time interval without including instantaneous values ​​from any other time interval, particularly the immediately preceding one. The start (restart) of the averaging for a given time interval can be synchronized with its beginning, for example, by starting with the first instantaneous value recorded for that interval, or only after a defined time period or a defined number of instantaneous values ​​have been recorded within that interval. The averaging process is thus decoupled from the grid consumption pattern in previous time intervals. This can be particularly advantageous for optimized peak shaving when the grid consumption pattern exhibits different phases with significantly varying average grid consumption.This can be the case, for example, when an industrial plant is gradually ramped up or shut down as a consumer, with the stages having different energy requirements.

[0034] According to one embodiment, the temporal profile of the compensation power is determined as a function of both the temporal profile of any deviation of the instantaneous value of the power consumption from the threshold value and the temporal profile of any deviation of the mean value of the power consumption from the threshold value. Considering a given point in time along these temporal profiles, the compensation power to be drawn in the future can thus be determined as a function of both any deviation of the instantaneous value of the power consumption at that point in time from the threshold value and any deviation of the mean value of the power consumption from the threshold value. This represents a particularly simple way to implement the method. In particular, the implementation of the method can be achieved using simple logic (e.g.,...).Logic circuitry) without the need for microprocessors, microcontrollers, or complex computer programs. 120639P1488PC.

[0035] According to one embodiment, the compensation power is determined over time as the larger of the two deviations or as a functional dependency on this larger deviation. Thus, simple difference calculations and a simple comparison of the determined deviations are sufficient to determine the compensation power as a function of the instantaneous value and the average value.

[0036] According to one embodiment, when determining the compensation power over time, one of the two deviations is weighted more heavily than the other using a weighting factor k. In particular, the deviation of the mean value from the threshold can be weighted more heavily than the deviation of the instantaneous value from the threshold, thereby further enhancing the effect of the mean-value-based peak shaving, which is an additional measure compared to classic peak shaving. In many cases of grid consumption patterns, this can further increase the total compensation power provided for peak shaving, thus achieving more effective peak shaving. This is especially true for grid consumption patterns where the grid consumption varies significantly over time, so that the smoothing effect of the mean-value-based component of the compensation power determination is amplified.

[0037] According to one embodiment, the weighting factor k is defined as a time-dependent quantity k(t). In this case, the weighting factor k(t) is not a time-invariant quantity, but rather a time-dependent quantity that can be used to vary the contribution of the mean-based component to the overall compensation power calculation over time, particularly in a different temporal dependency than that of the grid reference, i.e., the sequence of instantaneous values. This allows the inertia in the contribution caused by the deviation from the mean to be reduced.This can be used, for example, to systematically increase the contribution as the time interval progresses, so that deviations from the mean value gain more and more weight compared to deviations from the instantaneous value, and towards the end of the time interval, even in the case of low instantaneous values, can still contribute to a significant reduction in the mean network reference value for the entire time interval. The time course of k(t) can, in particular, be chosen to be the same in each time interval.

[0038] In particular, according to one embodiment, the time dependence of the weighting factor k(t), especially for each time interval, can be defined by a time-dependent, monotonically increasing function of one of the following function types or by a combination of at least two of these function types:

[0039] - linear function,

[0040] - Step function with a finite number of steps,

[0041] - A non-linear, monotonically increasing function with positive or negative curvature. Therefore, k(t) is defined here due to its monotonicity such that the weighting factor reaches its largest value towards the end of the time interval. In particular, the monotonicity of the function can be strict monotonicity not only in the case of a linear function (straight line), but also in the case of a non-linear function.

[0042] According to one embodiment, the threshold is defined as a time-dependent threshold function Tr(t), particularly for each time interval. When using a constant threshold T, power peaks at the beginning of peak shaving, for example in the first part of the time interval, are generally compensated very well. In contrast, power peaks that occur later, for example towards the end of the time interval, cannot always be compensated with a constant threshold T. By introducing a time dependency of the threshold based on the threshold function Tr(t), the grid consumption can be overcompensated at the beginning. At the end of the time interval, the value of the variable threshold Tr(t) can then correspond, in particular, to the target value for the upper limit of the grid consumption, globally or individually for the respective time interval.Overcompensation at the beginning can help keep the average grid consumption, for example over the entire time interval, at or below the target value if peak loads only occur towards the end, when otherwise there might be little or even too little time for adequate compensation. 120639P1488PC.

[0043] According to one embodiment, the time dependence of the threshold function is defined, in particular for each time interval, by a time-dependent, monotonically increasing function of one of the following function types or by a combination of at least two of these function types:

[0044] - linear function,

[0045] - Step function with a finite number of steps,

[0046] - non-linear, monotonically increasing function with positive or negative curvature.

[0047] While a linear function allows for a moderate adjustment of the threshold value, and a step function, due to its many degrees of freedom but simple definition, offers high flexibility in terms of adjustment, non-linear functions are well suited for defining continuously progressive adjustments of the threshold value that can be effectively adapted to typical grid reference profiles for a given consumer. In particular, choosing a monotonically increasing function with positive curvature (i.e., the second derivative is positive) is especially suitable for building up a large compensation buffer at the beginning of the time interval, in order to effectively compensate for any load peaks later in the interval.

[0048] A second aspect of the proposed solution concerns a device for peak shaving of a consumer's power consumption via a grid connection point of an electricity network. The device includes a control unit configured to execute the procedure described in the first aspect, thereby determining the electrical compensation power and drawing it from the energy storage system, bypassing the grid connection point.

[0049] All statements regarding the procedure according to the first aspect apply accordingly to the device according to the second aspect.

[0050] According to one embodiment, the device further includes a measuring device for recording the instantaneous values ​​of the power consumption by the consumer via the grid connection point. 120639P1488PC

[0051] According to one embodiment, the device also includes the energy storage device.

[0052] According to one embodiment, the device further comprises an energy generator or energy converter for providing energy to charge the energy storage device and / or for feeding energy into the power grid via the grid connection point.

[0053] The device can therefore include not only the control unit but also one or more additional components that contribute to implementing peak load shaving. This allows for the creation of an integrated system with this functionality.

[0054] The method and the device are described herein, in particular, in connection with the compensation of peak loads during positive grid draw, i.e., when drawing electrical power from the grid. However, it is equally possible to use the method and / or the device for compensating peak loads during negative grid draw, i.e., when feeding electrical power into the grid via the grid connection point. The power to be fed in can be provided, in particular, by an energy supplier of the consumer, such as a photovoltaic system and / or an energy storage system, for example, within the framework of a virtual power plant.

[0055] Further advantages, features and application possibilities of the present solution will become apparent from the following detailed description in the context of the figures.

[0056] This shows:

[0057] Fig. 1 shows an exemplary embodiment of a consumer with a device for peak load shaving,

[0058] Fig. 2 shows an example of a grid consumption curve and threshold for peak shaving, 120639P1488PC

[0059] Fig. 3 shows a consideration of the network usage according to the division into time intervals (billing intervals),

[0060] Fig. 4 shows an exemplary curve of grid consumption with limited power of the energy storage system,

[0061] Fig. 5 shows an exemplary curve of a grid consumption with overcompensation from the energy storage system,

[0062] Fig. 6 shows exemplary curves of a network reference and an associated moving average,

[0063] Fig. 7 shows exemplary curves of a network reference and an associated time interval-related mean value.

[0064] Fig. 8 shows an exemplary scenario for peak shaving with curves of a grid draw, an associated moving average Gav, the consumption at the load of the consumer and a constant threshold value,

[0065] Fig. 9 shows various exemplary courses of the time-dependent

[0066] Weighting factor k(t),

[0067] Fig. 10 shows various exemplary courses of the time-dependent

[0068] Threshold function Tr(t) and

[0069] Fig. 11 shows an exemplary scenario for peak load shaving with

[0070] The curves of a grid reference, an associated moving average Gav, the consumption at the load of the consumer and a time-dependent threshold function.

[0071] Figure 1 schematically illustrates an exemplary embodiment of a consumer 1. The consumer 1 is connected to a power grid 3 via a grid connection point 2 in order to draw electrical energy from it. It can also feed surplus electrical energy into the power grid 3 via the grid connection point 2. 120639P1488PC

[0072] Consumer 1 has at least one electrical load 4 configured to receive and consume electrical power 1. Examples of such loads 4 are systems or devices that require electrical energy for their operation. Furthermore, consumer 1 has an energy storage device 5 capable of receiving and releasing energy as electrical energy. An example of this is an electrochemical energy storage device 5, such as a rechargeable battery (accumulator), for example, a lithium-ion battery.

[0073] The energy storage device 5 is electrically connected to the grid connection point 2 via a current-limiting valve 6, such as a circuit breaker. For controlling the current-limiting valve 6, the load 1 includes a control unit 8, which can be designed as hard-wired logic or microprocessor-based with associated programming. The function of the control unit 8 will be explained in more detail below with reference to various exemplary embodiments of the solution method.

[0074] Furthermore, the consumer 1 optionally also includes an energy converter, such as a photovoltaic system, as an energy supplier 9 for electrical energy. The load 4 and the energy supplier 9 are also electrically connected to the grid connection point 2, in particular via one or more power lines. Finally, the consumer 1 also contains a measuring device 10 for determining the instantaneous power (instantaneous values ​​11 of the grid consumption 12) drawn from the power grid 3 via the grid connection point 2. This measuring device is connected to the control unit via a signal to transmit the recorded instantaneous values ​​11. Due to its function, the control unit can be described as a peak shaving device, either alone or in combination with the energy storage device 5 and / or the energy supplier 9.

[0075] Figures 2 and 3, with reference to Figure 1, illustrate the principle of classic peak shaving using an exemplary time profile of grid consumption 12 ("consumption"), which expresses the electrical power P drawn by a consumer 1 via an assigned grid connection point 2 120639P1488PC. A constant threshold value T is defined here as a target value in the sense of an upper limit for the average grid consumption 13, determined as a mean value over a defined time interval (billing interval). In the present example, the threshold value T is at an electrical power of P = 60 kW. The time interval can be, for example, 15 minutes, as illustrated in Figure 3, and is usually defined by the electricity supplier delivering electricity via the grid 3 as the billing interval and contractually agreed upon with the electricity customer.

[0076] Classic peak shaving takes place in the consumer 1 by the control unit 7 whenever the instantaneous value 11 of the current grid consumption 12 transmitted by the measuring device 10 is above the threshold value T, the flow control unit 7 controls the flow control valve 6 so that electrical compensation power is supplied from the energy storage device 5 (if available) in order to at least partially meet the instantaneous energy demand of the consumer 1 and thus reduce the current grid consumption 12 to or below the threshold value T, or to keep it there, as far as possible. In Figure 2, the compensation power drawn from the energy storage device 5, shown as black areas below the curve of the grid consumption 12, is set to its maximum; that is, it fills the entire area below the curve and simultaneously above the threshold value T.

[0077] However, it is also conceivable that the compensation power is chosen to be lower and thus only partially fills the area above the threshold T below the grid consumption curve 12. Depending on the profile of the grid consumption 12, this may still be sufficient to keep the average grid consumption 13 at or below the threshold T. The overall goal is to limit the average grid consumption 13 per time interval as much as possible to the threshold T as the upper limit for the average grid consumption 13. The areas above the grid consumption 12 and simultaneously below the threshold T represent charging energy with which the energy storage device 5 is charged. This charging energy can then be made available again in other sections of the grid consumption curve 12 that lie above the threshold T to provide compensation power. 120639P1488PC

[0078] In particular, the compensation power can be adjusted so that the resulting average grid reference 13 corresponds to the threshold value T, as shown in Fig. 2.

[0079] In order to determine the storage power B within the framework of classical peak shaving (in the case of discharge it corresponds to the compensation power), in particular an iterative procedure can be used in which, in a sequence of steps, each of which concerns an assigned short time period in the time interval, the power of the energy storage 5 is determined for the assigned time period; the time periods can, for example, each be one second long.

[0080] The change in the power output of the energy storage device 5 for the time interval between measurement point n for the current instantaneous value 11 (at step Sn) and the next measurement point n+1 (at step Sn+1) of the grid consumption 12 corresponds here to the difference between the grid consumption 12 corresponding to the respective step and the threshold value T. If the grid consumption 12 is above the threshold value T, the energy storage device 5 is discharged and the compensation power is provided from it; if, on the other hand, the grid consumption 12 is below the threshold value T, the energy storage device 5 is charged so that it can later provide compensation power again. The power output of the energy storage device 5 is determined at the shortest possible intervals, e.g., at intervals of seconds.

[0081] In a steady, steady state, the following applies (see Fig. 1):

[0082] Gn = Cn - En + Bn. (1 )

[0083] G: Power at grid connection point 2 (G < 0: feed-in, G > 0: grid consumption) C: Consumption 1 (C > 0)

[0084] E: Production of energy supplier 9, e.g. PV generation (E > 0)

[0085] B: Energy storage capacity 5 (“storage capacity” B; B < 0: discharge, B > 0: charge) 120639P1488PC

[0086] In a typical configuration, power measurement is performed directly at the grid connection point 2 using the measuring device 10. However, other configurations are possible, e.g., power measurements at the consumer 1, at the energy supplier 9, or at the energy storage device 5. The power at the grid connection point 2 can then be calculated using the relationship (1). However, this does not affect the principle of peak shaving.

[0087] An example of the aforementioned iterative procedure is explained in detail below, where an arbitrary step Sn+i in the sequence of iterative steps Sl ,... , Sn-1 , Sn, Sn+1 , ... with n > 1 is considered:

[0088] Starting from the current state of consumer 1 in the preceding step Sn (or from an initial state of consumer 1 if the first step Si of the sequence is considered), the change B to be initiated in the next step Sn+i is first determined. n +i of the storage power B is calculated: Bn+1 = T - Gn (2)

[0089] The symbols used here have the following meanings:

[0090] T: Threshold (T > 0)

[0091] G n : Instantaneous value 11 of the power at grid connection point 2 in step Sn AB n +i : Change in storage performance B relative to the previous step Sn

[0092] Bmax: maximum storage power available as compensation power B) Bn: storage power B (B<0: discharged, therefore B=compensation power) in step Sn

[0093] B1 n +i : Target value of the storage power B (compensation power) based on Gn n : Index to the considered step Sn n+1 : Index to the next step following Sn Sn+i

[0094] From this, the target value B can be determined. n +i for memory performance B is determined as follows: 120639P1488PC

[0095] Substituting relationship (1) into (2) yields:

[0096] ABn+1 = T - Cn + P 'n ~ Bn (3)

[0097] ABn+1 + B n = T - Cn + Pn (4)

[0098] B n +1 = Bn+1 + Bn — Bmax (5)

[0099] Figures 3 to 5 illustrate various scenarios within the framework of classic peak shaving, which can lead to unsatisfactory results.

[0100] In the example shown in Fig. 3, the grid reference curve "Reference 1" refers to a first time interval 14 and the grid reference curve "Reference 2" to a second time interval 15. The reference is shown here without compensation power from the energy storage system 5. However, if the energy consumption of consumer 1 in the respective time interval is too high to be adequately compensated by energy from the energy storage system 5 using classic peak shaving, a first average grid reference 13a for the first time interval 14 results, which lies above the threshold T and can therefore trigger additional electricity costs if it cannot be adequately compensated by peak shaving. In contrast, the second average grid reference 13b for the second time interval 15 is already below the threshold T even without compensation and would not need to be compensated to comply with the threshold T.

[0101] Figure 4 illustrates another scenario in which the compensation power from the energy storage unit 5 is insufficient to fully compensate for even large load peaks. Rather, the maximum compensation power Bmax is below at least one load peak in the grid consumption 12, so that a contribution to increasing the average grid consumption 13 remains, which, depending on its magnitude, may be sufficient to raise the average grid consumption 13 above the threshold value T over the associated time interval. 120639P1488PC

[0102] Figure 5 illustrates a further scenario in which, conversely, short power peaks are fully compensated, but this unnecessarily reduces the average grid consumption 13 for the corresponding time interval. In particular, the situation can arise that the average grid consumption 13 would already be below the threshold T without compensation power, and it is unnecessarily reduced even further by means of the compensation power, thereby unnecessarily burdening the energy storage 5.

[0103] Now, with reference to the device shown in Fig. 1, exemplary embodiments of the method according to the first aspect are explained. The following example builds upon and modifies the iterative process previously described for classical peak shaving.

[0104] For this purpose, a step Sn is considered again, in which the grid consumption 12 is higher than the threshold value T and thus a compensation power is taken from the energy storage 5 to compensate for the difference grid consumption 12 - threshold value T.

[0105] To calculate the target value B to be determined n +i for the compensation performance, a first component B1 is used. n +i is determined in the same way as previously described for classical peak shaving according to relations (1) to (5). Thus, according to relation (5):

[0106] B1n+1 = ABn+7 + Bn < Bmax with B n +1 = T - Gn (6)

[0107] In contrast to classic peak shaving, however, a second component B2 is now added. n +i in the calculation of the target value B n +i is added, which is determined on the basis of the course of the mean value of the power of the grid consumption 12.

[0108] B2n+1 =BdVn+1 Bn — Bmax with BdVn+1 — T - GdVn (7) 120639P1488PC

[0109] The newly added symbols used here have the following meaning:

[0110] Gav n : average power consumption at grid connection point 2 in step Sn;

[0111] ABavn+7: Change in storage performance B relative to the previous step n, based on Gavn;

[0112] B2 n +i : Target value of the compensation power based on averaged power values.

[0113] Various methods can be used to calculate the average, i.e., to determine Gavn. Fig. 6 illustrates an example of the behavior of a network reference 12 with the determination of a moving average Gavn of the network reference 12 as a basis for determining the setpoint B2n+1 according to equation (7). The average for step Sn can be determined, in particular, by averaging, e.g., arithmetic averaging, over the m preceding instantaneous values ​​Gn-m, Gn-m+1, ..., Gn-1. This method of calculating the average Gavn n is independent of the time interval boundaries and therefore avoids jumps caused by the time interval boundaries (represented as vertical dashed lines).

[0114] Fig. 7 illustrates, according to an embodiment of the method, an exemplary course of a network reference 12 with a new calculation of the mean value Gav for each time interval. nThe grid reference 12 serves as a basis for determining the compensation power within the framework of peak shaving. The mean value for step Sn can be determined, in particular, by averaging, e.g., arithmetic averaging, over all preceding instantaneous values ​​11 in the same time interval. In the present example in Fig. 7, the length of the time interval is 15 time units (e.g., minutes), and a new averaging process begins after each 15 time units have elapsed. With this method of calculation for the mean value Gav n By definition, there is a dependency on the time interval boundaries, which can often lead to jumps or sharp changes in the slope of the mean curve. 120639P1488PC

[0115] Now, from the two calculated target values ​​B1 n +1 and B2 n +i , which were initially determined as intermediate results, the determination of the target power B n+i of the energy storage unit 5 (compensation power) will be performed.

[0116] The following methods, in particular, can be used for this purpose: a) Minimum value method

[0117] In this case, the smaller of the two target values ​​is used (It should be noted that when providing the compensation power, by definition the values ​​of B1 n+1 and B2n+1 and therefore also the resulting value Bn+1 are each negative):

[0118] B n +i = Min(B1 n +i, B2 n +i) (8)

[0119] Since the value Bn is added in both cases when calculating the target values ​​B1n+1 and B2n+1, the following relationship can also be used to determine Bn+1:

[0120] Bn+i = Min( B n +i, ABav n +i) (9)

[0121] Fig. 8 illustrates an exemplary scenario with curves of a grid draw 12, an associated moving average Gav, the consumption 1 at the load 4 of consumer 1, and a constant threshold T. The time interval is again chosen here as an example of a length of 15 time units, e.g., minutes. The counting of the time units, which can correspond in particular to the steps Sn, starts again in each time interval. With regard to the device from Fig. 1, it should be noted that in this scenario the energy supplier 9 is not present or at least does not participate.

[0122] In the time range between t = 0 and t = 9, the grid consumption 12 remains almost constant and coincides with the consumption 1. The average grid consumption 12 largely follows these patterns, and all three patterns are below the threshold value T. The power drawn from the grid 3 is therefore completely consumed in load 4 1, and the energy storage 5 neither supplies power nor absorbs any.

[0123] From t = 9, the consumption 1 and the grid feed-in 12 increase simultaneously and steeply up to the threshold value T. The moving average, on the other hand, is smoother and flatter. Since the threshold value T has not yet been exceeded, no peak shaving occurs by drawing power from the energy storage 5. This holds true until the end of the first time interval 14 at t = 15.

[0124] However, this changes in the subsequent second time interval 15 at t = 3 (new count of t), where consumption 1 and, simultaneously, grid consumption 12 exceed the threshold value T, thus triggering peak shaving by drawing compensation power from the energy storage system 5. Until t = 8, when consumption 1 falls back to the threshold value T, the (instantaneous) grid consumption 12 is above both the average value and the threshold value T, so that the first (smaller) setpoint B1 is used here to determine the compensation power. n +1 dominates, while the (larger) second setpoint B2 n +i does not come into play. At t = 9 and t = 10, the mean value has already exceeded the threshold T, but it is still below the network reference 12, so the first setpoint B1 n+1 also dominates here. However, at t = 11 this changes, so that the mean value, which is still above the threshold T, is now higher than the instantaneous grid reference 12, which has since fallen, and thus the (now smaller) second setpoint B2. n +i instead of B1 n +i comes into play. Towards the end of the second time interval 15, peak shaving continues to reduce the average power consumption for the entire time interval, even though the current grid consumption 12 has already fallen below the threshold T. b) Constant weighting factor for the average-dependent setpoint

[0125] The minimum value method from section a) is, depending on the course of the network reference 12, often still relatively sluggish, which affects the second setpoint B2 n +i is particularly relevant in scenarios such as the one in Fig. 8, where the network reference 12 increases over time and B2 n+i the network reference 12 then lags behind. 120639P1488PC

[0126] However, this can lead to a situation where, especially towards the end of the time interval, even if the second target value B2 is reached, n +i comes into play, its contribution to peak load reduction is too small to reduce the average grid consumption 13 Bav over the entire time interval to the threshold T.

[0127] This can be countered by calculating B2 n The function +i is extended by introducing a weighting factor k. If k is chosen to be constant over the time interval, then k > 1.

[0128] Relationship (7) can then be modified as follows:

[0129] B2 n +i = k *ABav n + 1 + Bn - Bmax (10)

[0130] For determining the target power of the energy storage unit 5 (compensation power) from B1 n +i and B2 n+i, relationship (9) can be used again. c) Time-dependent weighting factor for the mean-dependent target value

[0131] The deviation of the average power at the grid connection (grid consumption 12) becomes increasingly important towards the end of the interval in order to achieve effective peak shaving. Therefore, in this embodiment, the weighting factor k is defined as time-dependent, i.e., k = k(t) or, in the iterative case, equivalently k = k(n).

[0132] The relationship (10) can then be modified as follows:

[0133] B2n+1 = k(t) *ABAVn+1 + Bn — Bmax (11 )

[0134] Here, k(t) > 0 is conveniently defined such that the value of k(t) is above average, and in particular maximal, towards the end of the time interval.

[0135] Figure 9 shows four different exemplary curves of k(t) within a time interval. A first curve, k1(t), is strictly monotonically increasing linear, and two further curves, k2(t) and k3(t), are each non-linear and strictly monotonic, with the second curve, k2(t), having positive curvature (i.e., a second derivative), and a third curve, k3(t), having negative curvature. The initial value K1 of k(0) at the beginning of the time interval is thus always smaller than the final value K2 = k(t=15) at the end of the time interval, where, in this example, the time interval is assumed to be 15 minutes.

[0136] An exemplary fourth curve k4(t) is monotonically increasing in steps, with a finite number of steps. For the parameters K1 and K2, K2 > K1 > 0, and in particular K2 > 1. K1 > 1 is also possible.

[0137] It is also conceivable that the behavior of k(t) varies across sections, for example, initially linear and then arc-shaped or step-like, etc. Even a break in the monotonicity is fundamentally possible, meaning that declining curve segments are not excluded. However, ascending curve profiles will generally lead to more effective solutions in terms of improvements over classical peak shaving or the aforementioned methods with constant k(t) = k = const.

[0138] In the embodiments described above, power peaks in the first part of the time interval are generally very well compensated. Power peaks that occur only towards the end of the billing interval cannot always be fully compensated.

[0139] This problem can be addressed by introducing a time-dependent threshold function Tr(t) instead of a constant threshold T, thereby overcompensating for the power at the grid connection at the beginning of the time interval. Here, Tr(t) > 0 is expediently defined such that the value of Tr(t) is above average, and in particular at its maximum, towards the end of the time interval. Thus, as illustrated, the value of Tr(t) at the end of the time interval can correspond to a threshold T defined as the target value of the average grid consumption, e.g., Tr(15min) = T.

[0140] As illustrated in Fig. 10, the behavior of the threshold function Tr(t) can be implemented, in particular, as a linear function Tr1(t), as a non-linear function Tr2(t) with positive curvature, or as a non-linear function with negative curvature. In particular, the behavior Tr4(t) of Tr(t) is also possible, which increases monotonically in steps with a finite number of steps. With the initial threshold TrO at the beginning of the time interval, Tr(t > 0) > TrO > 0.

[0141] It is also conceivable that the behavior of Tr(t) varies across sections, for example, initially linear and then arc-shaped or step-like, etc. Even a break in the monotonicity is fundamentally possible, meaning that downward-sloping curve sections are not excluded. However, upward-sloping curves will generally lead to more effective solutions in terms of improvements over classical peak shaving or the aforementioned methods with constant Tr(t) = T = const.

[0142] Fig. 11 shows an exemplary scenario with a variable threshold function Tr(t). In a first time interval 14, from t = 0 to t = 9, the consumption 1 by the load 4 is below the grid consumption 12, and the grid consumption 12 is below the threshold function Tr(t). Therefore, the energy storage 5 is charged, and no compensation power is drawn, as there are no load peaks to reduce. However, this changes at t = 10, where both the consumption 1 and the grid consumption 12 exceed the value of the threshold function Tr(t), resulting in a load peak that lasts until t = 14 and is compensated by drawing compensation power from the energy storage 5. Between t = 14 and the end of the first time interval 14 at t = 15, the consumption 1 and the grid reference 12 follow the course of the threshold function Tr(t), so that no compensation takes place here.During the entire first time interval 14, the mean grid consumption 13 is below the instantaneous grid consumption 12, so that where peak shaving is carried out by compensation, the instantaneous grid consumption 12 is used instead of the mean grid consumption 13 according to relationship (8) or (9).

[0143] From the beginning of the second time interval 15, the threshold function Tr(t) decreases, thus resulting in compensation, even though consumption 1 and grid reference 12 initially remain constant and stay above the threshold function Tr(t) until t = 6. Furthermore, consumption 1 exceeds grid reference 120639P1488PC from t = 2 onwards.

[0144] 12, so that consumption 1 must be partially drawn from energy storage 5 or energy supplier 9 for this reason alone. At t = 5, the average grid consumption 13 exceeds the instantaneous grid consumption 12 and therefore now takes its place until time t = 9 according to equation (8) or (9). From t = 10 onwards, both the instantaneous grid consumption 12 and the average grid consumption 13 are below the threshold function Tr(t), so that no further compensation is required until the end of the second time interval 15. However, since consumption 1 remains above grid consumption 12, it must continue to be partially drawn from energy storage 5 or energy supplier 9.

[0145] In the second time interval 15, the average grid consumption 13 was thus used for determining and retrieving the compensation power from the energy storage system 5 in the time range t = 5 to t = 9, whereas no compensation would have taken place in this time range with conventional peak shaving. Overall, this resulted in a contribution to more effective peak shaving.

[0146] While at least one exemplary embodiment has been described above, it should be noted that a large number of variations exist. It should also be noted that the described exemplary embodiments are merely non-limiting examples, and it is not intended to restrict the scope, applicability, or configuration of the devices and methods described herein. Rather, the preceding description will provide the person skilled in the art with guidance for implementing at least one exemplary embodiment. It is understood that various modifications to the function and arrangement of the elements described in an exemplary embodiment can be made without deviating from the subject matter defined in the appended claims. 120639P1488PC

[0147] Reference symbol

[0148] 1 Consumer 2 Grid connection point 3 Power grid 4 Load 5 Energy storage 6 Flow valve 7 Control 8 Control device 9 Energy supplier 10 Measuring device 11 Instantaneous values ​​12 Grid consumption 13 Average grid consumption 13a First average grid consumption 13b Second average grid consumption

[0149] 14 first time interval 15 second time interval

[0150] B Storage capacity

[0151] Bmax maximum compensation power k1 (t) first curve weighting factor k2(t) second curve weighting factor k3(t) third curve weighting factor k4(t) fourth curve weighting factor T threshold Tr(t) threshold function Tr1 (t) first curve threshold function Tr2(t) second curve threshold function Tr3(t) third curve threshold function Tr4(t) fourth curve threshold function

Claims

120639P1488PC Patent claims 1. Method for peak load shaving of a consumer's power consumption (1 ) via a grid connection point (2) of an electricity grid (3), wherein the method comprises: Recording a time course of the power consumption of the consumer (1) via the grid connection point (2); Determining a time course of an average benefit receipt from the recorded time course of benefit receipt; Determining a time course of an electrical compensation power to be drawn from an energy storage device (5) assigned to the consumer (1) as a function of a certain threshold value (T) for the power draw and both the time course of the power draw and the time course of the mean value; and Arranging for at least partial compensation of the power consumption by taking the electrical compensation power from the energy storage (5) bypassing the grid connection point (2) according to its specific temporal profile.

2. Method according to claim 1, characterized in that the mean value of the performance reference is determined as a moving mean value of the performance reference.

3. Method according to one of the preceding claims, characterized in that a sequence of time intervals is defined or is defined along the temporal course of the power consumption, for which, within the framework of the method, peak shaving is carried out with the aim of keeping the average grid consumption (13) for the time interval at or below the threshold value (T) on the basis of the compensation power withdrawn from the energy storage (5) in the respective time interval; and 120639P1488PC the respective mean value of the benefit is determined as the mean value over a sequence of instantaneous values ​​(11 ) of the benefit, all of which are assigned to the same time interval.

4. Method according to one of the preceding claims, characterized in that the temporal course of the compensation performance is determined as a function of both the temporal course of a deviation of the instantaneous value (11 ) of the power reference from the threshold value (T) and the temporal course of a deviation of the mean value of the power reference from the threshold value (T).

5. Method according to claim 4, characterized in that, over time, the compensation performance is determined as the larger of the two deviations or in functional dependence on this larger deviation.

6. Method according to claim 4 or 5, characterized in that, in the course of time, when determining the compensation performance, one of the two deviations is weighted more heavily than the other deviation using a weighting factor k.

7. Method according to claim 6, characterized in that the weighting factor k is defined as a time-dependent quantity k(t).

8. Method according to claim 7, characterized in that the time dependence of the weighting factor k(t) is defined by a time-dependent, monotonically increasing function of one of the following function types or by a combination of at least two of these function types: - linear function, - Step function with a finite number of steps, - non-linear, monotonically increasing function with positive or negative curvature. 120639P1488PC 9. Method according to one of claims 4 to 8, characterized in that the threshold (T) is defined as a time-dependent threshold function Tr(t).

10. Method according to claim 9, characterized in that the time dependence of the threshold function (Tr(t)) is defined by a time-dependent, monotonically increasing function of one of the following function types or by a combination of at least two of these function types: - linear function, - Step function with a finite number of steps, - non-linear, monotonically increasing function with positive or negative curvature.

11. Device for peak load shaving of a consumer's (1) power consumption via a grid connection point (2) of a power grid (3), wherein the device has a control unit (8) configured to perform the method according to one of the preceding claims and to determine the electrical compensation power according to the method and to initiate its consumption from the energy storage device (5) bypassing the grid connection point (2).

12. Device according to claim 11, further comprising a measuring device (10) for recording the instantaneous values ​​(11) of the power consumption of the consumer (1) via the network connection point (2).

13. Device according to claim 11 or 12, further comprising the energy storage device (5).

14. Device according to one of claims 11 to 13, further comprising an energy generator or energy converter for providing energy for charging the energy storage device (5) and / or for feeding energy into the power grid (3) via the grid connection point (2).

Images